Handbook of Behavioural Change and Public Policy 9781785367854, 9781785367847

Table of contents :
Front Mattter
Copyright
Contents
List of contributors
1 Introduction to the Handbook of Behavioural Change and Public Policy
Part I: Knowing and changing behaviour: The co-evolvment of behavioural science and behavioural public policy
2 Nudging before the nudge? Behavioural traffic safety regulation and the rise of behavioural economics
3 Economic expertise and public policy
4 A bibliometric analysis of behavioural studies in economics and public policy journals
5 The concepts of nudge and nudging in behavioural public policy
6 Behavioural considerations in public policy: matching policy tools and their targets
Part II: Nudging around the world: Proliferation and professionalization
7 Nudging around the world: a critical geography of the behaviour change agenda
8 The application of behavioural insights to policy in Europe
9 Behaviour experts in government: from newcomers to professionals?
10 Nudge and the European Union
11 Nudging citizens’ knowledge in knowledge-based EU: the case of breast cancer screening programmes and participatory rights in choice architectures
12 Behaviouralizing Europe: behavioural economics enters EU policy-making
Part III: Behavioural insights and instruments in policy-making
13 The enzymatic effect of behavioural sciences: what about policy-makers’ expectations?
14 The politics of nudge and framing behaviour change in health
15 Nudging to sustainability? Critical reflections on nudging from a theoretically informed sustainability perspective
16 Encouraging longer working lives: a behavioural perspective
17 Behavioural economics and development policy
18 Steering the behaviour of young people: the EU’s policy approach to promote employment
19 The governance of behavioural taxation: moralization and the new modes of tax collection
Part IV: The governance of behaviour: Normative ideals, critical perspectives and political consequences
20 Which nudges do people like? A national survey
21 Constitutional limits to regulation-by-nudging
22 Nudging: ethical and political dimensions of choice architectures
23 Macro libertarianism and micro paternalism: governance in an age of nudging
24 Behaviour change: extralegal, apolitical, scientistic?
Index

Citation preview

HANDBOOK OF BEHAVIOURAL CHANGE AND PUBLIC POLICY

M4716-STRAßHEIM_9781785367847_t.indd 1

25/04/2019 13:32

HANDBOOKS OF RESEARCH ON PUBLIC POLICY Series Editor: Frank Fischer, Rutgers University, New Jersey, USA The objective of this series is to publish Handbooks that offer comprehensive overviews of the very latest research within the key areas in the field of public policy. Under the guidance of the Series Editor, Frank Fischer, the aim is to produce prestigious high-quality works of lasting significance. Each Handbook will consist of original, peer-reviewed contributions by leading authorities, selected by an editor who is a recognized leader in the field. The emphasis is on the most important concepts and research as well as expanding debate and indicating the likely research agenda for the future. The Handbooks will aim to give a comprehensive overview of the debates and research positions in each key area of focus.   Titles in the series include: International Handbook on Ageing and Public Policy Edited by Sarah Harper and Kate Hamblin Handbook on Complexity and Public Policy Edited by Robert Geyer and Paul Cairney Handbook of Critical Policy Studies Edited by Frank Fischer, Douglas Torgerson, Anna Durnová and Michael Orsini Handbook of Public Policy Agenda Setting Edited by Nikolaos Zahariadis Handbook of Policy Formulation Edited by Michael Howlett and Ishani Mukherjee Handbook of European Policies Interpretive Approaches to the EU Edited by Hubert Heinelt and Sybille Münch Handbook on Participatory Governance Edited by Hubert Heinelt Research Handbook on Street-Level Bureaucracy The Ground Floor of Government in Context Edited by Peter Hupe Handbook of Behavioural Change and Public Policy Edited by Holger Straßheim and Silke Beck

M4716-STRAßHEIM_9781785367847_t.indd 2

25/04/2019 13:32

Handbook of Behavioural Change and Public Policy

Edited by

Holger Straßheim Bielefeld University, Germany

Silke Beck Helmholtz-Centre for Environmental Research – UFZ, Leipzig, Germany

HANDBOOKS OF RESEARCH ON PUBLIC POLICY

Cheltenham, UK • Northampton, MA, USA

M4716-STRAßHEIM_9781785367847_t.indd 3

25/04/2019 13:32

© Holger Straßheim and Silke Beck 2019 Cover image: Skye Studios on Unsplash All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical or photocopying, recording, or otherwise without the prior permission of the publisher. Published by Edward Elgar Publishing Limited The Lypiatts 15 Lansdown Road Cheltenham Glos GL50 2JA UK Edward Elgar Publishing, Inc. William Pratt House 9 Dewey Court Northampton Massachusetts 01060 USA

A catalogue record for this book is available from the British Library Library of Congress Control Number: 2018967820 This book is available electronically in the Social and Political Science subject collection DOI 10.4337/9781785367854

ISBN 978 1 78536 784 7 (cased) ISBN 978 1 78536 785 4 (eBook)

06

Typeset by Servis Filmsetting Ltd, Stockport, Cheshire

M4716-STRAßHEIM_9781785367847_t.indd 4

25/04/2019 13:32

Contents List of contributorsviii  1  Introduction to the Handbook of Behavioural Change and Public Policy Holger Straßheim and Silke Beck

1

 NOWING AND CHANGING BEHAVIOUR: THE PART I  K CO-EVOLVEMENT OF BEHAVIOURAL SCIENCE AND ­BEHAVIOURAL PUBLIC POLICY  2  Nudging before the nudge? Behavioural traffic safety regulation and the rise of behavioural economics Rüdiger Graf  3 Economic expertise and public policy Frédéric Lebaron

23 38

 4  A bibliometric analysis of behavioural studies in economics and public policy journals49 Stuti Rawat  5  The concepts of nudge and nudging in behavioural public policy Pelle Guldborg Hansen  6  Behavioural considerations in public policy: matching policy tools and their targets Michael Howlett

63

78

NUDGING AROUND THE WORLD: PROLIFERATION AND PART II  PROFESSIONALIZATION  7  Nudging around the world: a critical geography of the behaviour change agenda90 Mark Whitehead, Rhys Jones and Jessica Pykett  8  The application of behavioural insights to policy in Europe Emanuele Ciriolo, Joana Sousa Lourenço and Sara Rafael Almeida

102

 9 Behaviour experts in government: from newcomers to professionals? Joram Feitsma and Thomas Schillemans

122

10 Nudge and the European Union Alberto Alemanno

138

v

M4716-STRAßHEIM_9781785367847_t.indd 5

25/04/2019 13:32

vi  Handbook of behavioural change and public policy 11  Nudging citizens’ knowledge in knowledge-based EU: the case of breast cancer screening programmes and participatory rights in choice architectures148 Luca Leone and Mariachiara Tallacchini 12 Behaviouralizing Europe: behavioural economics enters EU policy-making P.W. (Peter-Wim) Zuidhof

163

PART III  BEHAVIOURAL INSIGHTS AND INSTRUMENTS IN POLICY-MAKING The enzymatic effect of behavioural sciences: what about policy-makers’ 13  ­expectations? Kathrin Loer 14 The politics of nudge and framing behaviour change in health Muireann Quigley and Anne-Maree Farrell Nudging to sustainability? Critical reflections on nudging from a 15  theoretically informed sustainability perspective Basil Bornemann and Paul Burger

180 195

209

16 Encouraging longer working lives: a behavioural perspective R. Kent Weaver

227

17 Behavioural economics and development policy Christian Berndt

242

18  Steering the behaviour of young people: the EU’s policy approach to promote employment Jale Tosun and Felix Hörisch

257

19  The governance of behavioural taxation: moralization and the new modes of tax collection Sebastian Botzem

272

PART IV  THE GOVERNANCE OF BEHAVIOUR: NORMATIVE IDEALS, CRITICAL PERSPECTIVES AND POLITICAL CONSEQUENCES 20 Which nudges do people like? A national survey Cass R. Sunstein

285

21 Constitutional limits to regulation-by-nudging Anne van Aaken

304

22 Nudging: ethical and political dimensions of choice architectures Mark D. White

319

23  Macro libertarianism and micro paternalism: governance in an age of nudging Roger Tyers

M4716-STRAßHEIM_9781785367847_t.indd 6

332

25/04/2019 13:32

Contents  vii 24 Behaviour change: extralegal, apolitical, scientistic? Robert Lepenies and Magdalena Małecka

344

Index361

M4716-STRAßHEIM_9781785367847_t.indd 7

25/04/2019 13:32

Contributors Anne van Aaken is Alexander von Humboldt Professor for Law and Economics, Legal Theory, Public International Law and European Law at the University of Hamburg, Germany. Alberto Alemanno is Jean Monnet Professor of EU Law and Global Clinical Professor at NYU School of Law. Sara Rafael Almeida is a Policy Analyst at the Foresight, Behavioural Insights and Design for Policy Unit, Joint Research Centre (European Commission). Silke Beck is Deputy Chair of the Department of Environmental Politics, HelmholtzCentre for Environmental Research – UFZ, Leipzig, Germany. Christian Berndt is Professor of Economic Geography at the University of Zurich. Basil Bornemann is a Post-doctoral Researcher at the Sustainability Research Group, Department of Social Sciences, University of Basel. Sebastian Botzem is a Visiting Professor of International Political Economy at the Free University Berlin and leader of the research group on Transnational Political Ordering in Global Finance at the University of Bremen, Germany. Paul Burger is Professor for Sustainability Research and Head of the Sustainability Research Group, Department of Social Sciences, University of Basel. Emanuele Ciriolo is a Behavioural Economist at the Foresight, Behavioural Insights and Design for Policy Unit, Joint Research Centre (European Commission). Anne-Maree Farrell is Professor of Law at Queens University Belfast, United Kingdom. Joram Feitsma (MSc.) is a PhD Candidate at Utrecht University School of Governance. Rüdiger Graf (PD Dr.) is Head of the Research Department on Economic History at the Center for Contemporary History at Potsdam. Pelle Guldborg Hansen is a Behavioural Scientist at Roskilde University, Denmark. Felix Hörisch (PD Dr.) is Lecturer for Political Science at the Institute of Political Science, Heidelberg University, Germany. Michael Howlett is Burnaby Mountain Chair in the Department of Political Science at Simon Fraser University in Vancouver, Canada and Professor in the Lee Kuan Yew School of Public Policy at the National University of Singapore. Rhys Jones is Professor of Political Geography at Aberystwyth University in Wales. Frédéric Lebaron is Professor at the Ecole normale supérieure Paris-Saclay and member of the Research Group Institutions et Dynamiques Historiques de l’Economie et des Sociétés (IDHE.S), ENS-Cachan. viii

M4716-STRAßHEIM_9781785367847_t.indd 8

25/04/2019 13:32

Contributors  ix Luca Leone is Research Fellow at the Università Cattolica del Sacro Cuore – Faculty of Economics and Law – Piacenza (Italy). Robert Lepenies is a Post-doctoral Research Scientist at the Helmholtz Centre for Environmental Research (UFZ) in the Department of Environmental Politics. Kathrin Loer is Research Fellow (Political Science) at the Fernuniversitaet in Hagen. Joana Sousa Lourenço is a Policy Analyst at the Foresight, Behavioural Insights and Design for Policy Unit, Joint Research Centre (European Commission). Magdalena Małecka is Academy of Finland Post-doctoral Researcher at TINT, Centre in the Philosophy of the Social Sciences, University of Helsinki. Jessica Pykett is a Senior Lecturer in Human Geography at the University of Birmingham. Muireann Quigley is Professor of Law, Medicine, and Technology at the University of Birmingham. Stuti Rawat is a PhD Candidate at the Lee Kuan Yew School of Public Policy, National University of Singapore. Thomas Schillemans is Professor of Public Governance with a focus on accountability, behavior and institutions at Utrecht University School of Governance. Holger Straßheim is Professor of Political Sociology at Bielefeld University and Guest Researcher at the Berlin Social Science Center. Cass R. Sunstein is the Robert Walmsley University Professor at Harvard University. Mariachiara Tallacchini is Professor of Philosophy of Law at the Università Cattolica del Sacro Cuore – Faculty of Economics and Law – Piacenza (Italy). Jale Tosun is Professor of Political Science at the Institute of Political Science, Heidelberg University, Germany. Roger Tyers is a Research Fellow in the Department of Sociology, Social Policy and Criminology at the University of Southampton. R. Kent Weaver is Professor of Public Policy and Government at Georgetown University and a Senior Fellow in Governance Studies at the Brookings Institution. Mark D. White is Chair of the Department of Philosophy at the College of Staten Island/ CUNY. Mark Whitehead is a Professor of Human Geography at Aberystwyth University in Wales. P.W. (Peter-Wim) Zuidhof is Associate Professor in European Political Economy at the Department of European Studies, University of Amsterdam, the Netherlands.

M4716-STRAßHEIM_9781785367847_t.indd 9

25/04/2019 13:32

M4716-STRAßHEIM_9781785367847_t.indd 10

25/04/2019 13:32

1.  Introduction to the Handbook of Behavioural Change and Public Policy Holger Straßheim and Silke Beck

Over the past decade behavioural public policy1 has spread inter- and transnationally (see the contributions in part II of this volume). Based on a survey conducted in 23 countries, the OECD (2017, p. 13) concluded that behavioural insights in public policy seem to be more than a ‘fashionable short-term foray’ and ‘have taken root in many ways across many countries around the world and across a wide range of sectors and policy areas’. The World Bank (2015), the European Commission (Lourenço et al. 2016) and the United Nations (2017) have all documented their efforts to implement behavioural insights and interventions in a wide variety of policy areas such as development, taxation, energy, mitigation of climate change, sustainable consumption, pensions, public health, employment, poverty, gender mainstreaming and anti-corruption policy. While the US, the UK and Singapore are among the forerunners, modes of behavioural public policy can nowadays be found in many countries across Western Europe, Central America, Asia, Africa and the Middle East. The motives and mechanisms behind this worldwide spread of behavioural interventions have been interpreted in different ways. Some observers are sceptical and compare the diffusion of behavioural instruments and regulations in a relatively short span of time to other fads and fashions of public sector reform such as the total quality movement in the 1980s or the new public management in the 1990s and early 2000s (Brunsson 2009). Some agree with the OECD (2015) in characterizing the use of behavioural approaches as a ‘paradigm shift’ in public policy, a change that is based on both the ‘behavioural revolution’ in economics and the social sciences and the capability of behavioural scientists to ‘translate’ their claims into easily understandable ideas and devices for public policy (John 2018). Others criticize the behavioural movement as the latest expression of neoliberal governmentality incorporating all the characteristics of a ‘post-democratic’ age such as scientization and technocracy, individualization and self-management, marketization and deregulation, depoliticization and economization (Jones et al. 2013; Whitehead et al. 2017). Behavioural public policy is not simply a tool. In many ways, it raises pressing questions about how we imagine the future relations between science, policy and society in a globalized world. Despite an ever-growing amount of literature, however, there are only few systematic studies on the historical roots and developments of the behavioural change agenda in public policy, the modes and varieties of behavioural public policy across jurisdictions and policy areas, the causes for its inter- and transnational spread as well as its ethical and normative implications and its political consequences.2 This handbook contributes to addressing this gap by mapping the terrain and bringing together scholarship from a wide variety of conceptual and analytic perspectives. Leading scholars in this field from various disciplines critically assess the co-evolvement of behavioural science and public 1

M4716-STRAßHEIM_9781785367847_t.indd 1

25/04/2019 13:32

2  Handbook of behavioural change and public policy policy from early experiments to recent strategies of behavioural public policy (part I), its proliferation and professionalization across countries (part II), the application of behavioural insights and instruments across multiple policy areas (part III), and the normative ideals and political pitfalls of behaviour change and public policy (part IV). The overarching goal of these combined efforts is to mirror the state of the art in current research with its different angles, approaches and normative assessments. Behavioural public policy is a contested and historically changing concept (see part I of this volume). Given the emergence of research on behaviour in different scientific fields, there is a variety of definitions of what behavioural insights actually are and should be. Moreover, there have been multiple attempts to define the application of these insights in different branches of government, policy-making, regulation and administration. A more restricted understanding boils it down to the concept of nudging presented in Thaler’s and Sunstein’s (2008) influential book. From a broader perspective, behavioural public policy includes every policy initiative that is tested, informed or at least aligned to evidence from behavioural research (Lourenço et al. 2016). A third approach emphasizes the ways behavioural public policies are embedded in a normative debate about paternalism and its role for the relationship between citizens and the state (Le Grand and New 2015). Behavioural public policy as understood in this handbook includes all means and modes of public policy aiming at influencing individual or collective behaviour3 by using insights from behavioural economics, behavioural sciences, psychology or neurosciences. This entanglement of scientific and political claims, of analysing and governing behaviour is a common feature of all forms of behavioural public policy (Straßheim et al. 2015; John 2018). In behavioural public policy, expertise has become an essential resource for governing as it is supposed to establish and provide the scientific evidence for the behavioural change agenda. The Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel in 2017, honouring the work of Richard H. Thaler, highlights the growing impact of behavioural economics in both science and policy. For behavioural science and expertise the spread of behavioural interventions across the world gives evidence of its reputation and its capability to translate research findings into policy recommendations. Given this symbiotic relationship, behavioural public policy is an influential and highly representative case for understanding the multiple ways in which science and policy are co-evolving and, in effect, co-producing social and political order (Jasanoff 2004b; Beck 2015). Beyond this common denominator, however, the landscape of behavioural public policy is extremely diverse (see part III in this volume). The spectrum of policy instruments is large, including efforts of political and administrative simplification to reduce the cognitive burden on citizens, education programmes for decisions under the conditions of risk and insecurity, techniques of social norms marketing, and behaviourally informed regulation to prevent industry from manipulating consumers (World Bank 2015; Oliver 2015). Nudges are arguably the most prominent subtype of behavioural public policy (Thaler and Sunstein 2008). Based on insights about behavioural biases and heuristics, nudges are specifically targeting the mechanisms of the cognitive system without resorting to either force or incentives. Since the 1970s, psychologists such as Amos Tversky, Daniel Kahneman, Baruch Fischoff, Paul Slovic and others have argued that mental shortcuts and other biases deviating from the model of rational behaviour are common in complex

M4716-STRAßHEIM_9781785367847_t.indd 2

25/04/2019 13:32

Introduction  3 situations (Kahneman et al. 1982; Sent 2004). Additional information is not always helpful because it might aggravate the problem. Nudges are supposed to make use of these biases to steer people’s behaviour in the direction of their self-defined interest (see Guldborg Hansen, Chapter 5 in this volume) – mostly by re-designing the decision-making environment (‘choice architecture’). Since choice architectures are ubiquitous and unavoidable, proponents argue that policy-makers should become designers of choice architecture (Thaler and Sunstein 2008). This is one of many reasons why behavioural public policy and approaches on policy design are closely interlinked (Jones et al. 2013). Today, many behavioural designs are commonplace: the rearrangement of food in cafeterias to focus our attention on healthy options; graphic warnings on cigarette packages; the fly in the urinal – these and other interventions have become part of everyday life (Peeters et al. 2013; Alemanno 2012; Shafir 2013). The taken-for-grantedness, even banality of some behavioural public policies suggests that ideas and instruments of behavioural change are already more widespread than it might seem. This could be one of the reasons why measuring their influence and impact is still a desideratum of research on governance and public policy. Especially the long-term consequences, both intended and unintended, of nudges merging with our physical and social environment are still unclear. While some argue that nudges are only ‘signposts’ leading the way without prescribing it,4 others insist that the behavioural agenda carries some more or less explicit normative understandings that could transform, if not undermine, the societal values of law, politics and democracy (see part IV of this volume). Indeed, the discussion on the libertarian and paternalist foundations of behavioural public policy revitalizes some classical reflections of political theory on problems such as self-binding and authenticity, autonomy and rationality, individual and collective wellbeing (Elster 2000; Offe 1992). This is not an abstract discussion: As a result, proponents of behavioural public policy have begun to argue for measures of happiness and subjective wellbeing as an alternative to structural indicators such as the GDP (Halpern 2009; O’Donnell et al. 2014). Today, wellbeing indicators have been implemented in the national statistics of more than 40 countries worldwide (Fleischer et al. 2016). It seems that the current struggle on how to define and measure political and societal progress as documented, for example, by the International Panel on Social Progress (IPSP) cannot be separated from the ways behavioural public policy is already influencing the debate on (re-)defining the purposes and goals of collective decisionmaking (Tomlinson and Kelly 2013; IPSP 2018). The following sections give a brief introduction into the main topics. The contributions in this handbook open up various conceptual and empirical angles on the history, variation, application, and normative underpinnings of behavioural public policy. They indicate that the link between behavioural change and public policy should not be taken for granted. Instead, we suggest tracing the multiple ways behavioural public policy is constituted, translated and legitimized across policy areas and jurisdictions. This, in turn, is a precondition for understanding its impact, its political consequences, its internal contradictions and its larger role in the transformation of public policy.5

M4716-STRAßHEIM_9781785367847_t.indd 3

25/04/2019 13:32

4  Handbook of behavioural change and public policy

PART I: KNOWING AND CHANGING BEHAVIOUR: THE CO-EVOLVEMENT OF BEHAVIOURAL SCIENCE AND BEHAVIOURAL PUBLIC POLICY Behavioural science and its use in public policy are often identified with the concept ‘nudge’ and the book carrying the same name, published by Thaler and Sunstein in 2008. Thaler and Sunstein have coined the concept and contributed to provide it visibility in scientific and public discourses around the world. Personalized by Thaler and Sunstein, the rise and uptake of behavioural public policy is often presented as a one-dimensional and linear trend. Although both contributed to shape the field and raise its visibility, the emergence of behavioural sciences and public policy has had many drivers and resulted in a variety of meanings and applications. One of those drivers is the Behavioural Insights Team in Britain (BIT). Founded in 2010 by the Cameron government, BIT has become a paradigmatic example for the translation of behavioural insights into public policy. In their detailed study on the rise of the ‘psychological state’, Jones and colleagues (2013) show that in the mid-1990s, think tanks such as Demos brought economists, psychologists, consultants and public policy experts from the US and the UK together in a series of workshops to search for modes and models of changing behaviour beyond the neoclassical understanding. After New Labour won the elections in 2001, insights from this exchange were integrated much more systematically and explicitly into the governmental strategy (COSU 2004). The Cabinet Office Strategy Unit (COSU) cooperated with multiple British ministries in experimenting with behavioural interventions. At the same time, Richard Thaler became an advisor to the Conservatives (Jones et al. 2013, p. 24). In 2010, MINDSPACE was published, an influential paper that deviates in both style and structure from the usual government reports. In order to set standards for future behavioural public policy, MINDSPACE combines insights from behavioural economics with principles for behavioural change interventions. By simplifying its message, using acronyms for identifying factors of behavioural influence and giving good practice examples, MINDSPACE is in itself a prototype of behaviourally informed design that has been of programmatic influence. After the victory of the conservative–liberal coalition at the elections in 2010, BIT was founded and integrated into the Cabinet Office. In the years to follow, it quickly expanded and has today more than 120 employees in London and in its branches in Manchester, Sidney, Singapore and New York. In 2014, it was privatized and is now jointly owned by the UK government, the NESTA foundation and its employees. BI Ventures, a team within BIT, develops digital products that address issues such as recruitment, online RCTs, designing behavioural interventions and individual goal attainment (‘good habit lab’). The case of the BIT is illuminating because it illustrates some of the mechanisms behind the rise of behavioural public policy. First, behavioural public policy has been and still is continuously co-produced by politico-administrative, scientific and other societal actors in boundary-crossing networks. The result is a circle of both collaborations as well as an emerging frame of reference for its members that help them to gain authority, reputation and relevance across multiple contexts. Second, organizations such as the BIT are ‘multitasking mediators’ (Botzem and Straßheim 2016). Professionalization is key. By bringing together expertise from various backgrounds such as behavioural sciences, psychology, economics, public policy, mathematics and cultural studies, they

M4716-STRAßHEIM_9781785367847_t.indd 4

25/04/2019 13:32

Introduction  5 cope with a multitude of tasks in different fields of specialization. In terms of structure and recruitment, many behavioural expert organizations resemble consulting agencies. Third, behavioural public policy reaches across a broad spectrum of political positions. Both Labour and Conservative governments have made use of behavioural insights. This ideological flexibility has certainly been helpful in spreading behavioural ideas. The ‘prehistory’ (Jones et al. 2013) of behavioural public policy, however, reaches far back to the second half of the twentieth century when behavioural sciences and public policies to regulate behaviour co-evolved (see Graf, Lebaron and Rawat, Chapters 2, 3 and 4 in this volume). The emergence of the concepts flagged under the umbrella term ‘behavioural insights’ (BI) is embedded in several historical trajectories and contexts rather than converged into one pattern. Scholars in the fields of economics and public policy have long recognized the importance of behavioural processes and designed policy tools to address behaviour change. From the beginning behavioural economics has always been systematically related to policy issues, using its key heuristics as a source for suggesting behaviour changing interventions. Behavioural economists have extensively developed corrective procedures, modes of removing biases and instruments of behavioural intervention to improve collective judgment heuristics and risk perceptions as a core challenge for public management (Kahneman et al. 1982; APSC 2007). The example of traffic safety regulation since the 1960s, however, shows that behavioural interventions evolved even before behavioural public policy or behavioural economics were actually conceptualized (for the following see Graf, Chapter 2 in this volume). Experts from the behavioural sciences and some early behavioural economists were influential in developing traffic safety regulation and road construction. Behavioural measures to influence and change people’s conduct were also discussed in other policy areas such as environmental protection and health policies. In the early 1980s, experts discussed ‘prompts and cues’ to influence environmental and health behaviour. Some of these behavioural approaches to regulation were already common practice in many areas when behavioural economics started to gain prominence. Kahneman and Tversky were so especially successful because they found a way back into mainstream economics by continuing to accept ‘homo economicus’ as a benchmark for economic behaviour (Sent 2004). While Herbert Simon had developed the alternative concept of ‘bounded rationality’, they kept the model of the homo economicus but focused on those instances in which empirically observable decision-making systematically diverged from it. As Graf argues, the decline of Keynesian approaches in the 1970s and the rise of neoliberal concepts in the 1980s, behavioural economics offered new opportunities for the state to intervene where markets had failed. From this long-term perspective, both behavioural economics and behavioural public policy need to be seen as part of a much broader historical movement to observe and regulate human behaviour. An essential element in this broader context is the rise of economics as professional expertise. As Lebaron argues (Chapter 3 in this volume), these changes were not unidirectional and consisted of multiple dynamics such as the spread of economic indicators, the ‘financialization’ of economies, the increase in think tanks and consultancies, and the combination of different types of political, economic and epistemic authority. Economics has become a scientific subsystem strongly structured around a set of institutions of evaluation and consecration, with the ‘Nobel memorial prize in economic sciences’, so-called ‘Nobel Prize’ at its top. While the financial crisis from 2007 created a new

M4716-STRAßHEIM_9781785367847_t.indd 5

25/04/2019 13:32

6  Handbook of behavioural change and public policy and ambivalent situation for economic expertise in general, it may have supported the proliferation of behavioural economics in the public sector. One of the reasons is that it offers a critique to and an alternative approach to orthodox economics whose hegemony was challenged during the financial crisis. It took behavioural economics several decades to gain acceptance in both economics and public policy research (for the following see Rawat, Chapter 4 in this volume). Between 1950 and 1999 public policy research even took the lead in studying behaviour. This is very much in line with the observation that studies on behaviour and behavioural interventions in the public sector were already common before behavioural economics started its career (Graf, Chapter 2 in this volume). It was only as late as in the 2000s that economics caught up. Driven by the rise of behavioural economics there was a massive increase in behavioural studies that surpassed public policy research. In the past few years, this changed again when public policy increasingly began to adopt experimental methods beyond the laboratory. Geographically, behavioural studies were for a long time heavily based in the US. In economics after 2000, studies seem to have spread towards other European and Asian nations, with the highest number of Asian and African studies set in India and Kenya. In public policy research between 2010 and 2016, studies outside of the US are mainly from Denmark, the Netherlands and the UK as main countries. At least for this period of time, ‘not a single study dealing with the behavioural aspects of public policy in the African continent exists’ (Rawat, Chapter 4 in this volume). Interestingly, Rawat observes that both economics and public policy studies seem to be showing some partial convergence: public policy studies are using more experimental methodologies and insights from behavioural economics while behavioural economics increasingly study applications in a public policy context. Today, behavioural public policy and related concepts such as nudge are at the core of a debate that challenges the foundational principles and traditional conceptualizations of public policy. As both Hansen and Howlett argue in this volume (Chapters 5 and 6, respectively), the established ways of thinking about public policy and its implementation have been congruent with a utilitarian, rationalist orientation ‘which has been generally pervasive in the policy sciences from the very founding of the discipline’ (Howlett, Chapter 6 in this volume). The result has been the inability of rational approaches to public policy and regulation to cope with complex problems, policy failures and variations in compliance. Alternative concepts of public policy are addressing these limits: nudging is only one of many efforts to better understand the complexity of individual and collective behaviour such as co-production, deliberation, faith-based public services or social marketing. What sets nudging apart from these approaches, however, is that it both presupposes and ‘exploits’ these seemingly irrational behaviours by making use of cognitive boundaries, biases and habits in order to influence people to act ‘in their own selfdeclared interests’ (Hansen, Chapter 5 this volume). Other strategies (see Hansen) involve the exaggerated use of traditional regulatory approaches in light of severe behavioural problems such as massive taxes on cigarettes (‘push’) or the regulation of activators of biases, for example by banning pre-ticketed boxes in online businesses (‘clear’). Whatever we might think of nudging or other behavioural approaches, they teach us that biases and bounded rationality have serious implications for both citizens and policy-makers. One of the key challenges for future developments may involve a constant evaluation not only of nudges and related instruments but also of the policy-makers themselves in order to

M4716-STRAßHEIM_9781785367847_t.indd 6

25/04/2019 13:32

Introduction  7 detect and avoid biases in decisions and policy-making. Moreover and in a more general sense, policy-makers need to be aware that the success (or failure) of governing depends on citizen’s perceptions of – and judgments about – their credibility, legitimacy, cupidity and competence (Howlett, Chapter 6 in this volume).

PART II: NUDGING AROUND THE WORLD: PROLIFERATION AND PROFESSIONALIZATION The very fact that behavioural science turned behaviour into a universally accepted concept (Graf, Chapter 2 in this volume) leads to the expectation towards integrated scientific approaches and standardized procedures gaining ground internationally. Expanding the base of shared knowledge will promote, as the argument goes, transnational convergence and uniformity. The emerging research on the spread and diffusion of behavioural public policy paints a more complicated picture. It also explores how behavioural science is operating in and embedded into complex regulatory regimes embracing multi-level, decentralized decision-making processes; interaction between states and networks of non-state actors; and modes of societal coordination (Jordan et al. 2015). Indeed, regulatory approaches to global issues, from climate change to food safety, are increasingly operating through pluralistic, overlapping, and fragmented regimes (Beck and Mahoney 2018; Keohane and Victor 2011; Abbott 2012; Kuyper et al. 2017). Science and multi-level governance is also a key theme in the study of the European Union. In relation to energy, taxation and land-use planning, Member States have insisted on preserving a high degree of autonomy – ‘subsidiarity’ in the language of the EU – meaning that the Commission’s competence to propose common policies is still relatively limited. As a consequence, it is an empirically open question whether the use of behavioural insights may override deeply seated and institutionalized modes of policy-making or create spaces for the protection of local values and local autonomy and thus contributes to institutional fragmentation (Jasanoff 2013). Indeed, recent research shows that the ‘varieties of behavioural public policy’ (Straßheim 2017a, p. 220; Lourenço et al. 2016; Whitehead et al. 2017) are multifaceted and complex. Utilizing a set of policy proxies and tropes, Whitehead, Jones and Pykett argue (Chapter 7 in this volume) that behavioural public policy in a broad sense (i.e. behaviourally-tested, behaviourally-informed and behaviourally-aligned policies) have travelled around the world to a great extent. According to their study, behavioural sciences have some impact on public policy delivery in 135 independent states (and Taiwan). Reaching across a broad spectrum of policy areas, more than a half of the states have developed centrally directed policy initiatives. In contrast, South (but not Central) America, parts of Eastern Europe and large portions of the Middle East are all characterized by an absence of behavioural public policies. Inter- and transnational organizations such as USAID, AusAID, UNICEF, the World Health Organization, and the United Nations Population Fund, but also NGOs and multinational organizations are playing an important role for the diffusion of behavioural approaches in regions such as Africa. A critical geography of behavioural public policy can highlight not only the geographical variations and adaptations, but also the forces shaping these differences. In a similar vein and based on their influential report ‘Behavioural Insights Applied

M4716-STRAßHEIM_9781785367847_t.indd 7

25/04/2019 13:32

8  Handbook of behavioural change and public policy to Policy’ (Lourenço et al. 2016) Ciriolo, Lourenço and Almeida (Chapter 8 in this volume) are reviewing behavioural public policies and institutional developments across 32 European countries. Behavioural insights, they argue, are not confined to nudges and may support a broad range of policy instruments including more traditional forms. As the example of road safety policy illuminates (see also Graf, Chapter 2 in this volume), European legislators may have taken results from behavioural studies into account when they decided in 2006 that wearing seatbelts for all vehicles across the EU should become compulsory. At the EU level, the integration of behavioural insights began in 2008 in consumer and competition policy. In 2010, the European Commission started to carry out studies and collected behavioural evidence in multiple policy areas such as financial services and tobacco, energy labelling, online gambling and food information. Since May 2017, the European Commission has been training public officials in behavioural sciences. Moreover, an overview on very recent applications of behaviourally-tested interventions in European countries shows that even in this narrow category of explicit behavioural public policies initiatives are covering a broad spectrum of areas and topics ranging from consumer protection and competition policy, energy and environment, finance and taxation to public health, welfare and employment. Again, these and other observations on the spread of behavioural public policies across Europe raise questions about the modes of diffusion, the motives of public policy professionals and the degree of uniformity. In countries such as the UK, Ireland, Norway or Denmark the adoption of behavioural approaches is proceeding at a speedier rate; in others such as Germany, the Netherlands, France or Italy the picture is more scattered and decentralized; a third group of countries such as Austria or Luxembourg is more or less absent from behavioural public policy-making (see Ciriolo et al. and Alemanno, Chapters 8 and 10, respectively, in this volume). Analysing the development of a new occupation of behavioural experts in specific institutional settings and the multiple ways these actors are interlinking science and policy may help to uncover some of the underlying mechanisms of this uneven spread. Feitsma and Schillemans (Chapter 9 in this volume) reveal that in the case of the Dutch government profession is still in its infancy and currently more fragmented than cohesive. Behavioural experts build and legitimize professional boundaries to develop their knowledge, position, standards and identities. Not all behavioural experts are organized in teams such as the BIT in the UK; as in the Dutch case, they may collaborate in a loose form of informal networks, knowledge exchange groups, working groups or strategic projects. Despite these less coherent organizational forms, experts draw on a cohesive core of ideas, books, reports, and role models. They share little uniform standards and struggle with the tension between autonomy and integration in existing structures. In the system of relatively autonomous government agencies in the Netherlands, behavioural expertise will most probably remain fragmented. These findings are important, because they open up venues for understanding the influence of different administrative cultures and government institutions on the professionalization of behavioural public policy in Europe and beyond (on the British and German case see Straßheim et al. 2015). In the multi-level system of the European Union the future developments and political consequences of both the proliferation and professionalization of behavioural public policies are difficult to assess (see Alemanno, Chapter 10, Leone and Tallacchini, Chapter 11, and Zuidhof, Chapter 12 in this volume). The European Commission’s in-house

M4716-STRAßHEIM_9781785367847_t.indd 8

25/04/2019 13:32

Introduction  9 science service, the Joint Research Centre (JRC), has tasked the EU Policy Lab, a multidisciplinary team, and its Foresight and Behavioural Insights Unit (FBIU) with exploring and re-examining policy issues in the fields of foresight, behavioural insights and policydesign. In order to fulfil these functions, the Policy Lab is supposed to coordinate existing behavioural research capacities at the JRC and at EU agencies and to take stock of past experiences with behavioural public policies. It organizes thematic workshops with different Commission services on policy issues from a behavioural perspective, offers training modules and engages with DGs to policy implementation problems and solutions from a behavioural perspective. It also works on including behavioural insights into the policy cycle and to develop standards for behaviourally informed policy-making and regulation as part of the Better Regulation initiative. The Better Regulation Toolbox, issued in 2015, is more or less systematically referring to behavioural insights as integral elements of evidence-based policy. As Alemanno argues, however, there are certain tensions and limitations constraining the use of behavioural public policy in the European Union. Given that efforts to nudge citizens in the direction of ‘good behaviour’ are already facing public criticism on the national level, their application on a supranational level might turn out to be even more difficult. The EU is an administrative entity that has virtually no direct contact to its citizens and it shares its citizenry with the Member States. As a result, ‘several visions of what a “good citizen” is – or ought to be – tend to compete’ (Alemanno, Chapter 10 in this volume). On the other hand, the EU has used instruments such as the internal market and EU citizenship to redefine what is good for citizens and to shape both national and EU lifestyle policies in relation to tobacco, alcohol and diets regardless from which country in Europe they are coming. The technocratic character of EU law making may also be conducive to the uptake of approaches informed by behavioural evidence. While the EU vision of citizenship has, at least in principle, been committed towards active citizen participation, the appeal of behavioural public policy to the EU might result from its implicit technocratic and top-down character. The fate of nudging in the EU, as Alemanno concludes, ‘appears as intertwined with its political dimension as it is with its social-scientific one’. Focusing on this contradictory meaning of European citizenship, Leone and Tallacchini show that behavioural public policies fail to take the full meaning of citizenship into account. In its transformation to a political community, the concept of citizenship has been considered as a key element in legitimizing innovation in the EU both from the epistemic and the normative point of view. This role, however, has remained mostly theoretical, if not rhetorical; nonetheless, the pressures to speed-up governance in Europe have led to new regulatory soft practices (such as nudge) that challenge citizenship. Two examples, dealing with health and digital privacy, where citizens are treated as ‘objects of concern and control’, reveal some of these ambiguities. In order to rethink nudging as a more legitimate form of normative innovation, Leone and Tallacchini propose the notions of Participatory Design and Rights-in-Design as ways to open up and reframe choice architectures by granting more robust epistemic and democratic credibility. This includes the reference to the term ‘citizen’ rather than ‘consumer’ in order to frame the discourse in a broad political and social perspective, where learning how to collectively contribute to a better society is a primary concern. If regulation is increasingly seen as a ‘learning process’, citizens should participate in designing this process. This suggestion

M4716-STRAßHEIM_9781785367847_t.indd 9

25/04/2019 13:32

10  Handbook of behavioural change and public policy might resonate with some more recent ideas on combining nudges with feedback mechanisms, greater transparency and modes of citizen deliberation (John 2018). Given that in the European Commission behavioural public policies typically move through expert channels such as the EU Policy Lab, such political considerations about citizen involvement might not resonate easily within the EU (see Zuidhof, Chapter 12 in this volume). One could indeed argue that behavioural interventions are produced by and at the same time reproduce a complex environment of interdependencies between administrative professionals and scientific experts, leading to a quiet depoliticization of policy-making and market regulation. Moreover, nudging and similar approaches prove to be especially suitable in contexts were the EU has only limited competences and would otherwise be confronted with considerable political obstacles. They may challenge the fragile democratic achievements of the EU. The recent call by the Group of Chief Scientific Advisors, however, for getting a better understanding of ‘biases in doing science’ and of situations in which ‘evidence utilisation can shift the political debate to particular questions or concerns in a non-transparent way’ could be interpreted as a first sign of a critical discussion on the technocratic foundations of EU policy-making from within the system of professionals (EU Commission 2018).

PART III: BEHAVIOUR INSIGHTS AND INSTRUMENTS IN POLICY-MAKING The contributions of part III in this volume explore the particular nature of policy tools based on behavioural insights and relate them to traditional policy instruments such as regulatory and market-based interventions. In order to capture their innovative character and identify their common features, research is conducted in a broad range of fields such as public health (Quigley and Farrell, Chapter 14, and Loer, Chapter 13 in this volume), sustainability (Bornemann and Burger, Chapter 15 in this volume), pension policy (Weaver, Chapter 16 in this volume), development policy (Berndt, Chapter 17 in this volume), employment (Tosun and Hörisch, Chapter 18 in this volume) and taxation (Botzem, Chapter 19 in this volume). They all address the question whether the behavioural approach to government tools is fit for function to address emerging challenges in an effective and legitimate way. They also raise questions as to how behavioural public policy impacts upon broader issues concerning evidence-based policy and political legitimacy and contributes to transforming politics and policies. Probably one of the more important advances in public policy studies was the notion that different programmes and tools of government are based on a limited number of mechanisms – mechanisms that form the very basis of public policy. A classical typology is the NATO scheme (Nodality, Authority, Treasure, Organisation) which has later been extended in numerous variations (see also Loer, Chapter 13 in this volume). While some argue that nudges and similar behavioural interventions are fundamentally different from this traditional set of instruments and therefore need to be conceptualized as a further extension of public policy, Peter John and others have pointed to the fact that all public policy interventions are embedded in an informational context (John 2013). From this view, behavioural public policy is a specific style of interpreting and restructuring the informational context of both novel as well as conventional instruments of governing.

M4716-STRAßHEIM_9781785367847_t.indd 10

25/04/2019 13:32

Introduction  11 The consequences are far-reaching as they do not only concern the potentials and limits of behavioural public policy but also the political and social perceptions of public policy in general. Kathrin Loer (Chapter 13 in this volume) echoes this debate by arguing that behavioural insights in public policy transform existing instruments in a way that changes our understanding of the underlying mechanisms. Very much in line with the contributions by Hansen and Howlett, Loer shows that most of the existing instruments in public policy have been conceptualized without taking into account the biases and bounded rationality of target behaviour. Behavioural insights, however, give a ‘spin’ (Loer, Chapter 13 in this volume) to policy instruments. Applying the findings of behavioural research in policy-making requires shifting the focus on the addressee and the question why they behave the way they do. They enhance the visibility of a variety of environmental and cognitive factors that influence people’s preferences and decisions. Heuristics such as mental accounting, sunk-cost effects, availability, salience or anchoring are relevant for policy design across the whole spectrum of instruments and instrument combinations and not only for nudging-type interventions. The most recent ‘sugar tax’ in the UK could be seen as a case in point (Hallsworth et al. 2016). In public health, the result of this ‘spin’ has been an increasing focus on promoting individual behaviour change in relation to ‘lifestyle risks’ caused by tobacco, alcohol, and dietary factors. Muireann Quigley and Anne-Maree Farrell (Chapter 14 in this volume) critically examine how the political context frames the use of nudges and thus impacts the use of behavioural insights in public policy. Drawing on health promotion and organ donation, they show how the use of behavioural insights in public policy has led to a mismatch between policy intent and implementation. Behavioural health promotion is focusing too much on individual lifestyle preferences, with insufficient account being taken of the wider social determinants of health. Policy-makers may be motivated by the fact that health interventions which focus on individual lifestyle risks may be quicker and easier to implement than the sustained effort required to alleviate poverty in vulnerable populations. Sustainability has become another policy area in which behavioural insights and interventions are used more frequently (Bornemann and Burger, Chapter 15 in this volume). The focus is on those (Western) consumption behaviours that are thought to be responsible for the unsustainable pathway of the global human society. Within subfields such as household energy use, mobility, food consumption, shopping behaviour, tourism or waste production interventions have been tested in randomized controlled trials, small-scale labs and field experiments. Bornemann and Burger argue that both advocates and proponents of behavioural interventions tend to overlook the normative and evaluative dimensions of multiple sustainability concepts such as justice-based, resilience-based or procedural sustainability. It turns out that the relationship between nudging and sustainability is filled with tensions and normative implications that vary between different sustainability conceptions. The authors argue that further research will have to scrutinize the potential and actual role of nudging in normatively ambitious and empirically complex governance settings. Scrutinizing complex governance settings is important for understanding the multiple barriers, side-effects and failures of behavioural public policy (Weaver 2016; Sunstein 2017). Weaver (Chapter 16 in this volume) illuminates this challenge by looking at labour

M4716-STRAßHEIM_9781785367847_t.indd 11

25/04/2019 13:32

12  Handbook of behavioural change and public policy market exits from a multidisciplinary behavioural perspective that places multiple determinants of the behaviour of a very heterogeneous set of actors at the centre of analysis. A variety of strategies are, as Weaver shows, available to governments to alter this behaviour, ranging from improved information about longevity risks to increasing minimum and ‘standard’ retirement ages. His analysis suggests that for extending working lives, as with most policies that have diverse target populations, it is critically important to look at multiple barriers to behaviour change such as complex incentives and a high degree of uncertainty that challenges the cognitive capacity of most workers, plus substantial peer effects – and to develop policy interventions that address those barriers. Knowledge about ‘compliance and enforcement regimes’ (Weaver, Chapter 16 in this volume) may be of strategic importance for the future of behavioural public policy. The complexity and political repercussions of translating behavioural insights into public policy are also highly relevant in the context of development policy (see Berndt, Chapter 17 in this volume). Scholars such as Banerjee and Duflo (Banerjee and Duflo 2012) have argued that poverty creates a cognitive burden for the poor that leads to disadvantageous behaviour caused by loss-aversion or procrastination. As Berndt shows, the ‘behavioural revolution’ (Datta and Mullainathan 2014) in designing development policy has profited from distancing itself from both neoclassical market-oriented policies as well as from traditional large-scale state interventions. Randomized field experiments play a crucial role in the spread of behavioural thinking in public policy. In the area of development, however, this has also created an asymmetric focus on the relationship between experts and the poor that is in many ways a social projection of the cognitive model of system 1 and system 2 (Kahneman 2011). Without complete turning away from homo economicus, this model is used as an ‘utopian yardstick’ in behavioural change interventions (see also Sent 2004). By implicitly sticking to the neoclassical discourse, development policies based on behavioural insight mobilize well-known imaginations of modernity and render development a technical problem of behavioural engineering. Barriers and possibilities of behavioural change interventions become especially visible in the case of employment policy. Focusing on policy instruments prescribed by the European Union to promote youth employment, Tosun and Hörisch (Chapter 18 in this volume) critically assess deliberative interventions that are discussed in behavioural research as ‘boost’ (Grüne-Yanoff and Hertwig 2016) or ‘think’ (John et al. 2009) strategies. In a complex area such as youth unemployment with its economic, social and political implications, the European Union has developed a multitude of approaches ranging from non-formal learning and intercultural dialogue to a complex coordination mechanism such as the open method of coordination. Based on a survey on the beliefs of young Germans between the ages of 18 and 35 on what the causes of youth unemployment are and on what they would be willing to do to get a (better) job, the authors show that these young Germans are willing to move within the country, learn new skills, or even get retrained. The majority of them, however, are unwilling to move to a different country. While EU instruments tackling youth employment are surprisingly well designed to cope with the complexities of compliance (see also Weaver, Chapter 16 in this volume), they do not include social protection or the demand-side of labour markets, thus leaving important dimensions of the choice architecture uncovered. Compared to areas such as public health or sustainability, behavioural public policies in tax collection are relatively recent (see Botzem, Chapter 19 in this volume). In taxation

M4716-STRAßHEIM_9781785367847_t.indd 12

25/04/2019 13:32

Introduction  13 policy, the behavioural ‘spin’ on instruments that has been observed by many authors of this volume takes a moralizing form. By placing emphasis on the moralizing character of official communications the relationship between the state and the individual is reinterpreted in terms of moral conduct and social norms. Most behavioural interventions in this area seek to discourage tax avoidance by tax payers. As Botzem shows, tax morale, trust and cultural determinants of tax compliance have become leading concepts in the past decades. With the rise of randomized controlled trials since the 2000s, studies on tax letters have been carried out across many countries in Europe, North America and Latin America. While the effects of behavioural interventions in terms of enhancing tax collection are limited, proponents emphasize the possibility of directly communicating the importance of social norms to the tax payer: ‘The state is using moralization as an approach to address its citizenry, invoking an imagined contract between taxpayers and tax authorities’ (Botzem, Chapter 19 in this volume). Very much in line with other contributions in this section of the handbook, Botzem underscores the complexities of (tax) compliance and the implicit normative assumptions and unintended side-effects that should not be underestimated when implementing behavioural public policy.

PART IV THE GOVERNANCE OF BEHAVIOUR: NORMATIVE IDEALS, CRITICAL PERSPECTIVES AND POLITICAL CONSEQUENCES The former chapters indicate that differences in the rise, diffusion and implementation of behavioural public policy are often rooted in more fundamental differences over world-making assumptions such as the boundaries of state intervention into the market and models of agency. These controversies also reflect crucial political and normative disagreements about the role of the (behavioural) science in policy-making or the legitimacy and accountability of regulatory agencies. The studies of policy areas also show that there is no single driver or trend, but rather a mutual interplay of several constellations in science and policy-making. Behavioural public policy is a totemic example of where science is asked to provide evidence for political decisions and where expertise has become a major source of political legitimation. The growing political demand for legitimation by behavioural experts for political choices raises novel challenges for science and expert bodies and realigns forms of scientific and political representation. The next series of contributions addresses – broadly speaking – the political impacts and implications of the rise, spread and institutionalization of BI (including sectoral, geo-political and cultural variations) from a variety of perspectives. They range from normative-conceptual considerations to empirical-descriptive findings and include ethical, normative and political, legal and institutional dimensions of using BI for and by public policy-making. They all focus on the changing relationship between state and non-state actors (including citizens). They all raise questions of agency, representation and their normative and epistemological implications under different, but complementary perspectives. Questions of legitimacy are addressed not only as ethical or normative ones but also as positive or empirical ones: Which nudges do people endorse, and which ones do they reject? They also address gaps such as the constitutional limits to nudging and they all reflect how behavioural public policies contribute to and reflect upon transformations

M4716-STRAßHEIM_9781785367847_t.indd 13

25/04/2019 13:32

14  Handbook of behavioural change and public policy of the political (politics, policy and polity), including forms of de- and re-politicization and changes in the structures of power and authority. Sunstein and Thaler’s argument that ‘libertarian paternalism is not an oxymoron’ (Sunstein and Thaler 2003) was the starting point for a heated debate on the ‘ethics of nudge’ (John 2018). The basic idea of this argument is that people are free to not follow a nudge. This libertarian notion is combined with the paternalist notion that if people choose to follow a nudge their behaviour is corrected from biases and problems of self-control in a way that is either socially desirable or in their own interest as judged by themselves: ‘choice architects can preserve freedom of choice while also nudging people in directions that will improve their lives’ (Thaler and Sunstein 2008, p. 252). Critics, however, have argued that it is highly unclear what is good for people, what they would choose for themselves and what they themselves would call a bias or problem of selfcontrol (Sugden 2016). A more recent turn in this debate is the focus on public opinions about nudges (see Sunstein, Chapter 20 in this volume and John 2018, pp. 117‒118). Studies have shown that people prefer open and transparent nudges to covert nudges; they choose nudges that support socially desirable goals over pro-self nudging (Felsen et al. 2013; Jung and Mellers 2016). There are also considerable differences in the approval of nudges across different nations (Sunstein et al. 2018). In his contribution to this volume, Cass Sunstein presents and discusses the results of a representative survey among American citizens involving thirty-four nudges. The data indicate a strikingly broad consensus, across partisan lines, that people dislike those nudges that promote what people see as illicit ends or are perceived as inconsistent with either the interests or values of most choosers. By contrast, there is a general approval of nudges that are taken to have legitimate ends and to be consistent with the interests and the values of most choosers. One main difference is that partisan affiliation seems to affect the level of support; even if the ends of nudges are supported Republicans are more sceptical than Democrats. While there is a clear majority in agreement about which nudges should be supported, these important findings about the influence of party affiliations point to a distinct political dimension of nudging that deserves greater attention in future research. In her contribution, van Aaken (Chapter 21 in this volume) enters the debate on libertarian paternalism not from an empirical, but from a legal standpoint. She distinguishes between paternalistic nudges, that is, nudges targeting the wellbeing of single individuals and non-paternalistic nudges, that is, nudges that target third-party externalities or the public good. Taking this difference as a starting point, she assesses the constitutionality of nudging regulations by identifying infringements of rights associated with nudges and possible justifications. Governments need to respect constitutional limits to nudging, van Aaken argues. Especially paternalistic nudges are problematic. Constitutional limits may also arise due to the design of nudges. A crucial point is the informed and free decision of citizens. Whenever an individual makes a decision based on informed preference, paternalistic nudges fail the test of constitutionality even if they are effective and necessary to protect the individual against self-damaging behaviour. In a complementary way to van Aaken’s legal perspective, White addresses epistemic, ethical and political problems of nudges from a philosophical perspective. Proponents of behavioural public policy are confronted with complex, multifaceted and subjective interests that are based on a myriad of choices in a specific situation; constructing their

M4716-STRAßHEIM_9781785367847_t.indd 14

25/04/2019 13:32

Introduction  15 ideal preferences is a counterfactual exercise; nudges collapse this complexity into one focused area without taking into account other dimensions. This epistemic problem is accompanied by ethical problems whenever policy-makers and experts impose their own conception of interests on others or violate liberal neutrality by preventing individuals from pursuing their own vision of the good life. In consequence, people may even be subject to ‘cognitive hazard’ in which they do not deliberate as carefully about decisions when they know someone will prevent them from making bad ones. Somewhat in line with van Aaken’s distinction between paternalistic and non-paternalistic nudges, White (Chapter 22 in this volume) argues that one possibility would be to base nudges explicitly on universal welfare standards or objective theories of the good, in which the broad goals of nudges, such as promoting health or wealth, are made clear and the goal of helping individuals pursue their own unique and subjective interests is dropped. Reviewing the literature on libertarian paternalism, Tyers (Chapter 23 in this volume) engages with libertarian paternalism by identifying two recurrent critiques: first, that it is too liberal and individualizes responsibility too much; and second, that it is too paternalistic, undermining individual autonomy and placing too much power and trust in the ‘nudger’. Based on recent theories on anti-politics, post-politics, and de-politicization, Tyers suggests a third version: nudges can be viewed not only as too liberal or too paternalistic, but both. From this perspective, nudges are an expression of contemporary governance which is liberal and ‘de-politicized’ at the macro-level of business regulation whilst also being paternalistic and ‘re-politicized’ at the micro-level of individual behaviour change. Macro-level issues such as the primacy of material consumption or poor quality housing are excluded from mainstream discourse. Micro-level issues such as consumers’ consumption decisions or individual savings are brought into the political discourse and become target of nudge-style interventions. Behavioural public policy, Tyers argues, reflect broader developments of contemporary governance which is increasingly ‘anti-political’ and ‘solutionist’ – reluctant to regulate capitalism at a macro-level but eager to manage consumers’ micro-level behaviours. Does the widespread adoption of behavioural public policy undermine broader institutions in contemporary societies? Lepenies and Małecka (Chapter 24 in this volume) discuss the challenges which behavioural policy solutions pose for law, politics and science based on an institutionalist perspective. As to the law, behavioural instruments are not introduced through parliamentary legislation, but by acts of administrative bodies. Even if behavioural instruments are disclosed and made fully transparent, the mechanisms by which they are chosen withstand such disclosure. Politically, behavioural public policy might become increasingly problematic if societies ‘inherit’ nudges from prior public administrations. Once installed, nudges could merge with the institutional setting and circumvent public deliberation. Finally, scientific evidence from the behavioural sciences that enters policy can be already value-laden. Therefore, apart from testing the effects of nudge interventions advocated by proponents of nudging, it seems to be important to be able to scrutinize the values that in various ways enter behavioural research. These values should be made subject to debate. In their contribution, Lepenies and Małecka make several suggestions to safeguard society from these problems: a nudging oversight body such as a ‘nudging ombudsman’ could be appointed by parliaments; expiration dates for nudges could force lawmakers to deliberate on implementation in regular intervals; requiring nudge units to install an expert in the philosophy of science could support

M4716-STRAßHEIM_9781785367847_t.indd 15

25/04/2019 13:32

16  Handbook of behavioural change and public policy meta-reflection on epistemic problems. The proponents of behaviour change, as Lepenies and Małecka conclude, have gotten carried away with the design and implementation of behavioural public policy without seeing the broader political, institutional and societal context in which it is advocated.

FUTURE RESEARCH PERSPECTIVES All in all, the findings of this edited volume paint a rich picture of the state of research in this field. They call for more comparisons across policy sectors, levels and countries in order to identify patterns of convergence and divergence and to systematically explain the varieties of behavioural public policy. Moreover, it remains open and contested what the political impacts and implications of these trends are and how they resonate in different institutional and cultural contexts under the conditions of the ‘post-national constellation’ (Habermas 2001). From the many crosscutting themes in this handbook we would like to highlight four key challenges that could inform future research in this field. First, more research is needed on the multiple ways of how behavioural approaches transform our understanding of public policy. Many contributions in this volume have shown that behavioural public policy does not simply add another set of tools to the already existing spectrum of policy instruments. It is still an open question how behavioural interventions interact with other tools of government and what the unknown side-effects of these interactions could be. Similarities and differences between different policy sectors and levels also raise novel question, such as about enabling and constraining context condition for the spread and integration of behavioural insights into policy-making. ‘[N]udge needs to break from the individualising focus stemming from its supporters in behavioural economics and psychology and give more attention to the way that collective and institutional settings help determine the success or failure of a nudge’ (John et al. 2009, p. 369). While a variety of studies highlight the influence and significance of the broader political, institutional and societal context in which behavioural change and public policy is advocated, it remains a black box and calls for future investigation about how and why ‘context matters’. Moreover, most authors in this handbook have argued that the debate on behavioural insights and public policy needs to break away from the focus on nudges. Alternative uses of behavioural insights include modes of risk-oriented education, behaviourally-informed regulations that are supposed to protect consumers and citizens and multiple strategies to establish institutional spaces for citizen’s engagement, consultation and deliberation (Gigerenzer 2015; Oliver 2015; Grüne-Yanoff and Hertwig 2016). Most importantly, the contributions in this handbook have shown that the specific frame of behavioural public policy has implications for our conceptualization of the underlying mechanisms of public policy instruments, for the ways we assess and criticize public policies and for the norms and values we associate with them. These findings across multiple policy areas invite to deeper investigation on whether and how the spread of behavioural insights affects the allocation of power and responsibilities between the state and non-state actors and how it affects the normative foundations of modern democracies. It is still an empirically open and contested question how it impacts questions of individual agency, whether it contributes to empowering citizens or whether its close relationship to evidence-based policy provides legitimation

M4716-STRAßHEIM_9781785367847_t.indd 16

25/04/2019 13:32

Introduction  17 for experts and tendencies of technocracy in public policy-making (Fischer 2009; Straßheim 2017b). Second, another key challenge for future developments involves a critical evaluation not only of nudges and related instruments but also of the behavioural policy-makers and experts. The ‘choice architects’ themselves have become objects of behavioural studies in order to detect and avoid biases in decisions and policy-making (BIT 2018; World Bank 2015; Lodge and Wegrich 2016). The findings in this handbook are very much in line with recent studies arguing that ‘elected and unelected government officials are themselves influenced by the same heuristics and biases that they try to address in others’ (BIT 2018, p. 3). Some have called this the ‘rationality paradox’ of behavioural public policy (Lodge and Wegrich 2016). Interestingly, the research by Kahneman and Tversky was originally motivated by the question on how expert intuition and professional decision-making works (Kahneman 2011). The ‘confirmation bias’ is probably one of the most frequent heuristics influencing the behaviour of policy professionals. It refers to the selective gathering (or the undue relevance) of information in order to support a previously held belief and to neglect other information. This affects the interpretation of data, the analysis of situations or the critical appraisal of public policies. The co-evolvement of behavioural sciences and behavioural public policy as mapped out in this volume suggests that both experts as well as policy-makers could even reinforce each other in ignoring some of the political consequences and normative implications of behavioural approaches. Understanding the mechanisms leading to failures and pathologies in mutual learning and co-production between science and policy might have far-reaching methodological and conceptual implications for the future integration and critical reflection of behavioural insights in public policy (Dunlop 2017; Jasanoff 2004a). Third, a variety of authors in this volume have raised questions on citizen’s autonomy and authenticity in behavioural public policy. Understandings of ‘libertarian paternalism’ vary and, as opinion polls show, citizens themselves prefer those nudges that are transparent, have legitimate ends and are consistent with the interests and the values of choosers. This debate is closely connected to discussions on public deliberation and the modes of ensuring individual as well as collective reflection on decision-making and democratic innovation in the face of complex problems (John et al. 2009; Fischer 2003; Dryzek 2005). It also resonates with recent reformulations of the basic ideas of behavioural public policy. Some authors have argued that the division between system 1 and system 2, between automatic and reflective thinking is much too strict (Fiedler and von Sydow 2015; Bago and Neys 2017; Gigerenzer and Gaissmaier 2011). Nudges supposed to trigger ‘fast and frugal’ heuristics involve some kind of thinking and awareness of choices. Many nudgetype policies such as commitment devices or health interventions are ‘thought-provoking’ in terms of individual and social implications (John 2018). They initiate a complex interaction and back-and-forth movement between system 1 and system 2. This might have larger conceptual consequences, because it signifies a shift from focusing on behaviour as a more or less unconscious and automatic pattern on the individual level to action as an anticipation of how both one’s own behaviour and the behaviour of others materialize and have social consequences in the future (Schütz 1959). Turning from behaviour to action might help behavioural approaches to overcome unnecessary dichotomies and open up interdisciplinary venues of combining cognitive and behavioural sciences with social theories of action (Elster 2015; Giddens 1984).

M4716-STRAßHEIM_9781785367847_t.indd 17

25/04/2019 13:32

18  Handbook of behavioural change and public policy Finally, one of the greatest challenges in this context is connected to the digitalization of behavioural public policy. Based on large amounts and multiple sources of data on individual behaviour and its contexts, machine learning algorithms and other modes of pattern recognition are increasingly used in decision-making and the ‘personalized’ design of instruments and regulations. They provide information for the prediction of individual behaviour and establish an infrastructure for the creation of targeted interventions (Andrews et al. 2017; Sunstein 2015; Hacker 2017). The consequences of these developments are largely unclear. Proponents argue that digitalization helps to reduce the cognitive burden on people and support individual decision-making. Confronted with increasing technological and social acceleration and complex informational environments people tend to make suboptimal decisions. From this perspective, digital behaviour change interventions could support people in winning back authority and authenticity (Sunstein 2015). Behavioural public policy based on data science might, however, even increase problems of individual choice (Margetts et al. 2014). Using citizens as ‘walking data generators’ could undermine their self-image (McAfee and Brynjolfsson 2012). Algorithmic regulation could limit the possibilities of individual and collective learning (Newell and Marabelli 2014). Behavioural strategies based on big data could perpetuate social asymmetries and lead to ‘smart discrimination’ (Hacker and Petkova 2017). Moreover, technologies of political micro targeting could open up new venues for the manipulation and the ‘redlining’ of specific groups such as non-voters (Bodó et al. 2017). The promises and pitfalls of digitalization point to the many open questions related to the ever-expanding research area of behavioural change and public policy.

NOTES 1. The term ‘behavioural public policy’ has been introduced by Adam Oliver (2013). It is also the name of a journal that is ‘devoted to behavioural research and its relevance to public policy’ (see www.cambridge. org/core/journals/behavioural-public-policy, last access 15 April 2018). In this Introduction, we use it as a synonym to similar terms such as ‘behavioural governance’ or ‘behavioural regulation’. 2. See the chapters in this handbook. On the rise of behavioural public policy in the UK see the excellent study by Jones et al. (2013) and John (2018). Whitehead et al. (2017) have recently presented a comparison of international case studies on what they call ‘neuroliberalism’. The European Commission’s report ‘Behavioural Insights Applied to Policy’ describes initiatives from member states of the European Union and their institutional context (Lourenço et al. 2016). Drawing on responses from a survey conducted among 60 public bodies in 23 OECD and partner countries and two international organizations the OECD report Behavioural Insights and Public Policy: Lessons from Around the World (OECD 2017), gives an overview on more than 100 applications. Based on an international survey, Sunstein et al. (2018) compare citizen’s attitudes towards ‘nudges’ covering eight countries. Alemanno and Sibony (2015) discuss the role of behavioural approaches in the European context. 3. We reject the notion to be found in some textbooks that the general purpose of public policy is to change behaviour. Public policy has many purposes ranging from organizational and institutional transformations and the enhancement of collective learning and reflexivity to the attainment of normative goals such as equal distribution or freedom of speech. Reducing it to behavioural change might already be a symptom of the increasing influence of behaviouralism since the 1960s. 4. Sunstein referring to Rawls in a presentation at the German Federal Ministry of Justice on 12 December 2014. 5. While research on governmentality in the tradition of Foucault has made major contributions to understanding the relationship between behavioural change and public policy (Miller and Rose 1994; Rose and Miller 2008), we are sceptical of a tendency in this literature to treat the ‘neoliberal regime of governmentality’ as an unquestioned vantage point of analysis. This tendency has been criticized from within the Foucauldian debate as ‘cookie cutter explanation’, that is, the reproduction of the same argument over and over again

M4716-STRAßHEIM_9781785367847_t.indd 18

25/04/2019 13:32

Introduction  19 (Brady 2014). Approaches such as this run the risk of underestimating the inherent contradictions, multiple dynamics and counter movements in complex configurations of governance.

REFERENCES Abbott K.W. (2012) The transnational regime complex for climate change. Environment and Planning C: Government and Policy 30: 571‒590. Alemanno A. (2012) Nudging smokers: the behavioural turn of tobacco risk regulation. European Journal of Risk Regulation 1/2012: 32‒42. Alemanno A. and Sibony A.-L. (2015) Nudge and the Law: A European Perspective. Oxford: Hart Publishing. Andrews L., Benbouzid B., Brice J., et al. (2017) Algorithmic Regulation (Discussion Paper 85), London: London School of Economics. APSC. (2007) Changing Behaviour. A Public Policy Perspective. Canberra: Australian Public Service Commission. Bago B. and Neys W.D. (2017) Fast logic?: examining the time course assumption of dual process theory. Cognition 158: 90–109. Banerjee A.V. and Duflo E. (2012) Poor Economics. A Radical Rethinking of the Way to Fight Global Poverty. New Delhi: Random House India. Beck S. (2015) Scientific experts. In: Lövbrand E. and Bäckstrand K. (eds) Research Handbook on Climate Governance. Cheltenham, UK and Northampton, MA: Edward Elgar Publishing, 286‒296. Beck S. and Mahoney M. (2018) The IPCC and the new map of science and politics. WIREs Climate Change https://doi.org/10.1002/wcc.547. BIT. (2018) Behavioural Government. Using Behavioural Science to Improve How Governments Make Decisions. London: Behavioural Insights Team. Bodó B., Helberger N. and de Vreese C. (2017) Political micro-targeting: a Manchurian candidate or just a dark horse? Internet Policy Review 6. Botzem S. and Straßheim H. (2016) On Her Majesty’s Service? The Behavioural Insights Team and its role in the global rise of behavioural change policies (Paper for the 44th ECPR Joint Sessions 2016, Workshop No. 2: Behavioural Change and Public Policy), Berlin/Bremen. Brady M. (2014) Ethnographies of neoliberal governmentalities: from the neoliberal apparatus to neoliberalism and governmental assemblages. Foucault Studies 18: 11‒33. Brunsson N. (2009) Reform as Routine: Organizational Change in the Modern World. Oxford: Oxford University Press. COSU. (2004) Personal Responsibility and Changing Behaviour: The State of Knowledge and its Implications for Public Policy (David Halpern and Clive Bates, Geoff Mulgan and Stephen Aldridge with Greg Beales and Adam Heathfield). London: Cabinet Office Strategy Unit. Datta S. and Mullainathan S. (2014) Behavioural design: a new approach to development policy. Review of Income and Wealth 60: 7‒35. Dryzek J.S. (2005) The Politics of the Earth: Environmental Discourses. Oxford: Oxford University Press. Dunlop C.A. (2017) Pathologies of policy learning: what are they and how do they contribute to policy failure? Policy & Politics 45: 19‒37. Elster J. (2000) Ulysees Unbound. Studies in Rationality, Precommitment, and Constraint. Cambridge: Cambridge University Press. Elster J. (2015) Explaining Social Behaviour. More Nuts and Bolts for the Social Sciences (revised edition). Cambridge: Cambridge University Press. EU Commission. (2018) Making sense of science under conditions of complexity and uncertainty. Scoping paper adopted by the High Level Group of Advisors (https://ec.europa.eu/research/sam/index.cfm?pg=science, last access 20 September 2018), 3‒4. Felsen G., Castelo N. and Reiner P.B. (2013) Decisional enhancement and autonomy: public attitudes towards overt and covert nudges. Judgement and Decision Making 8: 202‒213. Fiedler K. and von Sydow M. (2015) Heuristics and biases: beyond Tversky and Kahneman’s (1974) Judgment under Uncertainty. In: Eysenck M.W. and Groome D. (eds) Cognitive Psychology: Revisiting the Classical Studies. Los Angeles/London/New Delhi/Singapore: Sage, 146‒161. Fischer F. (2003) Reframing Public Policy. Oxford: Oxford University Press. Fischer F. (2009) Democracy and Expertise. Reorienting Policy Inquiry. Oxford: Oxford University Press. Fleischer L., Smith C. and Viac C. (2016) A Review of General Social Surveys (OECD Statistics Working Papers 2016/09), Paris: OECD Publishing. Giddens A. (1984) The Constitution of Society. Cambridge: Blackwell Publishers. Gigerenzer G. (2015) Risk Savvy: How To Make Good Decisions. New York: Penguin Books.

M4716-STRAßHEIM_9781785367847_t.indd 19

25/04/2019 13:32

20  Handbook of behavioural change and public policy Gigerenzer G. and Gaissmaier W. (2011) Heuristic decision making. Annual Review of Psychology 62: 451–482. Grüne-Yanoff T. and Hertwig R. (2016) Nudge versus boost: how coherent are policy and theory? Minds and Machines 26: 149–183. Habermas J. (2001) The Postnational Constellation. Cambridge: MIT Press. Hacker P. (2017) Personalizing EU private law: from disclosures to nudges and mandates. European Review of Private Law 25: 651‒677. Hacker P. and Petkova B. (2017) Reining in the big promise of big data: transparency, inequality, and new regulatory frontiers. Northwestern Journal of Technology and Intellectual Property 15. Hallsworth M., Snijders V., Burd H., et al. (2016) Applying behavioural insights. Simple ways to improve health outcomes (Report of the WISH Behavioral Insights Forum 2016), Doha: World Innovation Summit for Health. Halpern D. (2009) The Hidden Wealth of Nations. Cambridge/Malden: John Wiley & Sons. IPSP. (2018) Rethinking Society for the 21st Century. Report of the International Panel on Social Progress (3 volumes). Cambridge: Cambridge University Press (im Erscheinen). Jasanoff S. (2004a) The idiom of co-production. In: Jasanoff S (ed.) States of Knowledge. The Co-production of Science and Social Order. London/New York: Routledge, 1‒12. Jasanoff S. (2004b) States of Knowledge. The Co-production of Science and Social Order. London/New York: Routledge. Jasanoff S. (2013) Epistemic subsidiarity: coexistence, cosmopolitanism, constitutionalism. European Journal of Risk Regulation 2/2013: 133‒141. John P. (2013) All tools are informational now: how information and persuasion define the tools of government. Policy & Politics 41: 605‒620. John P. (2018) How Far to Nudge? Assessing Behavioural Public Policy. Cheltenham, UK and Northampton, MA: Edward Elgar Publishing. John P., Smith G. and Stoker G. (2009) Nudge nudge, think think: two strategies for changing civic behaviour. The Political Quarterly 80: 361‒370. Jones R., Pykett J. and Whitehead M. (2013) Changing Behaviours. On the Rise of the Psychological State. Cheltenham, UK and Northampton, MA: Edward Elgar Publishing. Jordan A.J., Huitema D., Hildén M., Van Asselt H., Rayner T.J., Schoenefeld J.J., Tosun J., Forster J. and Boasson E.L. (2015) Emergence of polycentric climate governance and its future prospects. Nature Climate Change 5(11): 977. Jung J.Y. and Mellers B.A. (2016) American attitudes towards nudges. Judgement and Decision Making 11: 62‒76. Kahneman D. (2011) Thinking, Fast and Slow. London: Penguin Books. Kahneman D., Slovic P. and Tversky A. (1982) Judgement Under Uncertainty: Heuristics and Biases. Cambridge/ New York/Sidney: Cambridge University Press. Keohane R.O. and Victor D.G. (2011) The regime complex for climate change. Perspectives on Politics 9: 7‒23. Kuyper J., Bäckstrand K. and Schroeder H. (2017) Institutional accountability of nonstate actors in the UNFCCC: exit, voice, and loyalty. Review of Policy Research 34: 88‒109. Le Grand J. and New B. (2015) Government Paternalism. Nanny State or Helpful Friend? Princeton/Oxford: Princeton University Press. Lodge M. and Wegrich K. (2016) The rationality paradox of nudge: rational tools of government in a world of bounded rationality. Law & Policy 38: 250‒267. Lourenço J.S., Ciriolo E., Almeida S.R., et al. (2016) Behavioural Insights Applied to Policy: European Report 2016. Brussels: European Commission. Margetts H., Hale S.A. and Yasseri T. (2014) Big data and collective action. In: Gragam M. and Dutton W. (eds) Society and the Internet: How Information and Social Networks are Changing Our Lives. Oxford: Oxford University Press, 223‒238. McAfee A. and Brynjolfsson E. (2012) Big data: the management revolution. Harvard Business Review 90: 60‒68. Miller P. and Rose R. (1994) On therapeutic authority: psychoanalytical expertise under advanced liberalism. History of Human Science 7: 29‒64. Newell S. and Marabelli M. (2014) Strategic opportunities (and challenges) of algorithmic decision-making: a call for action on the long-term societal effects of ‘datification’. The Journal of Strategic Information Systems 24: 3‒14. O’Donnell C., Deaton A., Durand M., et al. (2014) Wellbeing and Policy. London: Legatum Institute. OECD. (2015) Behavioural Insights and New Approaches to Policy Design. Summary of an International Seminar. Paris: OECD Publishing. OECD. (2017) Behavioural Insights and Public Policy: Lessons from Around the World. Paris: OECD Publishing. Offe C. (1992) Bindings, shackles, brakes: on self-limitation strategies In: Honneth A., McCarthy T., Offe C., et al. (eds) Cultural-Political Interventions in the Unfinished Project of Enlightenment. Cambridge, MA: MIT Press, 63‒95. Oliver A. (2013) Behavioural Public Policy. Cambridge: Cambridge University Press.

M4716-STRAßHEIM_9781785367847_t.indd 20

25/04/2019 13:32

Introduction  21 Oliver A. (2015) Nudging, shoving, and budging: behavioural economic-informed policy. Public Administration 93: 700‒714. Peeters M., Megens C., van den Hoven E., et al. (2013) Social stairs: taking the piano staircase towards long-term behavioral change. In: Berkovsky SFJ (ed.) Persuasive Technology. PERSUASIVE 2013 (Lecture Notes in Computer Science, vol 7822). Berlin/Heidelberg/New York: Springer, 174‒179. Rose N. and Miller P. (2008) Governing the Present: Administering Economic, Social and Personal Life. Cambridge/Malden: Polity Press. Schütz A. (1959) Tiresias, or our knowledge of future events. Social Research 26: 71‒89. Sent E.-M. (2004) Behavioral economics: how psychology made its (limited) way back into economics. History of Political Economy 36: 735‒760. Shafir E. (2013) The Behavioral Foundations of Public Policy. Princeton: Princeton Univ. Press. Straßheim H. (2017a) Die Globalisierung der Verhaltenspolitik. Jahrbuch Normative und institutionelle Grundfragen der Ökonomik Band 16: Kapitalismus, Globalisierung, Demokratie: 211‒242. Straßheim H. (2017b) Trends towards evidence-based policy formulation. In: Howlett M and Mukherjee I (eds) Handbook of Policy Formulation. Cheltenham, UK and Northampton, MA: Edward Elgar Publishing, 504‒521. Straßheim H., Jung A. and Korinek R-L. (2015) Reframing Expertise: the rise of behavioural insights and interventions in public policy. In: Berthoin Antal A, Hutter M and Stark D (eds) Moments of Valuation. Exploring Sites of Dissonance. Oxford: Oxford University Press, 249‒268. Sugden R. (2016) Do people really want to be nudged towards healthy lifestyles? International Review of Economics 64: 113‒123. Sunstein C.R. (2015) Choosing Not to Choose: Understanding the Value of Choice. Oxford/New York: Oxford University Press. Sunstein C.R. (2017) Nudges that fail. Behavioural Public Policy, 1: 4‒25. Sunstein C.R., Reisch L.A. and Rauber J. (2018) A worldwide consensus on nudging? Not quite, but almost. Regulation and Governance 12: 3‒22. Sunstein C.R. and Thaler R.H. (2003) Libertarian paternalism is not an oxymoron. The University of Chicago Law Review 70: 1159‒1202. Thaler R.H. and Sunstein C.R. (2008) Nudge. Improving Decisions About Health, Wealth and Happiness. London: Penguin Books. Tomlinson M.W. and Kelly G.P. (2013) Is everybody happy? The politics and measurement of national wellbeing. Policy and Politics 41: 139‒157. United Nations. (2017) Consuming Differently, Consuming Sustanably: Behavioural Insights for Policymaking. New York: United Nations Environment Programme. Weaver K. (2016) Getting people to behave: research lessons for policy makers. Public Administration Review 75: 806–816. Whitehead M., Jones R., Lilley R., et al. (2017) Neuroliberalism: Behavioural Government in the Twenty First Century. New York/London: Routledge. World Bank. (2015) Mind, Society and Behavior. Washington: World Bank.

M4716-STRAßHEIM_9781785367847_t.indd 21

25/04/2019 13:32

PART I KNOWING AND CHANGING BEHAVIOUR: THE CO-EVOLVEMENT OF BEHAVIOURAL SCIENCE AND BEHAVIOURAL PUBLIC POLICY

M4716-STRAßHEIM_9781785367847_t.indd 22

25/04/2019 13:32

2.  Nudging before the nudge? Behavioural traffic safety regulation and the rise of behavioural economics Rüdiger Graf

1. INTRODUCTION Over the last decade, behavioural economists or, more generally, experts with a background in the behavioural sciences have made significant inroads into political consulting and policymaking – both nationally and internationally (The Cabinet Office 2010; van Bavel et al. 2013; Whitehead et al. 2014). Advocates of behavioural approaches to public policy usually claim that their strategies differ fundamentally from older instruments designed to influence citizens’ behaviour. Traditionally, the ‘regulatory toolbox’ consisted of laws, financial incentives and educational measures, all addressing the rational decision-maker who was supposed to avoid punishment, realize gains or act in accordance with certain principles (Hood 1983; Lascoumes and Le Galès 2007). By contrast, behavioural experts draw on ‘academic research in the fields of behavioural economics and psychology which show how often subtle changes to the way in which decisions are framed can have big impacts on how people respond to them’ (Sunstein 2017, p. 199). They usually also claim that their behavioural interventions are quite new, require little investment and are empirically testable (World Bank 2015, p. 13). Protagonists of the academic subdiscipline of behavioural economics, which has emerged since the late 1970s, have usually emphasized the political importance of their research. Editing the first Handbook of Behavioral Economics in 1986, Benjamin Gilad and Stanley Kaish argued that, by analysing ‘real behaviour with its foundation of real motives, real aspirations, and empirically determined dynamics, we have a better chance of judging whether public policy achieves the positive goals its proponents intended’ (Gilad and Kaish 1986, p. xx). According to this view, behavioural economics suggests a ‘new rationale for government intervention in the economy, given the failure of markets to promote a classical optimization due to individual judgment bias’ (Gilad and Kaish 1986, p. xx). Despite these early claims for political relevance by behavioural economists, the current situation should not be described as one of renewed economic imperialism (Radnitzky and Bernholz 1987). It is rather a case of a multi- and transdisciplinary field of experts on behaviour offering advice to policy makers (Shafir 2013). The tendency to describe their interventions as inspired by behavioural economics is most likely an attempt to invest psychological tactics with the some of the lustre of academic economics, which is still held in high esteem. The academic diversity of behavioural change policies notwithstanding, they all share a similar conceptualization of the citizens whose behaviour is to be regulated. They do not understand people as fully self-transparent decision-makers who are able to live up to the ideal of homo economicus that dominated neoclassical economics as well as early 23

M4716-STRAßHEIM_9781785367847_t.indd 23

25/04/2019 13:32

24  Handbook of behavioural change and public policy decision theories (Wilson and Dixon 2012). Since the 1950s, and increasingly since the late 1970s, behavioural economists have rejected both the empirical adequacy and theoretical usefulness of homo economicus. They maintain that ‘the standard economic model of human behaviour includes (at least) three unrealistic traits: unbounded rationality, unbounded willpower, and unbounded selfishness’ (Mullainathan and Thaler 2001, p. 1095). Behavioural economists, by contrast, only credit individuals with ‘bounded rationality’, using an expression coined by Herbert A. Simon (Simon 1972). In their view, people make decisions on the basis of their limited intellectual capacities within restricted time-frames while concrete circumstances influence the outcome in ways that are not necessarily accessible to them. According to Daniel Kahneman and Amos Tversky, these decision-making processes follow certain ‘heuristics and biases’ that lead to systematic and predictable diversions from the rules of economic logic, which, however, they still uphold as a normative standard (Tversky and Kahneman 1974). On the other hand a group of authors, like for example Gerd Gigerenzer, reject the universality of the economic standard of rationality arguing that the fast and frugal heuristics that human beings actually use could and should be examined and improved (Grüne-Yanoff and Hertwig 2016). The ‘heuristics and biases’ approach currently dominates. Above all, the broader public debate on behavioural policy tools concentrates on Richard Thaler and Cass Sunstein’s concept of ‘nudging’ and their political theory of ‘libertarian paternalism’ (Thaler and Sunstein 2003; 2009). Nudges differ from other policy tools inasmuch as they are not supposed to convince the rational decision-maker by giving him or her reasons to act. In order to function they may – though they do not have to – enter people’s explicit deliberative processes. Generally speaking, laws address the rational, self-controlled agent threatening him or her with punishment; economic incentives speak to the selfish utility-maximizer promising gains or threatening losses; and educational measures are dedicated to the enlightened and strong-willed individual who can live up to certain principles. By contrast, behavioural policy instruments such as nudges address the decision-maker in more subtle ways. They are designed to influence the outcome of decision-making processes by framing them in particular ways, setting default options and changing the architecture or environment of the decision. These changes may be made explicit – and there is a long debate on whether they should be and how that influences their effectiveness – but, unlike laws, incentives and education, their functioning does not necessarily depend on the actors’ awareness (Kemmerer et al. 2016). In the first part of this chapter, I will briefly describe the emergence of behavioural economics as an academic subdiscipline in order to accentuate the difference between behavioural and other approaches to influencing human activity. Second, I will show that behavioural approaches to public policy are not as novel as their advocates commonly claim.1 Focusing on the example of traffic safety regulation, I will show that subtle strategies to influence behaviour, which could be described as nudges, were widely used in the second half of the twentieth century. They were part of a broader trend to observe and regulate human behaviour that stemmed from the rise of the behavioural sciences more generally. Moreover, and in contrast to the current utopias and dystopias of an impending ‘nudge-world’ (Saint-Paul 2011), the example of traffic safety regulation also suggests that behavioural strategies will remain simply one instrument among others in the regulatory toolbox. Behavioural economics and behavioural regulation have been co-evolving since the late 1970s. In the concluding section, I will argue that they became increasingly

M4716-STRAßHEIM_9781785367847_t.indd 24

25/04/2019 13:32

Nudging before the nudge?  25 attractive because they offered new regulatory strategies in an age of deregulation and marketization (Rodgers 2011). Finally, describing and analysing the difficulties that economic actors have in behaving as entrepreneurial selves in accordance with the demands of neoliberalism, behavioural economics casts doubt on Michel Foucault’s analysis of the ascent of a neoliberal governmentality (Bröckling 2012; Rose 1989).

2. THE EMERGENCE OF BEHAVIOURAL ECONOMICS AS AN ACADEMIC FIELD The spread of behavioural insights teams in many governments and debates on the importance of behavioural economics for political regulation have also sparked an interest in its history as a discipline (Sent 2004; Heukelom 2014). The term was coined in the context of George Katona’s research on consumer behaviour at the University of Michigan in the 1940s (Earl 1988, p. 5; Gilad et al. 1988, p. 438). Using tests and survey methods, Katona’s research projects (re)introduced social psychology into economics as he tried to determine the psychological factors that influenced patterns of consumption (Sent 2004, p. 740). Whereas Katona’s concept of behavioural economics was confined to his research on consumer behaviour, Herbert A. Simon’s approach was broader, amounting to a fundamental transformation of economics. Being based at the University of Chicago, Simon embodied the transdisciplinarity that characterizes the behavioural paradigm in general. Starting as a political scientist in the 1930s, he worked in organizational sociology, psychology, economics and computer science in order to ‘set forth a consistent body of theory of the rational and nonrational aspects of human behaviour in a social setting’ (Simon 1957, p. VII). For this endeavour, homo economicus did not seem to be a very promising starting point. In Simon’s view, human decision-makers should be understood less as gods and more as rats, modelled on the species of laboratory animals preferred for use in behaviourist experiments (Crowther-Heyck 2005, p. 6; Lemov 2005). The conceptual starting point for the analysis of an actor’s decisions as well as for economics should be his or her ‘bounded rationality’: ‘We substitute for “economic man” or “administrative man” a choosing organism of limited knowledge and ability. This organism’s simplifications of the real world for purposes of choice introduce discrepancies between the simplified model and the reality; and these discrepancies, in turn, serve to explain many of the phenomena of organizational behavior’ (Simon 1955, p. 114). Rather than performing calculations with an omniscient utility maximizer, Simon suggested that the actual decision-making processes of real people in concrete situations should be scrutinized. In line with the trends of ‘Cold War science’ and ‘Cold War rationality’, however, Simon did not focus on human beings but on ‘decision-making organisms’, looking for universal principles of behaviour that could be found in both animals and humans as well as constructed in machines (Erickson et al. 2013). In his search for principles of behaviour, Simon was a towering figure in the social sciences in the United States, but his influence on economics was, at first, comparatively small (Camerer et al. 2004, p. 6; Pooley and Solovey 2010). Professional economists like his Chicago colleague and even housemate Milton Friedman mostly rejected Simon’s frontal attack on homo economicus (Simon 1991, p. 93). Individual economists, however, took up Simon’s ideas and pursued the line of experimental economics further (Mirowski

M4716-STRAßHEIM_9781785367847_t.indd 25

25/04/2019 13:32

26  Handbook of behavioural change and public policy 2002), even internationally, like the German economists Heinz Sauermann and Reinhard Selten (Ockenfels and Sadrieh 2010; Selten 1995; Tietz 1990). In addition, and at first independently of the developments in the United States, the German ordoliberal economist Günter Schmölders developed a programme of ‘economic behavioural research’ founding the Institute for empirical socio-economics, Forschungsstelle für empirische Sozialökonomik, at the University of Cologne in 1953 (Schmölders 1953, 1963). As one of West Germany’s most eminent taxation experts, who advised the Ministry of Finance from 1947 to 1972, Schmölders surveyed people’s financial behaviour in order to develop a ‘financial psychology’ that could inform taxation policy (Schmölders 1975). Despite all their differences in intellectual scope, ambition and style, Simon’s and Schmölders’ behavioural economics had many commonalities. Both rejected the ideal of homo economicus as inadequate, while suggesting that scholars empirically analyse the decisions of real people in actual situations. Deviating from the methodological individualism that focused on the individual choice act, they looked for principles of behaviour that could be found in larger groups and made individual actions predictable (Schmölders 1971, p. 13). As they did not consider these behavioural patterns specific to human beings, both looked at neighbouring disciplines for methodological help and theoretical inspiration. Schmölders, for example, wanted to open economics up to psychology, biology, neurology, sociology, history, social anthropology, linguistics and even animal ethology (Schmölders 1953). Whereas both Simon’s and Schmölders’ influence remained limited, behavioural economics took off as an academic subject from the second half of the 1970s. The term ‘bounded rationality’ appeared more frequently in academic journals (Klaes and Sent 2005), the number of publications on behavioural economics increased rapidly, the first handbooks came out, new journals appeared and papers on behavioural economics entered mainstream journals. While Schmölders took no part in these developments, Simon, who won the Alfred Nobel Memorial Prize in Economics in 1978, was commonly considered a forerunner and gave his credentials to the new discipline (Kahneman et al. 1982; Simon 1986). Crucial for the emergence of the field, however, were the psychologists Daniel Kahneman and Amos Tversky who collaborated first at the Hebrew University of Jerusalem and later in the United States. Retrospectively, younger behavioural economists describe the publication of their ‘heuristics and biases’ approach as well as their more narrowly economic Prospect Theory as seminal events. With Thomas Kuhn’s theory of scientific paradigm shifts in mind, Richard Thaler remembers: ‘I was constantly confronted with the contrast between the models my colleagues were constructing and the behavior I was so frequently observing.’ Prospect Theory then offered the new paradigm: ‘The theory of decision making that was proposed there … could make sense out of many of the examples I was carrying around in my head’ (Thaler 1991, p. xiif.; 2015). While this narrative may retrospectively glorify the events, Kahneman and Tversky clearly did create one school of behavioural economics that empirically and experimentally scrutinized the ways in which human decision-making deviated from the standard of economic rationality. Current strategy papers on the application of behavioural insights to public policy draw, above all, on five of these heuristics and biases. First, the Kahneman and Tversky school argues that people tend to miscalculate probabilities systematically and predictably, not because of negligence but because their calculus is influenced by factors that should be irrelevant while relevant factors are overlooked (Tversky and Kahneman 1971; 1974; 1981). Second, in many concrete choice situations people tend to maintain

M4716-STRAßHEIM_9781785367847_t.indd 26

25/04/2019 13:32

Nudging before the nudge?  27 the status quo. Hence, the design of a choice problem and the default option influence the outcome of the decision (Samuelson and Zeckhauser 1988). This would not happen to homo economicus who would not overestimate his or her own chances in a concrete situation either. Yet, studies by behavioural economists show, third, that overconfidence and undue optimism are common phenomena, influencing concrete decisions (Oskamp 1982; Weinstein 1989). Fourth, it is not easy to reconcile altruism and fairness with the selfish profit maximization of homo economicus. Yet, experiments seem to establish that perceptions of fairness can supersede profit orientation and often do so (Fontaine 2007). Finally, people choose inconsistently over time when costs and benefits are distributed unequally in the nearer and more distant future (Loewenstein and Elster 1992). Why have these issues been increasingly interesting topics for economists since the 1980s? Why did behavioural economics emerge then and not earlier – for example back in the 1950s in reaction to the work of Simon or Schmölders? From a history of science perspective, there were several reasons that encouraged a paradigm shift in the 1970s. As Floris Heukelom argues, Kahneman and Tversky were more easily accessible for mainstream economists because they continued to accept homo economicus as a yardstick for correct economic behaviour (Heukelom 2012b; 2014). Whereas Simon had wanted to abolish homo economicus altogether, they retained the model but focused on the areas where actual decision-making systematically diverged from it. The influence of this ‘normative-descriptive shuffle’ was all the bigger because Kahneman and Tversky spent much time and effort designing clever and intuitively appealing examples to reveal the rules behind unexpected behavioural patterns in the tradition of social psychology (Heukelom 2014, pp. 130‒132; Ross et al. 2010). Moreover, Heukelom suggests that the set of conferences and the special programme for behavioural economics that the Alfred P. Sloan and Russell Sage Foundations enabled in the 1980s endowed their participants with a ‘sense of mission’ (Heukelom 2012a). Apart from these factors, Heukelom also alludes to the more general rise of the behavioural sciences and the concept of behaviour in the second half of the twentieth century. After the Second World War, the Ford Foundation had developed a large funding scheme for the social sciences that was supposed to reconceptualize large areas of them as behavioural sciences (Pooley 2016; Pooley and Solovey 2010). Explicitly distancing this attempt from narrower concepts of behaviourism, members of different disciplines were supposed to dedicate their research to the explanation of human behaviour. While dreams of a comprehensive, singular ‘behavioural science’ were never realized (Alexander et al. 1956), this so-called behavioural revolution reconfigured the landscape of the social sciences in the United States. At the end of the 1960s, the programme’s chief administrator – who had been advised by Herbert A. Simon – judged in the International Encyclopedia of the Social Sciences that the term ‘Behavioural Sciences’ seemed ‘to be here to stay. The new field of the behavioural sciences will in all probability be ranked among the important intellectual inventions of the twentieth century’ (Berelson 1968, p. 44). Indeed, the most recent comparable standard work of reference was entitled International Encyclopedia of the Social & Behavioral Sciences. Most behavioural scientists in the second half of the twentieth century did not subscribe to the negation of mental states proposed by classical behaviourism. Nevertheless, their focus on ‘behaviour’ implied an approach to human activity that differed significantly from one using the concept of ‘action’. To quote Max Weber’s famous distinction in

M4716-STRAßHEIM_9781785367847_t.indd 27

25/04/2019 13:32

28  Handbook of behavioural change and public policy Wirtschaft und Gesellschaft [Economics and Society], the term behaviour encompasses all activities of a living organism, while action denotes only the type of behaviour that is endowed with a subjective meaning (Weber 1976). Accordingly, the concept of action refers to intentional actors and a first-person perspective, whereas behaviour privileges external observation. Actions have reasons; behaviour has causes. Developing sophisticated techniques of measurement and analysis, behavioural scientists naturalized older normative conceptions of behaviour. Moreover, most of them worked with a concept of rationality that bridged the difference between intended and unintended activity as well as those between humans and animals, questioning human exceptionalism (Catton Jr. and Dunlap 1978). During the heyday of the behavioural science movement from the 1950s to the 1980s, however, economists largely stood aside. It was not until the behavioural paradigm had become all-pervasive in the 1980s that economists increasingly started to focus on actual behaviour in concrete settings. Due to the earlier rise of the behavioural sciences, however, behavioural approaches were by that time already common practice in certain realms of political regulation (Jann 1981; Schuppert 2016). As the example of road safety will show, insurance companies, consumerist movements and public administration officials used biases brought about by availability, the status quo and overconfidence in order to reduce traffic fatalities and injuries long before behavioural economists coined the terms and nudging became a hotly debated topic.

3. NUDGING DRIVERS? PASSIVE SAFETY AND DRIVERS’ BEHAVIOUR IN TRAFFIC REGULATION Individual mass motorization over the twentieth century came at a high cost. In fact, ‘it is difficult to think of another technology that has, in the same short span, had such an impact on human life and death without having been intentionally designed to cause harm’ (Esbester and Wetmore 2015, p. 308). In its current ‘decade of action for road safety’ the World Health Organization has set itself the goal of saving five million lives by reducing the kinds of behaviour that are conducive to accidents and injuries (World Health Organization 2013). Especially during the post-war economic boom, rising automobilization caused a massive increase in the number of accidents, injuries and deaths in Western industrialized countries. From the invention of the automobile, its safety had been an important concern, yet the content of safety discourses and practices changed significantly over the course of the twentieth century. In general, four traffic safety regimes can be distinguished. (1) Until the end of the 1920s, drivers were held solely responsible for accidents. (2) From the 1920s to the 1960s, traffic experts tried to avoid accidents by a threefold strategy of highway engineering, education and enforcement (Norton 2015, pp. 321, 326). (3) A major shift occurred in the 1960s, when, in the face of dramatically increasing fatalities, experts started to focus on how to mitigate the effects of accidents on the occupants of vehicles (Norton 2015, p. 328; Trischler and Steiner 2008; Wetmore 2004). As a consequence of these debates, governments introduced various measures to increase the passive safety of cars in order to reduce the impact of the ‘second collision’ between the occupant and his or her own car, improving the ‘crashworthiness’ of the cars. However, only some of these measures were foolproof (Stieniczka 2006), while others

M4716-STRAßHEIM_9781785367847_t.indd 28

25/04/2019 13:32

Nudging before the nudge?  29 depended on the occupants’ behaviour so that the driver’s responsibility continued to be an important issue in traffic safety discourses. (4) Throughout the second half of the twentieth century, psychologists, social psychologists, behavioural scientists and behavioural economists discussed how to influence the drivers’ behaviour in order to reduce the number as well as the effects of accidents. On a global level, the World Health Organization (WHO) started to deal with traffic accidents in the 1950s because they constituted one of the most significant risks to health. Dedicating its annual World Health Day to road accidents in 1961, the WHO determined that automotive vehicles were the primary cause of accidents (World Health Day 1961). In the following year, it commissioned the British physician and government expert on Occupational Health and Traffic Safety, Leslie George Norman, to conduct a study on the Epidemiology, Control, and Prevention of accidents. Norman calculated that traffic accidents were the most frequent cause of death among men between the ages of 15 and 30 in highly motorized countries. He demanded more safety measures in order to avoid ‘loss to the community, for their education and training has been wasted’ (Norman 1962, pp. 9, 13). According to him, one of the main reasons for the increase in traffic accidents that occurred in the aftermath of the Second World War was what behavioural economists would describe as a lack of ‘availability’ (Tversky and Kahneman 1974). Traffic accidents were not mentally available to drivers because accidents did not receive the media coverage they deserved. In contrast to plane crashes and shipwrecks, they were widely dispersed and each of them produced comparatively few casualties. Therefore, drivers were overly confident that they would not be involved in an accident and drove less carefully. Personal experience could sustain this overconfidence for a long time as, statistically, an American car owner had to drive 400 000 miles before being involved in an accident. Therefore, Norman concluded, many drivers exhibited an ‘“it can‘t happen to me” attitude, which is not conducive to safety’ (Norman 1962, p. 11). In the light of the relative rarity of fatal traffic accidents, Norman suggested applying methods from epidemiology. Safety campaigns were supposed to function as a ‘vaccine’ against accidents by raising awareness of their possibility. While these campaigns were often classical educational measures, informing rational individuals about the risks they were taking, they also used availability heuristics, trying to make the abstract risk of traffic accidents more concrete. For example, it was common practice to place pictures of damaged cars next to statistics concerning the number of traffic accidents (World Health Day 1961, pp. 4, 12). Grassroots movements used similar strategies. On the occasion of World Health Day 1961, the journal of the United Nations Educational, Scientific and Cultural Organization (UNESCO) praised US citizens who had tried to visualize the annual death toll from traffic accidents in their county by lying down at a crossroads: ‘The sight of all these bodies sprawled on the highway around a battered automobile brings home perhaps more vividly than could any statistics the full meaning and measure of death on the road’ (World Health Day 1961, p. 19). In terms of behavioural economics, the aim of this performance was to increase the availability of traffic accidents in people’s minds in order to influence their assessment of the probability of being involved in an accident and to induce them to drive more safely. Simultaneously, however, psychologists questioned the positive effect of fear-arousing measures on drivers’ behaviour (Pfafferott 1973, p. 32). Shifting the focus from drivers’ behaviour to the technology they used, passive safety instruments dominated the discourse in the 1960s. In 1966, the US Congress issued the

M4716-STRAßHEIM_9781785367847_t.indd 29

25/04/2019 13:32

30  Handbook of behavioural change and public policy National Highway and Traffic Safety Act, institutionalizing an administration of the same name in order to evaluate and introduce safety measures. Probably most important were restraint systems such as the safety belt. Niels Bohlin’s three-point safety belt in particular decreased the risk of severe injuries significantly. As early as the 1950s, it was possible to install safety belts in cars, yet this required a conscious decision by the buyer and implied extra cost. Volvo was the first company to include safety belts as standard in their models and other, but not all, companies followed suit (Bergmann 2009; Weishaupt 1999). The US Traffic Highway and Safety Administration, however, made seatbelts mandatory for cars to be licensed and, in 1974, the West German government did the same. In beha­ vioural economic terms, first private companies and then the state changed the default for a driver’s decision to put on a belt. Drivers no longer had to make a conscious decision to have a seatbelt installed at extra cost; the seatbelt was already available for their use. This change in the status quo, however, did not change behaviour to a satisfactory degree. On the contrary, many studies for different countries show that, even after the mandatory installation of safety belts, the number of people actually using them remained low (Bundesministerium für Verkehr 1984, p. 12; Pfafferott 1973). In West Germany, a psychological study on ‘seatbelt behaviour’ (Gurtanlegeverhalten) concluded that, even with a changed default option and seatbelts in every car, only 10 per cent of drivers would use seatbelts when driving in the city and 30 per cent on the open road (Kroj et al. 1974, p. 8). Hence, increasing the passive safety of the cars did not abolish the need to influence drivers’ behaviour. In order to increase safety belt usage, the West German government chose a two-tier approach. On the one hand, it introduced a law in 1976 making the use of seatbelts mandatory. Due to intense opposition to seatbelts by drivers who saw their personal freedom as being illegitimately reduced, however, the government refrained from introducing a fine. As it turned out, making seatbelt use mandatory without fining people in case of noncompliance had a very limited behavioural effect. Accordingly, in order to achieve beha­ vioural change, the West German government commissioned psychologists to analyse the reasons for seat belt resistance and develop strategies to change seat belt behaviour. The German Council for Road Traffic Safety, in which insurance companies, car manufacturers and state traffic agencies worked together, conducted campaigns to influence people’s driving and seatbelt behaviour (Berger et al. 1973; Kroj et al. 1974; Volks 1978). Despite the questionable effects of deterrence, they still used TV spots with crash-test-dummies with and without seatbelts in order to demonstrate the risk of injuries in accidents (Saupe 2010, p. 112). Yet, as psychological studies had shown that drivers’ overconfidence in their ability to drive better than others was a major cause of their reluctance to use seatbelts, other campaigns directly addressed this bias. Campaign posters showed men and women wearing safety belts with slogans like ‘Successful people avoid risk’ or ‘Experts wear a seatbelt’ (‘Erfolgreiche meiden das Risiko’, ‘Könner tragen Gurt’). In the United States, the government even experimented with techniques to further reduce the human factor in the use of safety techniques. In 1973, ‘a federal standard required a belt system that allows the automobile to start only under certain conditions’ (Robertson 1975, p. 1320). There were two solutions: the interlock system made it impossible to start the car unless the seat belts were fastened, an alternative system used lights and noises to remind drivers to buckle up. While the interlock system was more successful in increasing the use of safety belts, it also met strong opposition. It was considered an

M4716-STRAßHEIM_9781785367847_t.indd 30

25/04/2019 13:32

Nudging before the nudge?  31 illegitimate infringement of personal freedom and many people tried to circumvent it. The buzzer/light system, however, only nudged drivers to fasten their seatbelts leaving them free to ignore the warning. Long before Kahneman and Tversky coined the notion, it was a ‘libertarian paternalistic’ alternative to both the paternalistic interlock system and a libertarian absence of any seatbelt rules (Latour 1996, pp. 28‒36). Avoiding the human factor and making safety completely passive was also the aim of the airbag, which became a standard piece of equipment in the 1980s. There was, however, a more radically libertarian opposition that questioned the life-saving effects of passive safety instruments altogether. In a study for the American Enterprise Institute, economist Sam Peltzman argued that the reduction in deaths on US roads could not be attributed to regulation by the Traffic and Highway Administration. Rather, he saw it as a consequence of demographic factors, a slower rise in income and car speed and a simultaneous increase in medical and car repair costs (Peltzman 1975, pp. 16f). Moreover, Peltzman discerned a risk of compensating behaviour among drivers: feeling safer with seatbelts, they took more risks while driving so that injuries and fatalities among pedestrians and cyclists increased (Peltzman 1975, p. 18; Robertson 1977). In West Germany, the combination of several measures to regulate drivers’ behaviour (changing the default option, making the use of seatbelts mandatory and addressing drivers’ overconfidence) did indeed change seatbelt behaviour. By the end of the 1970s, 45 per cent of drivers used seatbelts in the city and as many as 85 per cent on the open road (Bundesministerium für Verkehr 1984, p. 13). A new study by the Federal Agency for Roads and Traffic on behaviour with respect to the ‘installation and use of seatbelts’ concluded that further campaigns would not significantly increase the number of seatbelt users (Volks 1978, p. 133). Reacting to pressure from insurance companies, the Federal Government then introduced a fine. Thus, after legislative and behavioural strategies had been deployed, an economic incentive was introduced with the aim of further increasing the use of seatbelts. Altogether these measures succeeded in achieving seatbelt use rates of over 90 per cent. A significant change in behaviour was realized in only twenty years through a combination of laws, fines and education – but also by methods that resembled nudging. To sum up, traffic safety regulation is an example of the use of nudging techniques before nudging was actually conceptualized as a policy tool. In general, experts from the behavioural sciences and even some early behavioural economists exerted a considerable influence on traffic safety regulation and road construction. Traffic safety was not the only policy field in which behavioural measures to influence and change people’s conduct were discussed. They became prevalent too in environmental protection and health policies. In the early 1980s, for example, experts discussed not only legislative measures and financial incentives but also non-economic ‘prompts and cues’ to influence environmental behaviour (Fisher et al. 1984). Over the course of the decade, ‘health behaviour’ also became an important paradigm in medical research and intervention (Anderson 1988; Gochman 1988). This behavioural reorientation of regulation occurred largely without any reference to the field of behavioural economics, which was then gaining prominence. As both focused on individual behaviour and subtle means of influencing it, I will try to determine the underlying developments that can explain the simultaneous changes in both regulatory policy and economic theory.

M4716-STRAßHEIM_9781785367847_t.indd 31

25/04/2019 13:32

32  Handbook of behavioural change and public policy

4. CONCLUSION: THE REGULATION OF BEHAVIOUR IN AN AGE OF DEREGULATION Considering the simultaneous rise of behavioural economics and behavioural regulation in various policy fields, it does not seem appropriate to describe the recent spread of behavioural insights teams as resulting from the findings of Kahneman, Tversky and their colleagues. Rather we have to look for common root causes that explain both the theoretical reorientation of a growing number of professional economists and the extension of the regulatory toolbox since the 1970s. To begin with, both resulted from the success of the behavioural sciences in establishing ‘behaviour’ as the generic concept to describe human activity. As has been explained, the behavioural science movement proliferated into various fields of knowledge since the 1950s. When economists began to focus on behaviour in the 1970s, however, the behavioural science label had already lost its appeal (Pooley 2016; Smelser and Gerstein 2000). Yet, this exhaustion indicates not the failure but rather the success of the behavioural science project. Around 1980, the concept of behaviour had become all-pervasive and common-sensical, thereby losing its distinctive value. Politicians were already busy asking psychologists and behavioural experts for advice in designing regulatory strategies in several fields when a growing number of economists became willing to question standard presuppositions of rationality. The rise of the behavioural sciences in the second half of the twentieth century as well as the proliferation of the concept of behaviour to describe human activity both indicated and induced a new understanding of rationality and the self. Cybernetics, systems theory and early behavioural science (in the singular) were attempts to develop a unified theory of the animate and non-animate world that would offer means to influence and control societies and economies. With the simultaneous advance of the human, neuro, and brain sciences as well as with the so-called psycho-boom of the 1960s and 1970s (Tändler 2016), naturalized understandings of rationality gained further acceptance. With its emphasis on cognitive biases and systematic errors, behavioural economics might appear, at first sight, as a humanization of formerly idealized understandings of rationality. From this perspective, by spreading conceptions of ‘bounded rationality’, behavioural economists helped to overcome a specific ‘Cold War rationality’ that had led to hypertrophic conceptions of planning and control (Erickson et al. 2013). This reading gets more complicated, however, when taking into consideration the second process that fuelled the co-emergence of behavioural economics and behavioural regulation. Behavioural economics only became fashionable from the late 1970s onward because of the simultaneous trend towards marketization that began around that time. After older conceptions of how to plan social and economic processes had seemingly failed in the crises of the 1970s, the expansion of market mechanisms and the retreat of the state promised a solution to many problems under constrained budgets (Rodgers 2011). Yet when welfare provisions that had formerly been administered by the state were privatized, knowledge of the heuristics and biases of individuals became increasingly important. Behavioural economists analysed whether and how far people were actually capable of acting as competent participants in markets and making decisions that were in their own interests, particularly in their own long-term interests. Analysing the heuristics and biases of decision-makers in the welfare markets, they simultaneously developed means to influence their behaviour. In other words, after the Keynesian conception of planning had

M4716-STRAßHEIM_9781785367847_t.indd 32

25/04/2019 13:32

Nudging before the nudge?  33 seemingly failed in the 1970s and neoliberal marketization became dominant in the 1980s, behavioural economists now offered a means to increase the state’s capacity to regulate and intervene in the economy again. Replacing homo economicus with a psychologically more realistic model of economic activity, they complemented what Michel Foucault had described as ‘neoliberal governmentality’. Using economic principles to describe all kinds of human behaviour is not an invention of behavioural economics. For example, Gary S. Becker had also argued that economic methods could be applied to explain human behaviour in general, assuming that people were rational actors trying to maximize utility with respect to a stable preference system in different markets (Becker [1976] 1983). Like Milton Friedman and others, Becker did not claim that homo economicus was a description of what people actually did but only that it produced good predictions of how people would behave. Michel Foucault mainly drew on Becker’s writings when he defined neoliberal governmentality. Economics, according to Foucault, adopted a number of techniques that were ‘contemporarily practiced in the United States … and that were called behavioral techniques’ (Foucault and Sennelart 2006, p. 370). He argued that neoliberal strategies tried to govern homo economicus by setting economic incentives. They conceptualized human beings as ‘entrepreneurial selves’ that were constantly trying to adjust to the demands of the market. As a normative ideal, this concept dominated the managerial literature in the last decades of the twentieth century (Bröckling 2007). Extending Foucault’s diagnosis from the 1970s into the present accounts for the processes of privatization and marketization that occurred over the last decades. In the light of the simultaneous rise of behavioural economics, however, it has to be nuanced. Behavioural economists deal with the failure of individual economic actors to act as entrepreneurial selves due to individual judgement biases. To come back to the example of traffic safety regulation, drivers failed to buckle up because they miscalculated the risk of traffic accidents and were overconfident that they were able to drive better than others. The same cognitive biases, however, were then used in order to influence their behaviour and increase seatbelt use. Similarly, many behavioural interventionists argued that, due to cognitive biases, actors do not act in accordance with their own best interests in the welfare markets, for example failing to save enough money for retirement. Simultaneously, they promised to provide state actors with instruments to influence human behaviour. Thus, behavioural economists tried to enlarge the realm of state intervention, further opening the design of markets and the architecture of decision-making processes to political regulation. They could not, therefore, be smoothly integrated into the narrative of neoliberal governmentality, and it comes as no surprise that many neoliberals rejected the concept of ‘libertarian paternalism’. The French economist Gilles Saint-Paul even warned: ‘If current trends continue, I foresee a gradual elimination of individual freedom as “social science” makes progress in documenting behavioral biases, measuring happiness, and evaluating the effects of coercive policies’ (Saint-Paul 2011, p. 4). In the light of the above analysis this fear is unwarranted. Nudges alone did not induce the fundamental behavioural change in the realm of traffic safety, they had to be accompanied by laws and fines. In general, nudges and other behavioural strategies seem to work best not on their own, but in combination with other instruments from the regulatory toolbox such as laws, financial incentives and educational campaigns. Yet, they promise to augment state interventional capacities especially in areas in which stricter regulation

M4716-STRAßHEIM_9781785367847_t.indd 33

25/04/2019 13:32

34  Handbook of behavioural change and public policy could easily threaten basic principles of liberalism. This makes them not neutral but highly political instruments, which are attractive throughout the political spectrum as they promise to achieve large effects by small means.

NOTE 1. Halpern (2015) acknowledges that nudging has been around for a long time and even refers to traffic regulation but, then, jumps abruptly to the early 2000s; for a critique of the claim to novelty see Schuppert (2016).

REFERENCES Alexander, Franz et al. (1956), ‘Editorial: Behavioral Science, a New Journal’, Behavioral Science 1, 1–5. Anderson, Robert (ed.) (1988), Health Behaviour Research and Health Promotion, Oxford: Oxford University Press. Becker, Gary S. ([1976] 1983), Der ökonomische Ansatz zur Erklärung menschlichen Verhaltens, Tübingen: Mohr. Berelson, Bernard R. (1968), ‘Behavioral Sciences’, in David Sills and Robert K. Merton (eds), International Encyclopedia of the Social Sciences, vol. II, New York: Macmillan, pp. 41–45. Berger, Hermann-Josef, Herhard Bliersbach and Rolff G. Dellen (1973), ‘Zur Psychologie des Sicherheitsgurts‘, in Eberhard Löw von and zu Steinfurth (eds), Für und wider Sicherheitsgurte Faktor Mensch im Verkehr H. (15/16). Frankfurt am Main: Tetzlaff, pp. 51–150. Bergmann, Heike (2009), ‘Angeschnallt und los! Die Gurtdebatte der 1970er und 1980er Jahre in der BRD’, Technikgeschichte 76, 105–30. Bröckling, Ulrich (2007), Das unternehmerische Selbst: Soziologie einer Subjektivierungsform, Frankfurt am Main: Suhrkamp. Bröckling, Ulrich (ed.) (2012), Gouvernementalität der Gegenwart: Studien zur Ökonomisierung des Sozialen, Frankfurt am Main: Suhrkamp, 6th edition. Bundesministerium für Verkehr, Abteilung Straßenverkehr (1984), Der Sicherheitsgurt: Fragen und Antworten, Bonn. Camerer, Colin, George Loewenstein and Matthew Rabin (eds) (2004), Advances in Behavioral Economics. The roundtable series in behavioral economics, New York, Princeton, NJ: Russell Sage Foundation; Princeton University Press. Catton Jr., William R. and Riley E. Dunlap (1978), ‘Theories, Paradigms and the Primacy of the HEP-NEP Distinction’, American Sociologist 13 (4), 256–259. Crowther-Heyck, Hunter (2005), Herbert A. Simon: The Bounds of Reason in Modern America, Baltimore: Johns Hopkins University Press. Earl, Peter E. (ed.) (1988), Behavioural Economics: Volume II, Aldershot, UK and Brookfield, VT: Edward Elgar Publishing. Erickson, Paul, Judy L. Klein, Lorraine Daston, Rebecca Lemov, Thomas Sturm and Michael D. Gordin (2013), How Reason Almost Lost Its Mind: The Strange Career of Cold War Rationality, Chicago: University of Chicago Press. Esbester, Mike and Jameson M. Wetmore (2015), ‘Introduction: Global Perspectives on Road Safety History’, Technology and Culture 56 (2), 307–318. Fisher, Jeffrey D., Paul A. Bell and Andrew Baum (1984), Environmental Psychology, Philadelphia: Holt Rinehart and Winston, 2nd edition. Fontaine, Philippe (2007), ‘From Philanthropy to Altruism: Incorporating Unselfish Behavior into Economics, 1961‒1975’, History of Political Economy 39 (1), 1–46. Foucault, Michel and Michel Sennelart (2006), Sicherheit, Territorium, Bevölkerung: Vorlesung am Collège de France, 1977–1978, Frankfurt am Main: Suhrkamp. Gilad, Benjamin and Stanley Kaish (eds) (1986), Handbook of Behavioral Economics: Behavioral Microeconomics, Greenwich, CT: Jai Press. Gilad, Benjamin, Stanley Kaish and Peter D. Loeb (1988), ‘From Economic Behavior to Behavioral Economics: The Behavioral Uprising in Economics’ in Peter E. Earl (ed.), Behavioural Economics: Volume II, Aldershot, UK and Brookfield, VT: Edward Elgar Publishing, pp. 437–458. Gochman, David S. (ed.) (1988), Health Behaviour: Emerging Research Perspectives, New York: Plenum Press.

M4716-STRAßHEIM_9781785367847_t.indd 34

25/04/2019 13:32

Nudging before the nudge?  35 Grüne-Yanoff, Till and Ralph Hertwig (2016), ‘Nudge Versus Boost: How Coherent are Policy and Theory?’, Minds & Machines 26 (1‒2), 149–183. Halpern, David (2015), Inside the Nudge Unit: How Small Changes Can Make a Big Difference, London: Allen. Heukelom, Floris (2012a), ‘A Sense of Mission: the Alfred P. Sloan and Russell Sage Foundations’ Behavioral Economics Program, 1984–1992’, Science in Context 25 (2), 263–286. Heukelom, Floris (2012b), ‘Three Explanations for the Kahneman–Tversky Programme of the 1970s’, The European Journal of the History of Economic Thought 19 (5), 797–828. Heukelom, Floris (2014), Behavioral Economics: A History. Historical perspectives on modern economics, Cambridge: Cambridge University Press. Hood, Christopher C. (1983), The Tools of Government. Public policy and politics, London: Macmillan. Jann, Werner (1981), Kategorien der Policy-Forschung (Speyerer Arbeitshefte no. 37), Speyer. Kahneman, Daniel, Paul Slovic and Amos Tversky (1982), ‘Preface’, in Judgment under Uncertainty: Heuristics and Biases, Cambridge: Cambridge University Press, pp. X–XIII. Kemmerer, Alexandra, Christoph Möllers and Maximilian Steinbeis (eds) (2016), Choice Architecture in Democracies: Exploring the Legitimacy of Nudging, Baden-Baden: Nomos. Klaes, Matthias and Esther-Mirjam Sent (2005), ‘A Conceptual History of the Emergence of Bounded Rationality’, History of Political Economy 37 (1), 27–59. Kroj, Günter et al. (1974), Psychologische Forschung zum Sicherheitsgurt und Umsetzung ihrer Ergebnisse, Köln: Bundesanstalt für Straßenwesen. Lascoumes, Pierre and Patrick Le Galès (2007), ‘Introduction: Understanding Public Policy through Its Instruments: From the Nature of Instruments to the Sociology of Public Policy Instrumentation’, Governance: An International Journal of Policy, Administration, and Institutions 20 (1), 1–21. Latour, Bruno (1996), Der Berliner Schlüssel: Erkundungen eines Liebhabers der Wissenschaften, Berlin: Akademie-Verlag. Lemov, Rebecca M. (2005), World as Laboratory: Experiments with Mice, Mazes, and Men, New York: Hill and Wang. Loewenstein, George and John Elster (eds) (1992), Choice over Time, New York: Russell Sage Foundation. Mirowski, Philip (2002), Machine Dreams: Economics Becomes a Cyborg Science, Cambridge: Cambridge University Press. Mullainathan, S. and Richard H. Thaler (2001), ‘Behavioral Economics’, in International Encyclopedia of the Social & Behavioral Sciences, vol. II, Amsterdam: Elsevier, pp. 1094–1100. Norman, L.G. (1962), Road Traffic Accidents: Epidemiology, Control, and Prevention, Geneva: World Health Organization. Norton, Peter (2015), ‘Four Paradigms: Traffic Safety in the Twentieth-Century United States’, Technology and Culture 56 (2), 319–334. Ockenfels, Axel and Abdolkarim Sadrieh (eds) (2010), The Selten School of Behavioral Economics: A Collection of Essays in Honor of Reinhard Selten, Berlin, Heidelberg, New York: Springer. Oskamp, Stuart (1982), ‘Overconfidence in Case-Study Judgments’ in D. Kahneman, P. Slovic and A. Tversky (eds), Judgment under Uncertainty, Heuristics and Biases, Cambridge: Cambridge University Press, pp. 287–294. Peltzman, Sam (1975), Regulation of Automobile Safety, Washington: American Enterprise Institute for Public Policy Research. Pfafferott, Ingo (1973), ‘Problemfeld Sicherheitsgurt’, in Eberhard Löw von und zu Steinfurth (ed.), Für und wider Sicherheitsgurte, Frankfurt am Main: Tetzlaff, pp. 5–50. Pooley, Jefferson (2016), ‘A “Not Particularly Felicitous” Phrase: A History of the “Behavioral Sciences” Label’, Journal for the Sociology and History of the Social Sciences 1 (1), 38‒81. Pooley, Jefferson and Mark Solovey (2010), ‘Marginal to the Revolution: The Curious Relationship Between Economics and the Behavioral Sciences Movement in Mid-Twentieth-Century America’, History of Political Economy 42, annual supplement, 199–233. Radnitzky, Gerard and Peter Bernholz (1987), Economic Imperialism: The Economic Approach Applied Outside the Field of Economics, New York: Paragon House Publishers. Robertson, Leon S. (1975), ‘Safety Belt Use in Automobiles with Starter-Interlock and Buzzer-Light Reminder Systems’, American Journal of Public Health 65 (12), 1318–1325. Robertson, Leon S. (1977), ‘A Critical Analysis of Peltzman’s “The Effects of Automobile Safety Regulation”’, Journal of Economic Issues 11 (3), 587–600. Rodgers, Daniel T. (2011), Age of Fracture, Cambridge, MA: Belknap Press of Harvard University Press. Rose, Nikolas (1989), Governing the Soul: The Shaping of the Private Self, London, New York: Routledge. Ross, Lee, Mark Lepper and Andrew Ward (2010), ‘History of Social Psychology: Insights, Challenges, and Contributions to Theory and Application’, in Susan T. Fiske, Daniel T. Gilbert and Gardner Lindzey (eds), Handbook of Social Psychology, Hoboken NJ: Wiley, 5th edition, pp. 3–50. Saint-Paul, Gilles.(2011), The Tyranny of Utility: Behavioral Social Science and the Rise of Paternalism, Princeton, NJ: Princeton University Press.

M4716-STRAßHEIM_9781785367847_t.indd 35

25/04/2019 13:32

36  Handbook of behavioural change and public policy Samuelson, William and Richard Zeckhauser (1988), ‘Status Quo Bias in Decision Making’, Journal of Risk and Uncertainty 1, 7–59. Saupe, Achim (2010), ‘Human Security and the Challenge of Automobile and Road Traffic Safety: A Cultural Historical Perspective’, Historical Social Research 35 (4), 102–120. Schmölders, Günter (1953), ‘Ökonomische Verhaltensforschung‘, Ordo. Jahrbuch für die Ordnung von Wirtschaft und Gesellschaft 5, 203–244. Schmölders, Günter (1963) ‘10 Jahre sozialökonomische Verhaltensforschung in Köln‘, Ordo. Jahrbuch für die Ordnung von Wirtschaft und Gesellschaft 14, 259–273. Schmölders, Günter (1971), Der verlorene Untertan: Verhaltensforschung enthüllt die Krise zwischen Staatsbürger und Obrigkeit, Düsseldorf: ECON. Schmölders, Günter (1975), Einführung in die Geld- und Finanzpsychologie, Darmstadt: Wissenschaftliche Buchgesellschaft. Schuppert, Gunnar F. (2016), ‘Nudging: Neither a Novelty, nor a Promising Lead – Unless in Context’ in Alexandra Kemmerer, Christoph Möllers and Maximilian Steinbeis, Choice Architecture in Democracies: Exploring the Legitimacy of Nudging, Baden-Baden: Nomos, pp. 333–338. Selten, Reinhard (1995), ‘Autobiography’, in Tore Frängsmyr, The Nobel Prizes 1994, Stockholm, http://www. nobelprize.org/nobel_prizes/economic-sciences/laureates/1994/selten-bio.html (last accessed 30 September 2018). Sent, Esther-Mirjam (2004), ‘Behavioral Economics: How Psychology Made Its (Limited) Way Back into Economics’, History of Political Economy 36 (4), 735–760. Shafir, Eldar (ed.) (2013), The Behavioral Foundations of Public Policy, Princeton, NJ: Princeton University Press. Simon, Herbert A. (1955), ‘Behavioral Model of Rational Choice’, Journal of Economics 69, 99–118. Simon, Herbert A. (1957), Models of Man, New York: Garland. Simon, Herbert A. (1972), ‘Theories of Bounded Rationality’, in C.B. McGuire and Roy Radner (eds), Decision and Organization: A Volume in Honor of Jacob Marschak, Amsterdam, London: North-Holland, pp. 161–176. Simon, Herbert A. (1986), ‘Preface’, in Benjamin Gilad and Stanley Kaish (eds), Handbook of Behavioral Economics: Behavioral Microeconomics, Greenwich, CT: Jai Press, pp. XV–XVI. Simon, Herbert A. (1991), Models of My Life, New York: Basic Books. Smelser, Neil J. and Dean R. Gerstein (eds) (2000), Behavioral and Social Science: Fifty Years of Discovery, Washington, DC: National Academy Press. Stieniczka, Norbert (2006), Das “narrensichere” Auto: Die Entwicklung passiver Sicherheitstechnik in der Bundesrepublik Deutschland, Darmstadt: Wissenschaftliche Buchgesellschaft. Sunstein, Cass R. (2017), ‘Behaviorally Informed Regulation: Part 1’, in Roger Frantz (ed.), Routledge Handbook of Behavioral Economics, London: Routledge, pp. 199–209. Tändler, Maik (2016), Das therapeutische Jahrzehnt: Der Psychoboom in den 1970er Jahren, Göttingen: Wallstein. Thaler, Richard H. (1991), Quasi Rational Economics, New York: Russell Sage Foundation. Thaler, Richard H. (2015), Misbehaving: The Making of Behavioral Economics, New York, London: W.W. Norton & Company. Thaler, Richard H. and Cass R. Sunstein (2003), ‘Libertarian Paternalism’, The American Economic Review 93 (2), 175–179. Thaler, Richard H. and Cass R. Sunstein (2009), Nudge: Improving Decisions about Health, Wealth and Happiness, London: Penguin Books. The Cabinet Office (2010), ‘Mindspace: Influencing Behaviour Through Public Policy’, http://www.institutefor​ government.org.uk/publications/mindspace (last accessed 30 September 2018). Tietz, Reinhard (1990), ‘On Bounded Rationality: Experimental Work at the University of Frankfurt/Main’, Zeitschrift für die gesamte Staatswissenschaft 146, 659–672. Trischler, Helmuth and Kilian J. Steiner (2008), ‘Innovationsgeschichte als Gesellschaftsgeschichte: Wissenschaftlich konstruierte Nutzerbilder in der Automobilindustrie seit 1950‘, Geschichte und Gesellschaft 34, 455–488. Tversky, Amos and Daniel Kahneman (1971), ‘Belief in the Law of Small Numbers’, Psychological Bulletin 76, 105–110. Tversky, Amos and Daniel Kahneman (1974), ‘Judgment Under Uncertainty: Heuristics and Biases’, Science 185, 1124–1131. Tversky, Amos and Daniel Kahneman (1981), ‘The Framing of Decisions and the Psychology of Choice’, Science 211 (4481), 453–485. Van Bavel, René, Benedikt Herrmann, Gabriele Esposito and Antonios Proestakis (2013), Applying Behavioural Sciences to EU Policy-Making, Luxembourg: Publications Office. Volks, Holger (1978), Einbau- und Anlegeverhalten Sicherheitsgurte: Einfluß von Rechtsnormen und Aufklärungskampagnen über den Sicherheitsgurt auf Informationsstand, Einstellungen und Verhaltensweisen der Pkw-Fahrerpopulation, Köln: Bundeanstalt für Straßenwesen.

M4716-STRAßHEIM_9781785367847_t.indd 36

25/04/2019 13:32

Nudging before the nudge?  37 Weber, Max (1976), Wirtschaft und Gesellschaft: Grundriss der verstehenden Soziologie, Tübingen: Mohr, 5th edition. Weinstein, Neil D. (1989), ‘Optimistic Biases About Personal Risk’, Science 246, 1232–1233. Weishaupt, Heike (1999), Die Entwicklung der passiven Sicherheit im Automobilbau von den Anfängen bis 1980 unter besonderer Berücksichtigung der Daimler-Benz AG, Bielefeld: Delius Klasing. Wetmore, Jameson F. (2004), ‘Redefining Risks and Redistributing Responsibilities: Building Networks to Increase Automobile Safety’, Science, Technology, & Human Values 29 (3), 377–405. Whitehead, Mark et al. (2014), ‘Nudging All over the World: Assessing the Global Impact of the Behavioral Sciences on Public Policy’, https://changingbehaviours.files.wordpress.com/2014/09/nudgedesignfinal.pdf (last accessed 30 September 2018). Wilson, David and William Dixon (2012), A History of Homo Economicus: The Nature of the Moral in Economic Theory, London, New York: Routledge. World Bank (2015), Mind, Society, and Behavior: World Bank Development Report, Washington, DC: World Bank. World Health Day (1961), ‘World Health Day 1961: Accidents Need Not Happen’, The Courier 14 (April). World Health Organization (2013), ‘Global Status Report on Road Safety’ http://www.who.int/violence_injury_ prevention/road_safety_status/2013/en/index.html (last accessed 30 September 2018).

M4716-STRAßHEIM_9781785367847_t.indd 37

25/04/2019 13:32

3.  Economic expertise and public policy Frédéric Lebaron

1. INTRODUCTION Economic science has penetrated the main places of power and feeds daily into public and private actions. It is involved in the construction of states and markets. Through the development of economic expertise in the steering bodies of the global economy, its status of ‘cameral science’ (for a history of this notion, see e.g. Laborier et al. 2011), both modern and universal, has continued to strengthen globally, so that the distinction between ‘power’ and ‘knowledge’ is barely operative about it. Economics has become a scientific subsystem strongly structured around a set of institutions of evaluation and consecration, with the ‘Nobel memorial prize in economic sciences’, so-called ‘Nobel Prize’ (Lebaron 2000, 2006; Offer and Söderberg 2016) at its top. This award illustrates the structuring force of an organizational construction operating globally through a particularly advanced and efficient unification process: linguistic unification with the absolute hegemony of English; theoretical unification (though potentially competing paradigms like behavioural economics are getting stronger) with the ‘neo-classical’ framing; methodological unification with the use of a standardized design of mathematical statistics; evaluative unification with the use of bibliometrics and quantitative assessments. It is in this space that controversies and theoretical innovations designed around individual labels and universe repositories can be deployed (Coats 1993, 1997). In the first section of this chapter we will analyse the educational and professional structuration of economic expertise and how it affects its relation to the evolution of public policy at various levels; we will then analyse the rise of the discipline in power institutions as a major component of this evolution, with the particular example of central banks; then, we rapidly discuss the changes which are occurring since 2008, showing a growing tension between internal trends and external dynamics, in a context of globalization of expertise.

2.  ECONOMICS AS A PROFESSIONAL EXPERTISE The training of ‘economists’, either ‘professionals’ (exercising the profession of economist, in all its diversity, within universities, public administration, professional organizations or businesses) or simply ‘economics graduates’ (working in various industries, primarily banking, finance and administration), is a major process through which economic expertise has been more and more structured over time (Klamer and Colander 1990). Through education, the economic reality becomes more easily a social reality that is taken for granted, almost natural, whose ‘laws’ are binding on all social actors, especially politicians. It serves as an implicit reference inside the political, administrative and economic institutions. 38

M4716-STRAßHEIM_9781785367847_t.indd 38

25/04/2019 13:32

Economic expertise and public policy  39 The field of economics is now structured by strong institutional settings and norms which organize its daily functioning. They relate to the content of curricula, the methodological norms of scientific validation, the language, and the scientific culture it defines in general. 2.1  Economics in the Educational System The economic culture may be analysed as a ‘third culture’, in a meaning close to what Wolf Lepenies (1988) associated to human and social sciences, next to the literate culture (literature, law, etc.) and to the scientific-technical culture. It also symbolizes, in many passages of The Protestant Ethic and the Spirit of Capitalism and other works by Weber, a ‘utilitarian’ culture oriented towards the business world, which will make it possible to expand the capitalist ethos and which is restricted in the traditional world, in favour, for example, of the traditional literary culture. In any case, one should not forget the pivotal role of the education system (specifically higher education) in the transformation and reconfiguration of public policy, and in particular the rise of economics and management as elite culture now inherent in the exercise of power. This idea is supported by the trends identified by Pierre Bourdieu since the 1970s (Bourdieu 2000), confirmed regularly since then, and illustrated by developments in different countries: the rise of the technocratic and business schools in the field of schools, transformation of elite education in the sense of an increased weight of economic knowledge at the expense of traditional legal or literary culture, and so on. The rise of economics is measured, more quantitatively, by the growing weight of the economic sectors in secondary and higher education in the long term. A limitation to this process could only come from deep disciplinary reconfigurations that would change the ‘alliance systems’ within higher education and research institutions. The ‘Economics and Management’ group of disciplines appears now clearly in a dominant academic position in many countries. It means that economics is a central part of the culture of public policy actors as well as some business leaders all over the world. 2.2  A Case of Strong Institutionalization and Professionalization The historical basis of this process has been described in various studies on the changes in the content of economics, especially around the imposition of microeconomics–macroeconomics–econometrics at the chore of the curriculum. The standardization of teaching contents and curricula is clearly part of the global enhancement of the discipline in the long run (Coats 1993, 1997; Loureiro 1995; Yonay 1998; Chmatko 2002; FourcadeGourinchas 2006; Herredia 2014): this standardization potentially provided the discipline with an (apparently, at least) universal set of intellectual resources and techniques. An aspect of this process is the algorithmic way of ‘proving’ in economics, classically based on the ‘fit-and-test’ practice in mathematical statistics. A simple chaining of rational behaviour, aggregated relations and econometric tests is the universal way of providing answers to scientific questions. This way of proving is now challenged by experimental procedures inspired by the natural sciences. A second aspect relates to the way the discipline has been organizing itself around

M4716-STRAßHEIM_9781785367847_t.indd 39

25/04/2019 13:32

40  Handbook of behavioural change and public policy a single set of evaluation instruments, symbolized by the ‘pyramid of prizes’ (from the ‘Nobel prize’, created by a central bank in 1968, to the national and sub-disciplinary ones), which are all connected through a similar space of measurement (‘space of equivalence’), based on the systematic use of bibliometrics. Bibliometrics provided economics an efficient instrument of strong standardization and unification. This has allowed the ordinary use of rankings as a way of evaluating hierarchies between individuals, departments, universities, countries and so on. This trend is of course in total coherence with the growing use of rankings in the context of new public management techniques in the public sector, and in particular in universities and research institutions, showing a particular affinity between the discipline and new trends in public organizations. This instrument relates to another vector of unification, namely the use of English as the universal language of economics, imposed through the domination of Englishspeaking journals and publications. In the process of imposition of English as the unique scientific language, economics has clearly been at the forefront. One of the consequences of the process is the relative decline in intellectual diversity which has been characterizing some of the national academic cultures in economics and political economy, though it is counter-balanced by a growing internal differentiation, exemplified by the rise of experimental economics, which challenges the established mainstream. 2.3  A Large Labour Market Economics as a discipline has gained influence, especially after 1945, by spreading into a large variety of professional sectors outside the university. In the post-war period, it relates to the development of macroeconomic forecasting in ministries of finance, central banks, planning and statistical institutes, which has been the basis of the historical influence of the discipline. Since the 1970s, the discipline has been more and more present as a training programme in the field of finance and the banking sector. But the presence of economics has also expanded in areas such as health, education, social policy and development policy. In these sectors, the role of academic knowledge has certainly been rising in the long-term, and this evolution clearly relates to a larger use of microeconomic reasoning, cost–benefit computations, quantitative assessments, in a double movement of scientization (in a broad sense of growing reference to scientific cognitive tools and culture) and economization. The globalization process has been accompanied by a certain blurring of the traditional divisions between different levels of action, from the local to the global. Economics provides a set of tools which may be relevant at the level of local as well as international organizations. As with the academic sector, certain segments of the labour market (like international organizations and central banks, Coats 1997) are already highly internationalized. 2.4  Towards the Model of Experimental Science? A recent evolution is characterized by the rise of experimental practices in the sector of public economics, especially as a tool for public policy evaluation in the context of

M4716-STRAßHEIM_9781785367847_t.indd 40

25/04/2019 13:32

Economic expertise and public policy  41 ‘evidence-based’ policy. This evolution is hardly a new feature, since it was experienced at least since the 1960s in the United States, but it has been more rapid and more visible since the 1990s, especially in development economics with the role of institutions such the Abdul Latif Jameel Poverty Action Lab in the US and globally (Jatteau 2016). A recent book written by two French economists (Cahuc and Zylberberg 2016) claims that economics has become a full experimental science, following the model of biology and medicine. The authors argue that the scientific study of public policy measures is now similar to evidence-based medical decisions. Evidence-based decisions are now strongly influencing public policies in various sectors. When looking at the polemical reception of the book, it is at least obvious that this claim is not universally followed. But the existence of such a claim testifies the growing influence of models of expertise derived from the natural sciences, which is coherent with the long trend of scientization, as we defined it broadly, of public policy in the economic sector. This change relates to a process of social ascension which we will now more specifically describe.

3.  A LONG TREND OF ASCENSION The social ascension of economics appears relatively obvious if we adopt both a transnational and long-duration perspective, even if it is not that easy to measure through simple indicators (Fourcade-Gourinchas 2006). It is a multidimensional process, in which various realities can be involved, as the multiple meanings of the word ‘economist’ itself show (Lebaron 2000). We can for example mention the accession of professional economists to power positions, especially inside international organizations, governments, central banks and more generally in public and private organizations; the professional structuration of the group all over the world, especially through a process of standardization and unification, showing an extension of jurisdiction, especially in comparison with Law (Fourcade-Gourinchas 2006); the growing visibility of economists in the media, in public debate, during electoral campaigns and through the publicity of ‘consecration institutions’ like the ‘Nobel prize’ (Lebaron 2006); the rise of economic discourse and culture, cost–benefit reasoning, in the daily life of organizations and social life, and so on. This set of trends may give the impression of an ineluctable process of evolution, led by strong social dynamics, which could easily be described as a mechanistic process, whereas it is at least partially historically contingent. We can descriptively identify two phases in the rise of economics as bureaucratic and political resources, which appears particularly valuable: the first (starting in 1930) is dominated by macroeconomics, economic planning, and is oriented towards public intervention. The second phase (from the late 1970s) is characterized by the widespread success of microeconomics and neoliberalism in a large variety of forms. If this process is so powerful in this second phase, it is certainly largely due to the fact that it also corresponds to a transformation of state institutions, which in parallel are moving away from the Weberian model, and are approaching a universal type of economic organization based on the model of private enterprise (for France, see Bezes 2009). The move towards ‘new public management’ is one aspect of this overall trend, even it has been followed by critiques, and certain moves towards more networked, information-based modes of organization. This phenomenon takes different forms in different periods and countries, but in all

M4716-STRAßHEIM_9781785367847_t.indd 41

25/04/2019 13:32

42  Handbook of behavioural change and public policy cases it tends to delegitimize the elected actors and to favour other forms of political authority, more focused on independent expertise and provided by rational economics, including behavioural economics which aims at integrating irrationality and cognitive biases as parts of the analysis. Another vector for the construction of a social order now more clearly focused on the market is the formation of a political and administrative expert group, composed of leading civil servants and politicians (see for example Georgakakis and Rowell 2013), more strongly influenced by economics and, through it, by its socialization to ‘market mechanisms’, which requires familiarity with certain categories of perception and legitimate action. 3.1  Economics, Law and the Changing Post-war State This process first refers to the rise of expertise at the expense of other forms of legitimacy, a long-term phenomenon, related to the historical transformation of the state, which upsets all the traditional structures of domination. The legal culture is, however, far from being unified and it tends to be closer to ‘modern’ economic science, especially under the influence of North American academic currents, such as the movement ‘law and economics’ and the progress of business law. Sociological research on the transformation of the political field and the administrative field thus shows a tendency to partial substitution of legal and administrative knowledge, threatened with obsolescence or simply strongly outperformed, by more clearly economic knowledge (for an example taken from the European Commission, see Georgakakis and de Lassalle 2007). Economic knowledge is composed of management and economics properly speaking (macroeconomics, national accounting, econometrics, microeconomics), even if the boundaries between these two subsets remain relatively fuzzy. With planning, national accounting, macroeconomics as typical or at least the most legitimate knowledge, political or administrative actors (the link between the two being very close) had in common not to foreground the universality of market processes, but rather to ‘over-value’ the ability to understand public policy generally and especially to transform – more or less strongly – the economic and social order. This is particularly the case in France, at least around 1945‒1947 and then in the 1960s, but also in many developing countries, as highlighted work that unambiguously converges around this ‘story’ of the transformation of public policy after the war. The Marxist, socialist, Keynesian inspirations strongly competed, but they were all fastened on a strong belief in the virtues and transformative possibilities of public action, a belief probably related to the need to ‘rebuild’ a world destroyed by war. The economic field was often conceived as a more or less complex system, but largely ‘malleable’ by political action, through laws with which the policy could be centralized or more localized in order to achieve the objectives centrally defined. This ability to transform public policy ‘outside’ the market order could be accompanied by the prospect of a systemic change, embodied in ‘socialism’ in its many forms (with the Marxist–Leninist voluntarism providing the working class the ability to radically transform the economic and social order to bring about a classless society based on a centrally planned economy). Across different countries and traditions, the degree of effectiveness associated with public policy remained extremely variable, as analyses in the field of development economics and economic history show (see for example Chang 2003): countries which

M4716-STRAßHEIM_9781785367847_t.indd 42

25/04/2019 13:32

Economic expertise and public policy  43 promoted a growth led by export in a very voluntarist way were particularly successful, compared to countries remaining politically weak and economically dependent. But the political and administrative actors trained in the post-war period often claimed a higher rationality than market players, and based this claim on various economic doctrines from soft Keynesianism to Marxism, through a variety of developmentalist structural conceptions. If companies were to search ‘competitiveness’, this ‘rationalizing’ approach resulted primarily in representations of a general systemic movement, in which the state leadership remained essential to a large majority if not the totality of countries during this ‘Keynesian’ period. 3.2  The Neoliberal Revolution Since the late 1970s, public action in the economic sphere has lost its obviousness in favour of a belief in the intrinsic virtues of the market, competition, and private management, under the influence of various intellectual groups: the notion of ‘planning’, but also to some extent that of ‘macroeconomic policy’ were partly countered or disqualified by the more widespread idea of a relative impotence of politics, of the state, especially because of the ‘globalization’ of markets. This process is seen as a quasi-natural force, against which adaptation – the integration of competitive logic – now appears to be the only ‘solution’ (for France, see Denord 2007). In some economics work, such as Finn Kydland and Edward Prescott (1997) on monetary policies (work that has been consecrated by the ‘Bank of Sweden Prize in Economic Sciences in Memory of Alfred Nobel’), the impotence of the State and political actors crystallizes in the form of an opposition between ‘discretionary policies’ intended to threaten the stability and efficiency of the economic order by their irrational submission to special interests, and the adoption of ‘rules’ of public action based on a superior rationality which is ultimately that of economics: steady growth in the money supply, insensitivity to inflationary pressures, and so on. Already present in the process of delegitimization of the elected actors, but this time at the advantage of ‘independent’ agencies rather than a strong ‘executive power’, bureaucratic actors were also disqualified because they were considered a source of arbitrariness and potential distortion of price mechanisms. The ‘public choice’ school, as an influential American intellectual movement, strongly fuelled this disqualification of bureaucratic actors, defined as maximizers of public spending and power, far away from the idealized representation of the search for a general interest superior to individual selfishness. The economic education of political and administrative actors, in the same historical movement, clearly moved to a frame of references more directly derived from the commercial sector. In this new framework, policy seeks only to create a legal and institutional environment conducive to the expansion of markets and the creation of competitive enterprises, limiting the introduction of political or social objectives such as reducing economic and social inequalities, social and territorial cohesion, and so on. The ‘Washington Consensus’ was the term used to describe this process in international institutions (Dezalay and Garth 1998), naming the rise in power of incontestable ‘market solutions’ ranging from ‘orthodox’ fiscal and monetary policies to the mass privatization of utilities and rapid integration into the global capital market. The effects of these policies on economic and financial instability, and social inequalities, have contributed

M4716-STRAßHEIM_9781785367847_t.indd 43

25/04/2019 13:32

44  Handbook of behavioural change and public policy to a gradual loss of credibility of these ‘solutions’, especially since the beginning of the ‘subprime’ crisis, but they continue to be the dominant epistemic framework of public policies at the global and national levels. The interpretation of this ‘consensus’ as a product of a relatively homogeneous ‘epistemic community’, structured public policy networks and dominates the neoinstitutionalist research on the subject. Critical studies (like those of Joseph Stiglitz) showed how this could lead to unintended social effects in economic policy and, ultimately, bring relative discredit upon both the ‘Washington Consensus’ and the Bretton Woods consensus in many countries. 3.3  The New Role of Experts and the States In state organizations, the traditional concept of public administration suffers from the consequences of the imposition of new scientific categories from economics and management: the ‘new public management’ brings the workings of state organizational aims close to those of companies in the commercial sector, is supposedly more efficient by introducing external evaluations of individualization processes, and incentivizes competition, as a practical application of the ‘principal–agent’ model. We saw the translation in the health sector, more strongly dominated by economic reasoning (Benamouzig 2005; Pierru 2007); the educational sector is subject to the same type of logic. More broadly, the social sectors characterized by forms of disinterestedness are affected by the application of criteria of ‘performance’ from the corporate world. Rather than market introduction, this is often called a ‘quasi-market‘ in the Anglo-Saxon world, the idea being to create many mechanisms similar to market mechanisms. Scientific research as an institution has become a particular frame of reference in the general economization of processes. Economics provides legitimacy to public institutions which are now primarily backed by market logic and which, conversely, are the inseparable institutional support for the historical canons of ‘neoliberalism’. The example of ‘independent’ central banks illustrates this new feature of public policy, strongly embedded in the market order (Lebaron 2006). Central banks are active by providing liquidities when the financial markets show serious signs of weakness, as in 2007 and 2008 during the crisis. Their actors are socially defined as ‘apolitical’ and their legitimacy ultimately rests on their rational expert status. Central bankers resist any form of politicization of public action, that is to say, looking for targets that differ from those very specific to central banks and defending their position when political actors jeopardize their legitimacy. We observe a rise of economic indicators used to define the objectives of the organizations: The increased domination of certain economic and financial indicators at all levels illustrates the growing dominance of criteria coming from the economic field. Financial criteria penetrate all commercial organizations, including industrial or commercial, the latter being dominated by profitability criteria from the financial sector. These same criteria exert a strong pressure on all social sectors, including those which have hitherto resisted the application of criteria and market standards, such as the cultural universe, the administration of education, health, social policy, and so on, through public policy and monitoring the effectiveness of public spending. Finally, we can mention the recent development of a market economic expertise (‘think tanks’), again based on the American model, which is observed in most countries of the world, especially in France. This movement also illustrates the peculiarity of this market: private organizations,

M4716-STRAßHEIM_9781785367847_t.indd 44

25/04/2019 13:32

Economic expertise and public policy  45 often supported by foundations, professional organizations or large corporations and think tanks are mainly oriented towards the reform of public action through evaluation methods (‘benchmarking’) and by making media visibility an instrument of action. Their contribution to the existence of a ‘pro-market’ economic discourse in the public sphere has become fundamental. It complements the work in the directions of economic studies of banks and financial institutions which supply the daily cyclical analysis, particularly for the business press. 3.4 Financialization The rise of economists is linked in multiple ways to the ‘financialization’ of the economy. In this context, the success of economic science that is visible in the curriculum of many elite schools from the 1970s and especially the 1980s, appears in fact to be closely related to the connections between the ‘science of finance’ and economics. Economics not only produces an interpretation of reality, but also very directly informs the practices and institutions. In any case, modern economics provides the most current legitimation of the major developments in finance as a social institution: globalization, flexible exchange rates, derivatives markets, and so on. These developments have resulted in the award of several Nobel prizes. Conversely, ‘critical’ economists find consensus in the denunciation of both the effects and failures of the financialization of Western economies, such as inequality, instability and other related trends. Works on the diversity of capitalism show the historical dominance of Anglo-Saxon capitalism, precisely characterized by the hegemony of finance, and this issue has become even more visible during the financial crisis of 2007 and 2008. 3.5  A Rise of Behavioural Economics in Public Policy These changes are not unidirectional and may be combined with new developments of expertise in public administrations, especially at the international level. With the rise and diversification of economics, such as the growing reference to experimental methods, one has also seen a complexification of the epistemological and methodological bases of the discipline, and a corresponding change of its relations to public policy. A central aspect of this evolution is the emergence and success of behavioural economics, especially since the founding works of Allais in France and Kahneman and Tversky in the United States. The ‘Nobel prize’ objectifies this ascension, with awards successively attributed to Daniel Kahneman (and to a lesser extent Vernon L. Smith, as experimental and neuro-economist) in 2002, and recently with Richard Thaler in 2017. In producing a new framework, behavioural economics helps to rethink the logics of public policy and the scientific basis of the economic discipline, in a more nuanced and modest way, and on an empirically sounder basis. They are nuanced in the sense that rational action is not considered independently from concrete observed behaviour, and modest in the sense that decisions should be more and more based on in-depth and systematically verified behavioural analyses. A remaining problem lies nevertheless in the interpretation of experimental and behavioural results. In France, with the recent polemical debate provoked by Cahuc and

M4716-STRAßHEIM_9781785367847_t.indd 45

25/04/2019 13:32

46  Handbook of behavioural change and public policy Zylberberg (2016), already mentioned above, a less modest and nuanced vision of experimental economics (defined as a mixture of laboratory, random and natural experiments) has been promoted and simultaneously contested from various sides. In their perspective, economics is seen as fundamentally similar to evidence-based medicine, and should share the status of this dominant sector of applied biology. In a clear top-down perspective, a high degree of both certainty and consensus is supposed to be obtained through experiments, while the media and political actors are expected to be informed about and aware of the resulting evidence. This vision is more the idealization of an expected future than an empirically-based description of the current state of economics: the real discipline is still composed of a lot of purely theoretical speculations clearly situated inside the neo-classical paradigm, with a very strong reference to unproven sets of theories, very classical or ‘new’ econometrics on observational data, which share the situation of all social-historical sciences. What is nevertheless clear is the fact that a priori dogmatic conceptions of economic knowledge tend to be in competition with more empirically based results and political advices connected to a diversity of expertise. 3.6  The Example of Central Bankers In a prosopographical survey – a collective biography – which has given birth to regular publications since the beginning of the 2000s (Lebaron 2000), we have established the process by which the social identity of central bankers is constructed and reproduced at the intersection of three major social worlds: the Academy, the State and finance. A large proportion of leading central bankers are also well-recognized economists, who have been working inside the academic field and have specialized on monetary issues. Other trajectories relate to the state (politics and/or bureaucracy) and others to financial markets. Remaining types of trajectories (like pure industrial careers) are rare and very specific. This triple source of legitimacy (which derives from science, law and the market) is illustrated by the mixture of professional profiles inside the leading committees of central banks (monetary policy committee, like the Federal Open Market Committee of the Federal Reserve Bank or the Governing Council of the European Central Bank). The social dynamics since the 1980s have led to a progression of academic legitimacy and have been accompanied by closer links to finance, as illustrated now by the regular debates about the ‘Goldman Sachs connection’ at the top of central banks. At the same time, some central banks, like the People’s Bank of China, the Bank of Japan or even the ECB remain highly embedded in particular political and bureaucratic fields, showing the specificity of national capitalisms inside the organization of the central bank itself. Central bankers very clearly illustrate a more general process of combination of types of authority, which is related to the global status of economics. Leaders of the international ‘Bretton Woods’ organizations (the International Financial Institutions), like the International Monetary Fund and the World Bank, also have a high level of economics symbolic and technical capital, and they are characterized by deep connections with national political and bureaucratic fields (as exemplified by the cases of Dominique Strauss-Kahn, professor of economics, high civil servant and politician, and Christine Lagarde, law professional and politician, former minister of finance). We might also find similar trends among financial ministers, who have a central position in terms of the concrete implementation of neoliberal policies. The example of the

M4716-STRAßHEIM_9781785367847_t.indd 46

25/04/2019 13:32

Economic expertise and public policy  47 ‘Eurogroup’ (that is the group of ministers of finance inside the Eurozone) shows the complex tension between academic legitimacy (mainly in economics, but competed by Law), political and bureaucratic embeddedness and, of course, the more hidden links to financial markets. It was particularly obvious in 2015, with the arrival of the Greek Yanis Varoufakis, who imported academic values of heterodox economics inside a bureaucratic and political field and faced harsh reactions from the insiders, who share a more legal and political background.

4. CONCLUSION: THE EFFECTS OF THE CRISIS AND NEW TRENDS IN PUBLIC ECONOMICS The 2007 crisis, often described as a failure of economic thought, hasn’t stopped this general evolution, though it has created a new situation where the strength of the discipline seems to be put into question. At the most ‘external’ level, the discipline has been questioned in the media and by social movements, but economic experts have been put at the centre of the crisis, especially in Europe after 2010. Inside the Eurozone, the political legitimacy of economic experts has even in a sense been reinforced by the general context of shift to orthodox (austerity) budgetary policies. The process may appear paradoxical, because many economists have expressed a strong dissent as regards this shift and its macroeconomic consequences, in Europe and elsewhere, but the social mechanisms which have allowed the European Central Bank, the Treasuries and ‘economic experts’ to maintain their orientations are indeed very strongly rooted in the institutional and political dynamics of the EU. Insofar as the crisis of economics is a crisis of belief, we can certainly compare it to the crises that regularly affect religious systems. The dynamics of the crisis illustrate the progressive disarticulation of this universe of belief. It is a moral crisis, which questions the links between business and the world of finance on one side and ethics on the other. It is also a crisis of the concepts and intellectual instruments provided by economics, with, for example, the debates on Gross Domestic Product questioning the fundamental belief in the value of wealth, accumulation of money, and so on. It is also more concretely a crisis of authority, which is fuelled by erroneous predictions, excessive optimism, professional arrogance and social negative consequences of ‘rational’ economic decisions. In this process, the emergence of fields such as behavioural economics can be seen as a scientific response to a crisis of dominating core beliefs – a crisis which possibly has more or less deep and fundamental consequences for both scientific practices and public policies. A second aspect here is the fact that the crisis has not really put into question the internal dynamics of the field: the standardization and unification processes have a strong inertia, which is based on the reproduction mechanisms of any social field. ‘Heterodox’ economists may at the same time be more visible in the public sphere (in France, for example through the ‘Association d’économie politique’ or the ‘Economistes atterrés’ movement, at a global level through the student’s critical movement) and may continue to lose positions and influence inside the academic institutions. This ‘external/internal’ tension creates a disarticulation which contributes to the dynamics of the crisis, making the most recent evolutions strongly ambivalent.

M4716-STRAßHEIM_9781785367847_t.indd 47

25/04/2019 13:32

48  Handbook of behavioural change and public policy

REFERENCES Benamouzig D. 2005. La santé au miroir de l’économie, Paris: PUF. Bezes P. 2009. Réinventer l’Etat. Les réformes de l’administration française 1962–2008, Paris: PUF. Bourdieu P. 2000. Les structures sociales de l’économie, Paris: Seuil. Cahuc P. and Zylberberg A. 2016. Le négationnisme économique ou comment s’en débarrasser?, Paris: Flammarion. Chang H.-J. 2003. Kicking Away the Ladder. Development Strategy in Historical Perspective, London: Anthem Press. Chmatko N. 2002. Les économistes russes entre orthodoxie marxiste et radicalisme libéral, Genèses, 47, 123‒139. Coats A.W. 1993. Economics as a Profession, in Coats A.W. The Institutionalization and Professionalization of Economics, Boulder/London: Routledge. Coats A.W. 1997. The Post-1945 Internationalization of Economics: Annual Supplement to Volume 28 of History of Political Economy, Durham, NC: Duke University Press. Denord F. 2007. Néo-libéralisme version française : histoire d’une idéologie politique, Paris: Demopolis. Dezalay Y. and Garth B. 1998. Le ‘Washington Consensus’. Contribution à une sociologie de l’hégémonie du néolibéralisme. Actes de la recherche en sciences sociales, 121 (2), 56‒75. Fourcade-Gourinchas M. 2006. The Construction of a Global Profession: The Transnationalization of Economics, American Journal of Sociology, 112 (1), 145‒195. Georgakakis D. and de Lassalle, M. 2007. Les très hauts fonctionnaires de la Commission européenne: genèse et structure d’un capital institutionnel européen, Actes de la Recherche en Sciences Sociales, 166‒167, mars, 39‒53. Georgakakis D. and Rowell J. (eds) 2013. The Field of Eurocracy. Mapping the EU Staff and Professionals, with Jay Rowell, London: Palgrave, coll. European Administrative Governance. Herredia M. 2014. A quoi sert un économiste ?, Paris: La Découverte. Jatteau A. 2016. Faire preuve par le chiffre? Le cas des expérimentations aléatoires en économie, these de doctorat université Paris-Saclay. Klamer A. and Colander D. 1990. The Making of an Economist, Boulder, CO: Westview Press. Kydland, F.E. and Prescott E.C. 1977. June. Rules Rather Than Discretion: The Inconsistency of Optimal Plans, Journal of Political Economy, 85, 473–992. Laborier, P., Napoli, P., Vogel, J. and Audren, F. 2011. Les sciences camérales: activités pratiques et dispositifs publics, Paris, PUF. Lebaron F. 2000. La croyance économique. Les économistes entre science et politique, Paris: Seuil. Lebaron F. 2006. Nobel Economists as Public Intellectuals: The Circulation of Symbolic Capital, International Journal of Contemporary Sociology, 43 (1, April), 87‒101. Lepenies W. 1988. Between Literature and Science: The Rise of Sociology, translated by R.J. Hollingdale, New York: Cambridge University Press. Loureiro M.R. 1995. L’ascension des économistes au Brésil, Actes de la recherche en sciences sociales, 108, 70‒78. Offer A. and Söderberg G. 2016. The Nobel Factor The Prize in Economics, Social Democracy, and the Market Turn, Princeton: Princeton University Press. Pierru F. 2007. Hippocrate malade de ses réformes, Bellecombe-en-Bauges: Croquant. Weber M. 2001. The Protestant Ethic and the Spirit of Capitalism, London: Taylor & Francis. Whitley R. 1984. The Intellectual and Social Organization of the Sciences, London and New York: Oxford University Press. Yonay Y. 1998. Struggles over the Soul of Economists: Institutionalist and Neoclassical Economists in America between the Wars, Princeton: Princeton University Press.

M4716-STRAßHEIM_9781785367847_t.indd 48

25/04/2019 13:32

4.  A bibliometric analysis of behavioural studies in economics and public policy journals Stuti Rawat

1. INTRODUCTION Scholars in the field of economics and public policy have long recognized the importance of understanding behavioural processes. Perhaps the earliest behavioural economist can be thought to be Adam Smith. Smith’s Theory of Moral Sentiments (1759) is full of insights that relate to aspects of individual preference and judgment that presage developments in contemporary behavioural economics (Ashraf et al. 2005). Two centuries later, Robert Dahl (1947) and Herbert Simon (1955) would write about the distance between knowledge of human choice processes and administrative theory, stressing the need for public administration to derive from, and contribute towards, an understanding of human behaviour. In the years since this early scholarship, academic writing on the topic has burgeoned, as has the use of behavioural change instruments and behaviourally informed policy (Lunn 2014; Van Bavel et al. 2013). Academic writing, in addition to the knowledge that it provides on its chosen topic of focus, serves a two-fold purpose. While providing information about the distance traversed by a discipline, it also serves as a signpost for the direction in which the discipline is likely to evolve. Based on the assumption that the increasing prominence of behavioural dimensions in economics and public policy is reflective of, and reflected by, academic writing on the topic, this chapter undertakes a bibliometric analysis of the leading economics and public policy journals between the years 1940 to 2016. The intent behind focusing on articles in high-ranked journals is to map the particular body of thought which plays a prominent role in driving forward the evolution of the discipline. Through this analysis three key questions are addressed. The first – is the rise of behavioural studies in economics and public policy a recent phenomenon or something that follows cycles of its own? Second – if a cyclical pattern can be identified, is this mirrored between the two disciplines? There is reason to believe in this possibility, given that behavioural economics is premised on violations of rational choice theory (Thaler and Mullainathan 2000) which in itself can be traced to Simon’s conception of ‘bounded human rationality’ from 60 years ago (Simon 1955, 1957).1 The final ­question – is there a sectorial or geographic dimension to these behavioural studies? The structure of the chapter is as follows: the next section discusses the increasing importance of behavioural instruments and policy, providing reasons for this. The aim of this section is to provide a description of real-world developments so as to juxtapose them against related academic writing in economics and public policy. The third section elaborates on the methodology undertaken for the bibliometric analysis. Section four discusses the findings and the major themes of analysis, answering the key research 49

M4716-STRAßHEIM_9781785367847_t.indd 49

25/04/2019 13:32

50  Handbook of behavioural change and public policy ­ uestions. The fifth section focuses on the implications of the findings and how they tie q in with real-world issues. This is followed by some concluding remarks.

2. PRACTITIONER PERSPECTIVE: RISE OF BEHAVIOURAL STUDIES IN ECONOMICS AND PUBLIC POLICY While there is no single accepted definition for behavioural policy, scholarly definitions for behavioural economics can be categorized into three broad groups. The first like Herbert Simon’s focuses on the violations of neoclassical assumptions of human rationality. According to Simon (1976), actual behaviour often falls short of objective rationality2 because of a lack of complete knowledge of consequences, possible alternatives and difficulties associated with anticipation. He thus defines behavioural economics as being concerned with the empirical validity of neoclassical assumptions about human behaviour and where they prove invalid, deriving empirically backed laws that would describe behaviour as correctly and adequately as possible. The second group of definitions focuses on the explicit incorporation of psychological phenomena. Thus, Thaler and Mullainathan (2000) define behavioural economics as the combination of psychology and economics that investigates what happens in markets when human agents exhibit limitations and complications. The third category of definitions focuses on the methods. Shiller (2005) defines behavioural economics as the application of methods from other social sciences, particularly psychology, to economics. Unlike orthodox economics that adopts a deductive approach, behavioural economics follows an inductive approach that is based on systematic and repeated experiment and observation. Observations are followed by models or theory and ‘method’ becomes crucial (Lunn 2013). The focus on methods has resulted in the behavioural umbrella being regarded as one which encompasses behavioural insights as well as experimental methods.3 The experimental method refers to the use of a control and treatment group with subjects randomly distributed between the two. But the idea of having a control and a treatment group is not particularly new, and can be traced as far back as 1747, when James Lind used it to demonstrate the benefits of citrus fruits (Thomas 1997). Between the 1960s and 1990s, randomized trials were used in government sponsored social experiments in the United States and Europe. Since then, they have been used in a variety of settings around the world (Levitt and List 2009), with behavioural experiments becoming increasingly common in the policy world and especially development policy (Datta and Mullainathan 2014). Countries such as United States, the United Kingdom (Lunn 2014), Australia (see Cooper 2010), Singapore (Low 2012) and in the European Union (Van Bavel et al. 2013) have concentrated efforts on the application and implication of behavioural insights for larger policy-making. Though the number of randomized controlled trials (RCTs) in public policy are still far fewer than those in the health sector (Shepherd 2007) yet even this limited use has been a major stimulus for empirical development economics (Basu 2014) and in policy designed towards tackling poverty (see Banerjee and Duflo 2012; Karlan and Appel 2011). A number of organizations are making efforts to make evidence stemming from these RCTs centrally available. Thus, evaluation databases of organizations like J-PAL (Abdul Latif Jameel Poverty Action Lab), World Banks’ Development Impact Evaluation (DIME),

M4716-STRAßHEIM_9781785367847_t.indd 50

25/04/2019 13:32

A bibliometric analysis of behavioural studies in economics and public policy journals  51 Coalition for Evidence-Based Policy, among many others (Glennerster and Takavarasha 2013) are contributing to the wider dissemination and spread of behavioural studies. 2.1  Reasons for the Rise This rising influence of what could be termed ‘applied behavioural science’ (Kahneman 2012) can be attributed to its successful application in achieving higher target compliance with relatively lower costs (Thaler and Sunstein 2008). This is done by affecting programme design in three ways – by changing how the scope of the problem is defined, how it is diagnosed and how solutions to it are designed. Further, embedding innovation into the design process leads to a greater chance of success (Datta and Mullainathan 2014). Another reason explaining the rising popularity of behavioural instruments is that it provides an alternative way of using traditional policy instruments. Policy instruments are generally separated into three categories: inducements (incentives and sanctions), regulations, and knowledge or capacity tools (Schneider and Ingram 1997; Stone 2002; see also Loer and Howlett, Chapters 13 and 6 respectively, in this volume). Theories of rational choice and utility maximization are the underlying assumptions on which these tools are traditionally premised (Stewart 1993; Stone 2002). Behavioural economics and behavioural policy make use of these same tools but in ways that seek to exploit the cognitive biases individuals consistently exhibit, instead of targeting the supposed utility maximization tendency of individuals. To take an example, Ferraro and Price (2013) find that informing individuals about their neighbours’ water consumption results in conserving water much more than providing them simple information or exhorting them to be thrifty water users. The same policy instrument, ‘knowledge’, is used to harness the tendency that behavioural economists have observed – that of conforming to the perceived ‘social norm’, resulting in greater success than the traditional use of this instrument. Lunn (2013) identifies three routes by which behavioural economics has influenced policy: (1) the theory of libertarian paternalism (or ‘nudges’); (2) the provision of toolkits for policy-makers seeking behavioural change; and (3) the expansion of the skill-set of applied economists (and scientists in related fields). Lunn’s first point is based on Nudge, a book by Thaler and Sunstein (2008) that focuses on the application of behavioural economics to policy-making. A nudge is defined as ‘any aspect of choice architecture that alters people’s behaviour in a predictable way, without forbidding any options or significantly altering their economic incentives’ (p. 6). Thus, using individuals’ cognitive biases, policy can be designed to influence people’s choices in a certain direction. They categorize this approach as ‘libertarian paternalism’ (Sunstein and Thaler 2003; Thaler and Sunstein 2008). The emergence of many government initiatives has been influenced by this, such as auto enrolment in pensions in the US (Sunstein 2011). With respect to his second point Lunn argues that while behavioural economics appears to provide a valuable toolbox on a context specific, case-by-case basis (Oliver 2013), there are many unresolved issues within the discipline itself. For instance, even among behavioural economists, there isn’t a consensus on which behavioural findings have the strongest impact and which ones should be used under which circumstances. This leads to Lunn’s third point that the best chance for further integration of behavioural economics with policy-making will come from building capacity to engage in applied experimentation, piloting and evaluation. In this, behavioural economics has the advantage that there

M4716-STRAßHEIM_9781785367847_t.indd 51

25/04/2019 13:32

52  Handbook of behavioural change and public policy exists a professional network within its broader discipline that is already a part of the policy-making process in many countries.4 Oliver (2013), on the other hand, attributes the rise of behavioural economics to the failings of the neoclassical model, citing two main reasons. First is the 2008 financial crisis that was caused by insufficient regulation of the financial sector, which brought the shortcomings of the neoclassical economics model to the fore. The second reason is the search by liberal-minded politicians to find ways to motivate people to change their self-and-societal harming behaviour, without imposing interventionist measures such as bans and regulations. This argument is similar to that made by Straßheim et al. (2015) when explaining the greater institutionalization of behavioural economics in the UK as compared to Germany.5 According to them, behavioural ideas are ‘reframed as solutions to recent policy failures, promising to reconcile opposing views on the relationship between science, politics and citizens’ (p. 1), under the heading of libertarian paternalism. So far behavioural interventions linked to policy have withheld any overhauling changes to the existing setup. Instead, they have concentrated on how information is presented to actors, its salience, the convenience of different options and what actors know about others’ decisions (Lunn 2014). Such an approach makes it amenable to adoption in varied sectors and areas. This is also reflected in academic writing on the subject as the results in this chapter reveal.

3. METHODOLOGY The bibliometric analysis was carried out in two stages. In the first stage, leading journals in the discipline of economics and public policy were determined. In the next stage, articles were shortlisted and then analysed to answer the three key research questions. 3.1  Determining the Top Journals Three commonly referenced sources were looked at – Web of Science rankings (Web of Science | Thomson Reuters 2016), SCImago Journal rankings (SJR) (SJR: Scientific Journal Rankings 2016) and Google Scholar Metrics (Google Scholar Metrics 2016). Comparing journal rankings over time across three sources serves as a robustness check in determining those journals that have been widely regarded as among the top in their discipline for a long period of time. Every journal that appeared in the rankings was accorded a unique id. With the help of this id, a mean rank value for every journal was calculated (by dividing the sum of the ranks received with its frequency) for each source. Based on their mean rank values, the journals were placed in ascending order and this determined their ‘adjusted rank’. Finally, a cumulative rank value was obtained which was calculated as the average of the adjusted rank of the three sources. Journals that appeared in the top ten of all three sources were given precedence and those with the lowest cumulative rank value among them were selected. Additionally, journals that self-reported limitations on their scope or dealt exclusively with a particular sector were dropped. The final top three economic publications were, Quarterly Journal of Economics, Econometrica and the Journal of Political Economy. While the leading economics journals

M4716-STRAßHEIM_9781785367847_t.indd 52

25/04/2019 13:32

A bibliometric analysis of behavioural studies in economics and public policy journals  53 dated back to before the 1940s,6 in case of the top three public policy journals only one dated back to the 1930s (Public Administration).7 Hence, in order to facilitate comparison over a longer period of time, the fourth ranked journal – Public Administration Review, which dates to 1940, was also included in the analysis. Thus, the leading public policy journals that were finally included for analysis were Journal of Public Administration Research and Theory, Governance, Public Administration and Public Administration Review. Since it was only post-1940 that at least two public policy journals were available for analysis, the year 1940 was taken to be the cut-off point from which the analysis commenced. With respect to rankings, two limitations may be mentioned. First, the journal rankings for public policy were found to include both public policy and public administration journals.8 Second, journal rankings were not available before the 1990s.9 In the absence of rankings for the entire period of study, it was assumed that the available rankings were reflective of performance in preceding years as well.10 3.2  Shortlisting Relevant Articles For the leading journals in both the disciplines, searches for the appearance of key words in either topic, title or article abstract between 1940 until the end of 2016, were undertaken. The key words were: behaviour, economics, experiment, prospect, randomize, trial. These key words and their variants were as broad as possible, so as to minimize the chance of missing a relevant article, and tailored to the journals’ search engine. For economics journals, the word ‘economics’ was dropped. Following this, a deeper inspection of the shortlisted articles was carried out employing a pre-determined decision rule. For an article to qualify as a behavioural study, it had to either (1) incorporate behavioural insights or; (2) be based on an experimental approach for behavioural issues or non-theoretical real-world issues or; (3) combine both. Using this decision rule precluded the inclusion of laboratory-based theoretical or technical experimental studies that did not, in any way deal with behavioural issues.11 Applying the decision rule to the preliminary shortlist resulted in 195 behavioural studies in the leading economics journals and 116 behavioural studies in the top public policy journals. For both economics and public policy, more than half of all the articles fit the behavioural tag by way of the methodological criterion alone. Among economics articles 20 per cent of them fulfilled the ‘insight criterion’ alone; while 25 per cent fulfilled both. For policy journals, the ratio was similar, with a higher number of studies fulfilling the insight criterion alone (22 per cent) and 17 per cent fulfilling both. Each article was then individually analysed to answer the three key research questions.

4. FINDINGS The rise of behavioural studies in economics and public policy is not a recent ­phenomenon as evident in Figure 4.1, which depicts the relative share of behavioural articles to the total number of articles that have appeared over time in the leading economics and public policy journals. Thus, notwithstanding changes in the overall publishing trends, a cyclical pattern associated with behavioural studies in economics and public policy, is evident.

M4716-STRAßHEIM_9781785367847_t.indd 53

25/04/2019 13:32

54  Handbook of behavioural change and public policy

12.00 10.00

Percentage of behavioural studies in the top economics journals Percentage of behavioural studies in the top public policy journals

8.00 6.00 4.00 2.00 0.00

1950–1959 1960–1969 1970–1979 1980–1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016

Behavioural studies as a percentage of total published articles

14.00

Figure 4.1 Behavioural studies as a percentage of all articles published in the same time period 4.1  Temporal Patterns According to Oliver (2013), a degree of acceptance of behavioural economics by the mainstream economics community came with the publication in 1979 of Kahneman and Tversky’s prospect theory. However, as Figure 4.1 shows, this ‘acceptance’ translated into greater academic activity on the topic gradually. While behavioural studies took about two decades to gain traction in economics, in case of public policy this transition period was even longer at three decades. For the former, it was after 2000 that the share of behavioural studies doubled and entered a higher trajectory altogether. For public policy journals, it was only after 2013 that behavioural articles were at a higher level as compared to before. Behavioural studies in both economics and public policy exhibit a cyclical pattern, as is evident in periods of rise followed by periods of relative wane, suggesting the existence of something akin to Down’s (1972) issue-attention cycle among academics. Based on these patterns, three distinct phases can be demarcated in the trajectory of behavioural studies in public policy and economics over the entire time period. Phase I, between 1950 and 1999, is a period when both public policy and economics were moving in a similar fashion with regard to behavioural studies. In fact, public policy scholars appear to have taken the lead with economics scholars mirroring the trend.12 This is not surprising given that the idea of bounded human rationality, the cornerstone of behavioural economics, is also rooted in the works of early public policy scholars, Herbert Simon being the most prominent (see Simon 1955, 1976). Others such as Verba (1961) and Lindblom (1958, 1959) also noted the inadequacies of the rational model of man in different contexts. In Phase II, between 1999 and 2004, the disciplines were moving in divergent directions each reflecting Down-esque cycles of its own, with their peaks and troughs roughly coinciding with one another. Around the year 2000 there was a steep rise in behavioural studies in economics. This rise may be explained by the increased use of experiments in

M4716-STRAßHEIM_9781785367847_t.indd 54

25/04/2019 13:32

A bibliometric analysis of behavioural studies in economics and public policy journals  55 economics, both in the laboratory as well as natural and field experiments. In contrast, this was a time when there was virtually no research on behavioural issues in the top public policy journals. Phase III, between 2006 and 2015, reflects similar trends for both public policy and economics, with the latter being at a much higher level overall. This is also a period when public policy increasingly began to adopt experimental methods that went beyond the laboratory. The relative scarcity of experiments in public policy journals until then, according to Margetts (2011), says as much about the discipline (vis-à-vis methodological innovation) as it does about the challenges of the experimental approach. However, she postulates that barriers identified by scholars in the past (see Bozeman and Scott 1992) related to the experimental methodology, such as lack of research at the level of the individual, the need to establish ‘credibility’ with policy-makers superseding the pursuit of ‘truth’ or theoretical abstractions, other ethical and logistical barriers, have become much lower than they were twenty years ago. Post 2015, the trend between the two disciplines is hard to establish. While behavioural studies seem to be going out of favour with economists, they appear to be becoming very popular in public policy. The absolute number as well as the relative share of behavioural studies in public policy is greater than that in economics in 2016. However, this result may be driven by one of the top four public policy journals (Public Administration Review) running a special symposium on ‘Experimental Public Administration’ in 2016. If this journal is excluded then different patterns are exhibited for the year 2016. Until 2015, the top four and the top three public policy journals move in the same direction. However, in 2016 the top three public policy journals (after excluding Public Administration Review) like the economics journals, indicate a wane in behavioural studies, thereby suggesting a declining interest in behavioural issues in both the disciplines. Given the lack of information for 2017 at the time this chapter was written, it may be premature to conjecture about the future direction of behavioural studies in the two disciplines. 4.2  Geographic Spread The geographic spread of behavioural studies is relevant for those studies that had an applied component either through involvement of participants or by being undertaken in a real-world setting (laboratory, field or natural).13 Behavioural studies for both economics and public policy are heavily focused on the United States. This was especially true in the past, as Figure 4.2 indicates, and more so for public policy than economics. In economics after 2000, there is a spread in studies towards other European and Asian nations. Post 2010, the number of behavioural economics studies situated in the US is reduced for the first time, however, it still towers over studies in other nations. Among developing countries, the highest number of Asian and African studies are set in India and Kenya respectively. In public policy, on the other hand, until 1999 not a single behavioural study was located outside of the US. This may be rooted in the fact that until then experimental methods in public policy were used mostly in a laboratory setting. Unlike economists, public policy scholars were slow in catching on to the trend of using field and natural experiments. The mushrooming of behavioural studies outside of the US is driven by their ascent in European countries (between 2010 and 2016), chief among which are

M4716-STRAßHEIM_9781785367847_t.indd 55

25/04/2019 13:32

56  Handbook of behavioural change and public policy

70

Africa

60

Australia

50

South America Asia

40

Europe

30

North America (excluding USA)

20

USA

Economics

2010–2016

2000–2009

1990–1999

1980–1989

1950–1979

2010–2016

2000–2009

0

1990–1999

10 1980–1989

Number of behavioural studies

80

Public Policy

Figure 4.2 Spatial and temporal spread of behavioural studies in economics and public policy Denmark, the Netherlands and the United Kingdom. While a smattering of Asian and South American nations have been studied, not a single study dealing with the behavioural aspects of public policy in the African continent exists. Part of the reason for this skewed regional trend in behavioural studies in public policy may be explained by the general tendency of the discipline of public policy to focus on western countries. In comparison, within economics, development economics constitutes a sub-field in itself. An offshoot of this has been the osmosis of behavioural studies in developing countries (Basu 2014). 4.3  Sectorial Spread Behavioural studies in both economics and public policy have looked at a wide spectrum of sectors (Figure 4.3). Studies focusing on psychological insights and behavioural interventions, the health and education sector feature prominently in both disciplines. Similarly, studies that look at decision-making under risk, uncertainty, trust, fairness and satisfaction are also common in both. In economics, the financial sector, identity and diversity are important areas of study. Tied to the spread of the RCT methodology in the development sector, studies in economics focus on issues such as migration, poverty, corruption, natural resources and labour. However, more abstract issues related to individual attitudes, beliefs and emotions, find place in public policy journals much more than economics. This may have to do with the nature of the discipline itself. Scholars of management and public administration, starting with Maslow (1943), have stressed on the role of individual attitudes, values and emotions. Disaggregated over sectors are numerous behavioural economics studies that have to

M4716-STRAßHEIM_9781785367847_t.indd 56

25/04/2019 13:32

57

M4716-STRAßHEIM_9781785367847_t.indd 57

25/04/2019 13:32

0

10

20 Number of behavioural articles

30

Figure 4.3 Comparison between selected areas of behavioural studies in economics and public policy journals

Psychology and nudges Miscellaneous

Financial: pensions, savings, taxation, budgeting Identity and diversity: gender, ethnicity, race Decentralization and local governance Performance management Participation: voice, co-production, collaboration and cooperation Health and education Trust, risk, fairness, uncertainity, satisfaction and acceptance Attitudes, emotions, values and beliefs

Justice: crime, policing, judiciary Youth, children and family Law, welfare and voting

Information: provision and usage Performance Public services: delivery, provision, quality Decision making, behavior: bureaucrats, managers, administrators

Accountability and transparency Public service motivation Inter sector (public, private, NGO): differences and integration

Migration, poverty and corruption Labour, employment and recruitment

40

Economics Public Policy

58  Handbook of behavioural change and public policy do with crime, policing, judiciary; housing, welfare; youth and children. Many of these studies, especially those that were based on natural experiments involving governmental action, could well have featured in public policy journals. This suggests that the two disciplines seem to be showing some convergence, public policy – through its use of the experimental methodology and insights from behavioural economics; economics – through its applications in a public policy context. In contrast to economics, mapping the decision-making behaviour (of bureaucrats, managers, administrators) is a major point of study for public policy. As is performance and performance management. Studies related to participation and governance, public versus private sector, information provision, public service motivation and accountability are also common in public policy. Many behavioural studies in public policy focus on the provision of services as a whole, instead of studying them on a sector-by-sector basis which economics tends to do.

5.  IMPLICATIONS, CONCERNS AND REAL-WORLD ISSUES The previous section showed that the number of behavioural studies in both disciplines moved to a higher trajectory after 2006. Associated with this was an increase in the number of articles that satisfied the ‘behavioural criterion’ on grounds of methodology. This raises some points of concern. First is the question whether ‘method’ alone qualifies for a study to be classified as behavioural (Galizzi 2014). Though this is widely prevalent, it opens the door for bringing in more theoretical clarity to the concept. It becomes especially essential to do so, because behavioural economics which has been regarded as the vehicle of change for behavioural policy, is itself criticized for its inability to produce a framework that can be easily applied to other circumstances (Shiller 2005). The second point relates to the experimental methodology itself. While experiments are regarded to have high internal validity (Shadish et al. 2001), this comes at the cost of external validity. This means that RCTs can play an important role in building scientific knowledge but not as a stand-alone approach. They need to be combined with other methods, including conceptual and theoretical development, in order to answer questions beyond ‘what works?’ to ‘why it works’ (Deaton and Cartwright 2016). Even answering, the ‘what works?’ requires constant experimenting and piloting in different contexts. This may be difficult to implement in certain public policy situations. Administrators may oppose random assignment because of ethical issues in not serving all who may be eligible for a programme. Random assignment could, along with changing the nature of the programme, also affect the clients who seek it (Manski and Garfinkel 1992). In addition, experiments are associated with problems of getting permission and cooperation, attrition and differences between treatments assigned and treatments delivered (Farrington and Welsh 2005). Experiments can also induce distortion in other ways, for instance an evaluation of a pilot project may not be able to capture systemic effects that would occur if the programme were instituted permanently or on a larger scale (Manski and Garfinkel 1992). The third point relates to the increased tendency of studies to position themselves as experimental. This pattern is visible in numerous articles that termed themselves as good as experiments because of events that rendered some degree of randomization among the subjects. In this context, transparency in reporting is crucial because a biased estimate of

M4716-STRAßHEIM_9781785367847_t.indd 58

25/04/2019 13:32

A bibliometric analysis of behavioural studies in economics and public policy journals  59 effect may be unquestioningly accepted if it is not possible to recognize a rigorous RCT from a weakly designed one (Boutron et al. 2010). Thus, it is important that researchers pay attention to design and implementation, with the communication of results in a way that enables further use by the research community (James 2011). The fourth point relates to the increase in behavioural studies that use Internet research panels for surveying experiments (see for instance Baekgaard and Serritzlew 2016; Riccucci et al. 2016), especially in public policy journals. Some of these research panels, such as CivicPanel project, are university affiliated (Riccucci et al. 2016). Others, such as Amazon.com’s Mechanical Turk (MTurk), are essentially an online labour market in which people receive payment for participating in surveys (Marvel and Girth 2016). It may be argued that there is room for bias as the subject pool comprises those who self-select themselves into participating in Internet-based surveying. Supporters, however, argue that the Internet can be a reliable way to recruit samples and reduce biases (Buhrmester et al.  2011; Gosling et al.  2004) and may even be more representative than in-person convenience samples which commonly rely on student participation (Berinsky et al. 2012). This raises questions about judging the ‘quality of a panel’ and differences that may result from panels that offer monetary compensation and those that don’t. But exploring this issue is beyond the scope of this chapter. The final point ties in with the debate relating to the use of ‘nudges’ in public policy. One of the criticisms directed at a nudging policy is that regulators and administrators are behavioural agents themselves and thus, their judgments may suffer from psychological biases as well (Viscusi and Gayer 2015). As the analysis showed, a number of studies in public policy explore decision-making among administrators and managers. This suggests that there is room to go even further by mapping psychological biases that relate directly to an administrator or regulator’s use of nudging tactics. Awareness of such biases with respect to ‘nudges’ could aid policy design.

6.  CONCLUDING REMARKS Through a bibliometric analysis of the leading economics and public policy journals, this chapter presented results related to the development of ‘behavioural’ studies over time. It found that behavioural studies in the two disciplines follow cycles of their own. Compared to economics, the geographic spread of behavioural studies in public policy is more restricted. Certain sectors feature in both disciplines, whereas others exhibited a notable disciplinary bias. To the extent this analysis is restricted to published journal articles (excluding openaccess writing), it may have limited the representativeness of academic thought that it seeks to collate. However, it can be argued that influential open-access writing usually finds its way to peer-reviewed journals, albeit with a time lag. This, for instance, is seen in the case of NBER papers, many of which result in published studies. Another limitation stems from the fact that the results may reflect a restricted universe of institutional thought and knowledge – one that is circumscribed by the search terms used, the varying efficacy of search engines and the subjectivity involved in judging whether an article qualified as a behavioural study. In order to counter this, the key words and the decision rule employed, were made as broad as possible.

M4716-STRAßHEIM_9781785367847_t.indd 59

25/04/2019 13:32

60  Handbook of behavioural change and public policy Finally, the underlying assumption of this chapter has been that the presence of behavioural studies in economics and public policy journals is reflective of, and reflected by, research interests and real-world relevance. However, academic writing is often influenced by factors such as funding grants, publishing and acceptance trends. It may be interesting to explore in future research the impact of such factors on the academic research agenda.

NOTES  1. Though it may be noted that authors such as Sent (2004) argue that behavioural economics in the Kahneman and Tversky tradition follows a model of rationality which is different from Simon’s because of the former’s implicit acceptance of the ‘homo economicus’ model as a benchmark for what they term ‘biases’.   2. Simon (1976) makes a distinction between different kinds of rationality. A decision is regarded as ‘objectively’ rational if it is the correct behaviour for maximizing given values in a given situation.   3. Though the inclusion of the methodology of randomized experiments under the ‘behavioural’ category has been regarded as contentious (see Galizzi 2014), this chapter justifies their inclusion because a set of economists regard the focus on methods to be a key aspect of behavioural economics (see Shiller 2005), much as many practitioners do (see Haynes et al. 2013).   4. For instance, Sendhil Mullainathan’s presence in the US Consumer Financial Protection Bureau, Sunstein and Thaler’s role as advisers to UKBIT respectively, at the time this chapter was written.   5. The other reasons they cite are: (i) behavioural economics gains politico-epistemic authority by combining political practicality with experimental expertise; and (ii) differences between the UK and Germany related to domestic cultures of expertise influence how knowledge claims are validated and valued.  6. Quarterly Journal of Economics, Econometrica and Journal of Political Economy go as far back as 1886, 1933 and 1892 respectively.  7. Journal of Public Administration Research and Theory and Governance date back to 1991 and 1988 respectively.  8. Only Google Scholar listed ‘Public Policy and Administration’ as a separate discipline. Both Web of Science and SJR listed public policy and public administration journals under the category – ‘Public Administration’. Thus, ‘public policy’ journals in this chapter technically refers to journals in both public policy and administration.   9. Web of Science ranked journals annually from 1997 to 2014. SJR ranked journals from 1999 to 2014. However, Google scholar rankings (which also included open source publications) were not available for every year. 10. This assumption may be reasonable, given that the first rankings whenever they came out would be based on journals’ performance and reputation as established in earlier years. 11. Studies that were technically quasi-experimental but argued that they involved processes that were almost equivalent to randomization, were accorded the benefit of doubt and included in the ‘behavioural category’. 12. Is the rise of behaviourism in public policy since the 1950s comparable to the ‘new’ behaviourism around 2006? See the chapter by Graf in this volume. 13. Studies that provided a theoretical exposition of behavioural components were not included.

REFERENCES Ashraf, N., Camerer, C. and Lowenstein, G. (2005), ‘Adam Smith, Behavioral Economist’, Journal of Economic Perspectives, 19 (3), 131–145. Baekgaard, M., and Serritzlew, S. (2016), ‘Interpreting Performance Information: Motivated Reasoning or Unbiased Comprehension’, Public Administration Review, 76 (1), 73–82. Banerjee, A. and Duflo, E. (2012), Poor Economics, London: Penguin Books. Basu, K. (2014), ‘Randomisation, Causality and the Role of Reasoned Intuition’, Oxford Development Studies, 42 (4), 455–472. Berinsky, A., Huber, G. and Lenz, G. (2012), ‘Evaluating Online Labor Markets for Experimental Research: Amazon.com’s Mechanical Turk’, Political Analysis, 20 (3), 351–368.

M4716-STRAßHEIM_9781785367847_t.indd 60

25/04/2019 13:32

A bibliometric analysis of behavioural studies in economics and public policy journals  61 Boutron, I., John, P. and Torgerson, D. (2010), ‘Reporting Methodological Items in Randomized Experiments in Political Science’, The Annals of the American Academy of Political and Social Science, 628 (1), 112–131. Bozeman, B. and Scott, P. (1992), ‘Laboratory Experiments in Public Policy and Management’, Journal of Public Administration Research and Theory, 2 (3), 293–313. Buhrmester, M., Kwang, T. and Gosling, S. (2011), ‘Amazon’s Mechanical Turk: A New Source of Inexpensive, Yet High-Quality, Data?’, Perspectives On Psychological Science, 6 (1), 3–5. Cooper, J. (2010), Review into the Governance, Efficiency, Structure and Operation of Australia’s Superannuation System, Canberra: The Review. Dahl, R. (1947), ‘The Science of Public Administration: Three Problems’, Public Administration Review, 7 (1), 1. Datta, S. and Mullainathan, S. (2014), ‘Behavioral Design: A New Approach to Development Policy’, Review of Income and Wealth, 60 (1), 7–35. Deaton, A. and Cartwright, N. (2016). Understanding and Misunderstanding Randomized Controlled Trials (No. w22595). National Bureau of Economic Research. Downs, A. (1972), ‘Up and Down with Ecology-the Issue-Attention Cycle’, Public Interest, 28 (Summer), 38. Farrington, D. and Welsh, B. (2005), ‘Randomized Experiments in Criminology: What Have We Learned in the Last Two Decades?’, Journal of Experimental Criminology, 1 (1), 9–38. Ferraro, P. and Price, M. (2013), ‘Using Nonpecuniary Strategies to Influence Behavior: Evidence from a LargeScale Field Experiment’, Review of Economics and Statistics, 95 (1), 64–73. Galizzi, M. (2014), ‘What Is Really Behavioral in Behavioral Health Policy? And Does It Work?’, Applied Economic Perspectives And Policy, 36 (1), 25–60. Glennerster, R. and Takavarasha, K. (2013), Running Randomized Evaluations, Princeton, NJ: Princeton University Press. Google Scholar Metrics. (2016), Scholar.google.com. Retrieved 1 April 2016, from https://scholar.google.com/ citations?view_op=top_venues. Gosling, S., Vazire, S., Srivastava, S. and John, O. (2004), ‘Should We Trust Web-Based Studies? A Comparative Analysis of Six Preconceptions About Internet Questionnaires’, American Psychologist, 59 (2), 93–104. Haynes, L., Service, O., Goldacre, B. and Torgerson, D. (2013), ‘Test, Learn, Adapt: Developing Public Policy with Randomised Controlled Trials’, UK: Cabinet Office Behavioural Insights Team. Retrieved 13 October 2015, from http://38r8om2xjhhl25mw24492dir.wpengine.netdna-cdn.com/wp-content/uploads/2015/07/TLA1906126.pdf. James, O. (2011), ‘Performance Measures and Democracy: Information Effects on Citizens in Field and Laboratory Experiments’, Journal Of Public Administration Research And Theory, 21 (3), 399–418. Kahneman, D. (2012), ‘Foreword’. In E. Shafir (ed.), The Behavioral Foundations of Public Policy (1st edn), Princeton, NJ: Princeton University Press, pp. vii–ix. Kahneman, D. and Tversky, A. (1979), ‘Prospect Theory: An Analysis of Decision under Risk’, Econometrica, 47 (2), 263. Karlan, D. and Appel, J. (2011), More than Good Intentions, New York: Dutton. Levitt, S. and List, J. (2009), ‘Field Experiments in Economics: The Past, The Present, and The Future’, European Economic Review, 53 (1), 1–18. Lindblom, C. (1958), ‘Policy Analysis’, The American Economic Review, 48 (3), 298–312. Lindblom, C. (1959), ‘The Science of “Muddling Through”’, Public Administration Review, 19 (2), 79. Low, D. (2012). Behavioural Economics and Policy Design: Examples from Singapore. Singapore; Hackensack, NJ: World Scientific. Lunn, P. (2013), ‘Behavioural Economics and Policymaking: Learning from the Early Adopters’, The Economic and Social Review, 43 (3, Autumn), 423–449. Lunn, P. (2014), Regulatory Policy and Behavioural Economics, Paris: OECD Publishing. Manski, C., and Garfinkel, I. (1992), Evaluating Welfare and Training Programs, Cambridge, MA: Harvard University Press. Margetts, H. (2011), ‘Experiments for Public Management Research’, Public Management Review, 13 (2), 189–208. Marvel, J.D. and Girth, A.M. (2016), ‘Citizen Attributions of Blame in Third-Party Governance’, Public Administration Review, 76 (1), 96–108. Maslow, A. (1943), ‘A Theory of Human Motivation’, Psychological Review, 50 (01/1943), 370–396. Oliver, A. (2013), ‘From Nudging to Budging: Using Behavioural Economics to Inform Public Sector Policy’, Journal of Social Policy, 42 (4), 685–700. Riccucci, N., Van Ryzin, G. and Li, H. (2016), ‘Representative Bureaucracy and the Willingness to Coproduce: An Experimental Study’, Public Administration Review, 76 (1), 121–130. Schneider, A. and Ingram, H. (1997), Policy Design for Democracy, Lawrence, KS: University Press of Kansas. Sent, E.M. (2004), ‘Behavioral Economics: How Psychology Made its (Limited) Way Back into Economics’, History of Political Economy, 36 (4), 735–760.

M4716-STRAßHEIM_9781785367847_t.indd 61

25/04/2019 13:32

62  Handbook of behavioural change and public policy Shadish, W., Cook, T. and Campbell, D. (2001), Experimental and Quasi-experimental Designs for Generalized Causal Inference, Boston, MA: Houghton Mifflin. Shepherd, J. (2007), ‘The Production and Management of Evidence for Public Service Reform’, Evidence and Policy: A Journal of Research, Debate and Practice, 3 (2), 231–251. Shiller, R. (2005), ‘Behavioral Economics and Institutional Innovation’, Southern Economic Journal, 72 (2), 268. Simon, H. (1955), ‘A Behavioral Model of Rational Choice’, Quarterly Journal of Economics, 69 (1), 99–118. Simon, H. (1957), Administrative Behavior: A Study of Decision-Making Processes in Administrative Organization, New York: MacMillan. Simon, H. (1976), Administrative Behavior (3rd edn), New York: Free Press. SJR: Scientific Journal Rankings (2016), Scimagojr.com. Retrieved 30 March 2016, from http://www.scimagojr. com/journalrank.php. Smith, A. (1759), The Theory of Moral Sentiments [1981]. Edited by D.D. Raphael and A.L. Macfie, Indianapolis, IN: Liberty Classics. Stewart, J. (1993), ‘Rational Choice Theory, Public Policy and the Liberal State’, Policy Sciences, 26 (4), 317–330. Stone, D. (2002), Policy Paradox, New York: Norton. Straßheim, H., Jung, A. and Korinek, R. (2015), ‘Reframing Expertise: The Rise of Behavioral Insights and Interventions in Public Policy’, Moments of Valuation: Exploring Sites of Dissonance, 249–268. Sunstein, C. (2011), ‘Empirically Informed Regulation’, The University of Chicago Law Review, 78 (4), 1349–1429. Sunstein, C. and Thaler, R. (2003), ‘Libertarian Paternalism is not an Oxymoron’, The University of Chicago Law Review, 1159–1202. Thaler, R. and Mullainathan, S. (2000), ‘Behavioral Economics’, In: International Encyclopedia of the Social and Behavioral Sciences (1st edn), London: Pergamon Press. Thaler, R. and Sunstein, C. (2008), Nudge, New Haven, CT: Yale University Press. Thomas, D. (1997), ‘Sailors, Scurvy and Science’, Journal of the Royal Society of Medicine, 90 (1), 50–54. Van Bavel, R., Herrmann, B., Esposito, G. and Proestakis, A. (2013), ‘Applying Behavioural Sciences to EU Policy-making’, Joint Research Centre Scientific and Policy Reports, 1‒21. Retrieved 13 October 2015, from http://www.capire.org/capireinforma/scaffale/30092013_jrc_scientific_policy_report_en.pdf. Verba, S. (1961), ‘Assumptions of Rationality and Non-Rationality in Models of the International System’, World Politics, 14 (01), 93–117. Viscusi, W. and Gayer, T. (2015), ‘Behavioral Public Choice: The Behavioral Paradox of Government Policy’, Vanderbilt Law and Economics Research Paper, 15–2, 972–1007. Web of Science | Thomson Reuters (2016), Thomsonreuters.com. Retrieved 1 April 2016, from http://thomson​ reuters.com/en/products-services/scholarly-scientific-research/scholarly-search-and-discovery/web-of-science. html.

M4716-STRAßHEIM_9781785367847_t.indd 62

25/04/2019 13:32

5.  The concepts of nudge and nudging in behavioural public policy Pelle Guldborg Hansen

1. INTRODUCTION The past four decades, advances in behavioural economics, cognitive and social psychology, henceforth the behavioural sciences, have revealed how human decision-making is boundedly rational (Simon 1955), systematically biased (Tversky and Kahneman 1974) and strongly habitual (Neal et al. 2011), owing to the interplay of psychological forces with what ought to be, from the perspective of rationality, irrelevant contextual features of complex decision-making (see Kahneman 2011; Thaler 2016). These behavioural insights teach us how people may systematically fail to act on their well-informed preferences (Ariely 2010) and thus fail to achieve their preferred ends. In addition, such advances also teach us how neglecting these insights can cause public policies to fail in achieving intended effects and why paying more attention to them may provide the key for dealing effectively with main challenges facing modern social organizations (Shafir 2012; Thaler and Sunstein 2009). At least this is what Richard Thaler and Cass Sunstein suggested in Nudge: Improving Decisions about Health, Wealth and Happiness (2008). In particular, they suggested that if an unfortunate decision-making pattern is the result of cognitive boundaries, biases or habits, this may be ‘nudged’ towards a more preferred one by integrating insights about the very same kind of boundaries, biases and habits into the choice architecture surrounding that behaviour – that is, the totality of physical, social and psychological aspects of the contexts that influence our choices (see also Thaler et al. 2010). They also argued that ‘nudges’ may avoid some of the challenges facing traditional public policies, such as costly procedures, ineffective campaigning, unintended effects of incentivizing behaviours and invasive choice regulation, such as bans. The advantage, they write, of applying nudges is that public policy-makers might thus supplement – or, perhaps, at times even replace (Thaler and Sunstein 2009, p. 14) – traditional regulation with nudges to influence citizens’ behaviour in cheaper, less invasive and more effective ways without further restricting freedom of choice, imposing mandatory obligations or changing incentives. In Thaler and Sunstein’s original definition of a ‘nudge’ the absence of these traditional policy strategies is even stated as a formal condition, saying that a nudge alters behaviour without forbidding any options or significantly changing economic incentives (Thaler and Sunstein 2009, p. 6). They also coined the seemingly oxymoronic term, libertarian paternalism, to characterize the behavioural public policy paradigm that arises out of this approach when enacted to serve the interests of the citizens as judged by themselves (see Thaler and Sunstein 2003, 2009; Hansen 2016). ‘Libertarian’ because nudges preserve the original choice sets of citizens and ‘paternalistic’ because nudges despite their freedom preserving nature, aim to alter the behaviour of citizens, albeit in their own interests. That 63

M4716-STRAßHEIM_9781785367847_t.indd 63

25/04/2019 13:32

64  Handbook of behavioural change and public policy is, two ideologies in public regulation that are usually seen as incompatible, but which the concept of nudge proves to be reconcilable. According to Thaler and Sunstein, however, this paradigm is not only possible, but also ethically non-controversial (see Thaler and Sunstein 2009, pp. 239‒254; Sunstein 2016). Since there are no situations without choice architecture and no choice architecture without contextual features that influence behaviour, nudges are unavoidable and hence an antinudge position a literal non-starter (Thaler and Sunstein 2009, p. 11). This may be called the argument of unavoidability. In addition, since nudges as libertarian paternalistic measures neither forbid options nor significantly change incentives, they are liberty preserving (Thaler and Sunstein 2009, p. 5) and hence may be regarded as trivially acceptable public policy strategy (see Hansen and Jespersen 2013). Nevertheless, the reception of Nudge has been distributed over a continuum. At one end policy-makers and consultants have enthusiastically accepted Thaler and Sunstein’s arguments as a matter of authority. Little attention have been paid to the fine print in academic discussions as if the points made there, are irrelevant and only of intellectual interest. At the other end academic critics have raised concerns about nudging and libertarian paternalism, the generic critiques being: that the concept of nudge is ill-defined (e.g. House of Commons 2011; Mongin and Cozic 2014) and a return to present-day Behaviourism (Burgess 2012), hence its adoption may easily lead to slippery slopes that ends with shoves (e.g. Adams 2014; Rizzo and Whitman 2009; Smith and Zywicki 2014; Tierny 2014); that nudges are short-term subconscious influences that work best ‘in the dark’ (e.g. Bovens 2008) and thus easily may be aimed at manipulating citizens’ choices in democratic societies (e.g. Dunt 2014); and yet others describe nudging as identical to libertarian paternalism (e.g. Lunn 2014), thereby unintentionally giving rise to musings about how this ideology may both legitimize overreach (mostly in the US) as well as a roll-back (mostly in Europe) of the state (see Horton 2009; Leggett 2014). This contrast of extremes reflects a natural state of confusion, which is only to be expected in a young and emerging field such as nudging. Of course, we may progress through academic writings like this as we have done so many times before. But in Behavioural Public Policy where stakeholders with different backgrounds adopt our ideas in real time, applying these to citizens, the obscurity of an immature paradigm has real consequences. Therefore we need to establish and communicate a clear grasp of and agreement on the basic concepts and their implications in order to safeguard citizens from undue interventions as well as protect our work from being rejected or even misused due to careless interpretation.

2. THE THEORETICAL FOUNDATIONS OF BEHAVIOURAL PUBLIC POLICY During the twentieth century the rational foundations of public policy came close to perfection. Stretching its roots back through the revolutions, royal courts and monasteries of past centuries, Law had already required even illiterate peasants to be as clever in their belief formation as a cleric bonus pater for a long time. But with the arrival of mathematical probability and decision theory in the late nineteenth and early twentieth century even the choices of citizens seemingly became predictably rational (see Dimand and Dimand

M4716-STRAßHEIM_9781785367847_t.indd 64

25/04/2019 13:32

The concepts of nudge and nudging in behavioural public policy  65 1996). Since then public policy has been ruled by economists and lawyers in what may be referred to as the paradigm of traditional public policy. According to this, human behaviour may be explained as the result of careful consideration of subjective expected utility of consequences of possible actions based on rational belief formation given the available information and well-defined preferences – if not as a matter of realism, then at least as a suitable working hypothesis that will come true in the long run (Friedman 1966). Not surprisingly then, policy strategies rising from traditional public policy were cut to suit the intellectual digestive system of homo economicus; necessary and sufficient for influencing this creature of imagination was the trinity of rational regulation, that is, the provision of information and rational persuasion, negative and positive incentivising and legal regulation. From the perspective of behavioural science rational regulation is by no means a coincidence. Leaving philosophical traces aside, Premack and Woodruff’s (1978) research on what is called Theory of Mind (ToM), has shown that it is an innate ability of primates to take the intentional stance (Dennett 1987) in the observation of behaviour, interpreting this by attributing mental states, such as ‘desires’, ‘beliefs’ and the like, with the consequence reconstructing behaviour in terms of ‘decisions’ and ‘actions’ rather than instincts. It is thus easy to see why traditional public policy is rooted in rationality. Given that an action is attributed intentionality, influencing behaviour by inducing beliefs through the provision of information, relating such information to higher order desires and attaching negative and positive sanctions as consequences of choice options to affect the construction of preferences in the deliberative process become necessary and sufficient tools for influencing our decisions and hence our resulting actions. However, this perspective is not the only contribution of behavioural science. Another one is that the appearance of behaviour as rational is just as often a cognitive illusion, as not. Thus, while rational regulation is deeply rooted in our intuitive experience of human behaviour as rational, it is also out of tune with the behavioural scientific understanding of human decision-making as somewhat more complex than what meets the mind’s eye. The result has been a lack of behaviourally informed foundations of public policy that may be suggested as part of what is behind the inability of rational regulation to effectively address problems that are ultimately rooted in non-rational aspects of human behaviour, such as the obesity epidemic (Hansen et al. 2016), behavioural market failures (Sunstein 2014), climate change (Bhargava and Loewenstein 2015) and social inequality (Mullainathan and Shafir 2013). While behaviour causing these problems are often intentional, psychological factors seem to cause an irrational departure from the norms of rational decision-making that we do not readily intuit. Fortunately, the behavioural sciences have proven not to be limited to diagnosing the failings of traditional public policy. They have also made marked progress in providing an alternative view of human behaviour providing the foundations of the new and emerging paradigm of Behavioural Public Policy. Already in the aftermath of the Second World War Cognitive and Social psychology had begun to uncover empirical anomalies from what the deductive theories of rationality at the heart of traditional public policy predict (see Hewstone et al. 2012). The theoretical impact of the observations of such anomalies was only of secondary interest to psychology, though. However, the systematic empirical study of the bounds of reason in the work of Simon (1955), the errors of judgment and decision-making anomalies in the work of

M4716-STRAßHEIM_9781785367847_t.indd 65

25/04/2019 13:32

66  Handbook of behavioural change and public policy Kahneman and Tversky (1979; Tversky and Kahneman 1974, 1981) and the birth of Behavioural Economics delivered by Thaler (1985, 1992) and others, came to pose, a challenge from within to rationality as the proper foundation of the study and regulation of human behaviour. The result was a behavioural approach that came to regard recurring departures observed from the predictions of rationality as systematic ‘bias’, rather than anomalies, resulting from cognitive mechanisms – hence the label, cognitive bias (see Gilovich et al. 2002). The picture emerging of human agency ultimately became one of bounded rational slow reflective agency imbedded within and fed by fast non-rational automatic processes activated and affected by subtle contextual features that should not matter ‘in principle’, that is according to the theory of rationality, but yet affect behaviour systematically ‘in practise’. Thus, humans became predictively irrational (Kahneman 2011). Academic discussion ensued in what is called the ‘Great Rationality Debate’ (see Tetlock and Mellers 2002); a debate about whether best to understand bias as cognitive errors of reasoning (Kahneman and Tversky 2000; Gilovich et al. 2002), or if they are better understood as optimized adaptations in reasoning heuristics, a kind of fast and frugal habits of the mind, making them ‘ecologically rational’ in a wider perspective (Gigerenzer 1991; Gigerenzer et al. 1999). Yet, consensus seems to be emerging that while reasoning should not be regarded as merely dysfunctional, nor should cognitive bias and heuristics merely be rationalized, since while any kind of behaviour may in principle be rationalized according to rationality theory, such immunization comes at the cost of a theory’s scientific status. Thus, from the perspective of behavioural science, cognitive biases do exist and are likely to cause irrational decision patterns, especially given the presence of subtle, from the perspective of rationality, irrelevant features in unfamiliar and complex decision-making contexts. It is in terms of our bounded rationality, bias, and habits, then, that the behavioural sciences provide a new foundation for public policy to approach societal problems as potentially caused by the fallibility of human reasoning. The theoretical underpinnings of Behavioural Public Policy, though, have also raised suspicion that we are dealing with a return of Behaviourism implying an approach to citizens as laboratory rats to be influenced by subconscious means. Looking closer at the psychology underlying behavioural public policy such suspicions may, however, easily be put at rest. This is because the interplay of non-rational features, automatic processes and reflective reasoning aspiring to the ideals of rationality is the scientific object of so-called Dual Process Theories (DPTs) (Gawronski et al. 2014). Originating in social psychology DPTs vary greatly, but generally share the overarching structure of positing two qualitatively different kinds of reasoning – automatic and non-automatic – in explaining and predicting behaviour (Evans 2008). It is by identifying processes according to this simplified distinction that DPTs seek to explain how the supposedly irrelevant features of decision-making contexts systematically influence behaviour (Kahneman 2011). Using Marr’s (1982) distinction between computational, algorithmic and implementational-level theories of psychology,1 the explanatory function of DPTs may be located at the algorithmic level of analysis, while Behaviourism is best viewed as a computational level theory (Gawronski et al. 2014). Hence, the rooting of Behavioural Public Policy in DPTs reveals that it is not Behaviourism.2 In addition, an important point to make in this connection is that while DPTs assert that automatic processes do not require conscious awareness or intentions to work, they do not necessarily manifest themselves as ‘subconscious’

M4716-STRAßHEIM_9781785367847_t.indd 66

25/04/2019 13:32

The concepts of nudge and nudging in behavioural public policy  67 influences. Different from reflexes some bias may actually be observed by introspection as well as blocked by means of self-regulation (Stanovich 2012). Hence, although some of the most memorable examples of bias are memorable due to their non-conscious nature such strategies should not be mindlessly characterized as subconscious influences – that is, as inaccessible to internal observation, monitoring and control – as sometimes done by scientists, journalists and policy-makers alike.

3.  THE DEFINITION OF NUDGE At the centre of Behavioural Public Policy has been the concept ‘nudge’ as coined by Thaler and Sunstein (2008). Yet, as mentioned above conceptual confusion has reigned in the wake of its introduction. The concept has often been treated as identical to that of ‘libertarian paternalism’ (see e.g. Lunn 2014; Sunstein 2014, p. 58)3 or said to be ill-defined (House of Commons 2011, p. 84; Mongin and Cozic 2014). This latter claim in turn has provided for claims such as that the nudge approach may easily lead to motivated slippery slopes (Rizzo and Whitman 2009; Smith and Zywicki 2014, p. 12) and public musings about how nudges may quickly turn into shoves (see e.g. Adams 2014; Tierny 2014). Initially these claims may easily be dismissed. Thaler and Sunstein seemingly provide a clear-cut definition: ‘A nudge, as we will use the term, is any aspect of the choice architecture that alters people’s behavior in a predictable way without forbidding any options or significantly changing their economic incentives’ (Thaler and Sunstein 2009, p. 6). Yet, as pointed out in Hansen (2016) looking closer at this definition one finds that the concept of nudge is not only ill-defined; it is actually not defined at all. As for being ill-defined, the definition is silent on several dimensions necessary for it to serve its declared purpose of capturing the behavioural insights challenging the foundations of traditional public policy, that is, as Thaler and Sunstein write, that of providing a definition according to which ‘a nudge is any factor that significantly alters the behavior of Humans, even though it would be ignored by Econs’ (Thaler and Sunstein 2008, p. 8). For instance, their definition does not rule out the addition of strictly dominating choice options to agents’ pre-intervention strategy sets as nudges. Thus, adding a strictly dominating strategy (such as a big cold Duff beer on the couch) to an agent’s options (taking out the trash versus cleaning the dishes) would qualify as a nudge on their definition. Second, their definition does not exclude material and social incentives, such as those imposed by time, trouble, social sanctions and sandwiches. Adding 100 volts electro-shock to the couch, a Prize Donut for choosing to take out the trash or public shaming if you do not clean the toilet all count as nudges on the definition above. While such examples obviously do not square with their intention, the shortcomings of the definition leading to these, even seem to mislead Thaler and Sunstein themselves at times. In later writings Sunstein thus asserts that shaming resulting from defecting from social norms constitutes a nudge (Sunstein 2016, p. 19). Yet, behaviour change strategies such as these are clearly rationality-based interventions aimed at changing the utility functions of citizens, albeit in creative ways. Third, their definition does not specify what is meant by ‘significantly changing economic incentives’. This seems problematic – after all, any change to economic incentives that alters people’s behaviour in a predictable way would seem to be significant by definition.

M4716-STRAßHEIM_9781785367847_t.indd 67

25/04/2019 13:32

68  Handbook of behavioural change and public policy As for the claim that the concept is not defined at all, Thaler and Sunstein’s definition suffers from the shortcomings just mentioned because it does not stipulate necessary and sufficient conditions for what counts as a nudge. It only defines what a nudge is not. That is, besides stipulating a nudge as any aspect of choice architecture that alters people’s behaviour in a predictable way, it only says that a nudge is not about forbidding any options and not about significantly changing economic incentives. Regarded as a definition it thus says little more about what a nudge is than a definition of ‘human’ saying that this is an animal that is not a mouse and not an elephant. 3.1  Approaching the Concept of Nudge As a result four different approaches to defining the concept of nudge have evolved. The first claims the authority of Thaler and Sunstein’s original definition. Despite that it is not a definition and that its adoption leads to inconsistencies, this is actually the path taken in most current academic papers. Whether out of convenience or social conformity, sloppiness with regards to the most basic concept has apparently become the descriptive norm in a scientific community studying the very dynamics of informational cascades and pluralistic ignorance. Different from the authority approach, the two next approaches see Thaler and Sunstein as defining nudges primarily by means of examples (see e.g. Bovens 2008; Hausman and Welch 2010; Mongin and Cozic 2017); an assumption that the authors themselves seem to affirm in later work (see e.g. Sunstein 2016, pp. 20‒21). The reconstructive approach then seeks to reconstruct a unified definition of the concept by examining the main structural similarities of the examples and semantics of usages of the concept in Nudge (see e.g. Bovens 2008; Hausman and Welch 2010). Contrary to this the descriptive approach argues that more than one concept of nudge is at play, none of which seemingly can or should be treated as primary (see Mongin and Cozic 2014, 2018). Finally there is the theoretical approach. This tries to provide a clear and consistent definition based on the theoretical underpinnings of behavioural science and the intentions of Thaler and Sunstein. This approach holds that a definition should be built upon the behavioural insights challenging the foundations of traditional public policy and contained in the principle that ‘a nudge is any factor that significantly alters the behavior of Humans, even though it would be ignored by Econs’ (see Hansen 2016). The argument for doing this is that (1) the principle is formulated by Thaler and Sunstein as that which the concept of nudge is trying to capture, (2) it expresses what ultimately defines insights from Behavioural Economics if not behavioural science per se, and (3) it results in clear answers to the inconsistencies noted relative to the original definition. It is to this approach we now turn. 3.2  Nudges Work by Exploiting Irrationality A structural similarity pointed out by the reconstructive approach is that the examples of nudges provided by Thaler and Sunstein (2008) influence people’s behaviour by exploiting the automatic psychological mechanisms explored by DPTs. Such mechanisms are noted by Bovens (2008) to give rise to patterns of irrationality, where by irrational is meant that ‘what is driving my action does not constitute a reason for my action – i.e. it is not a

M4716-STRAßHEIM_9781785367847_t.indd 68

25/04/2019 13:32

The concepts of nudge and nudging in behavioural public policy  69 feature of the action that I endorse as a feature that makes the action desirable’ (Bovens 2008, p. 210). That is, nudges seem to work by exploiting cognitive boundaries, bias and habits. Although this does not fit all examples mentioned by Thaler and Sunstein, it does fit the basic principle that the concept of nudge is trying to capture as well as the findings of behavioural sciences. A reason for the theoretical approach adopting this feature as a necessary condition in the definition of ‘nudge’ is provided by its ability to answer some of the questions facing Thaler and Sunstein’s original definition. First, adding an option to the existing choice architecture is excluded as a nudge as it causes people to change behaviour purely on rational grounds. Second, it shows why non-economic incentives like electric shocks, public beatings or funny piano stairs do not count as nudges. These types of interventions constitute reasons for action that people readily endorse as features that make the actions desirable or undesirable. This point is not even neglected by standard economists who readily incorporate non-economic incentives into the utility functions of agents. Third, the introduction of the condition of the exploitation of irrational psychological mechanisms also provides a way to make sense of the concept implicit in Thaler and Sunstein’s definition of ‘insignificant’ changes to economic incentives that may yet influence behaviour ‘significantly’. Some changes in economic incentives studied in behavioural science, for example, charm prices (Thomas and Morwitz 2005), influence behaviour in ways or to an extent that may not be accounted for purely in terms of rationality. Interestingly, these points make for additional insights. In discussing the addition of choices Hansen (2016) thus not only finds that examples exist where adding an option may (e.g. by causing asymmetric dominance effects) or may not (e.g. adding a dominant strategy) qualify as a nudge; but also that examples exist where the removal of choice options qualifies as nudges because these options are rationally irrelevant, that is, their removal does not constitute a rational reason for action, but still affect behaviour change. Hence, contrary to Thaler and Sunstein’s definition one may actually nudge by removing an option, albeit usually in highly artificial circumstances. Also, with regards to the third point Hansen (2016) argues that it shows that a nudge should not be mistaken with the general intervention. Instead it should be regarded as the attempt to influence behaviour through the activation of a psychological mechanism or function by means of contextual features so as to cause patterns of irrational behaviour. This is important since it implies that such contextual features may in fact be bans, mandates, incentives or some other intervention providing rational reasons for behaviour change, yet at the same time also activate psychological mechanisms giving rise to patterns of irrationality. This is, for example, the case if introducing an incentive in terms of a lottery. In this way, a theoretical basis for the often-observed conclusion that nudges may be combined or integrated with traditional public policy is provided – one does not preclude the other. 3.3  Nudges Presuppose Irrationality Another structural similarity of the examples provided by Thaler and Sunstein is that of nudges being applied to existing patterns of irrational behaviour. This was noted by Hausman and Welch (2010) who even stipulate as part of the definition of nudge, irrationality, or as they call it ‘flaws in individual decision-making’, as a precondition for the application of nudge-interventions, which in turn delineates a particular practical field of application. As

M4716-STRAßHEIM_9781785367847_t.indd 69

25/04/2019 13:32

70  Handbook of behavioural change and public policy it happens this is not only a structural similarity of paradigm cases of nudges. It actually follows if behavioural science is taken as the theoretical foundation of nudges, since the basic theoretical principle that the concept is trying to capture points out that in so far as people are acting rationally relative to a particular context, that is, acting as Econs, then there is no reason to apply nudges, as Econs are supposed to ignore nudges by this principle. Thus, as formulated by Hansen (2016), nudges, when intentional, are motivated because of cognitive boundaries, biases and habits in individual and social decision-making posing barriers for people to perform rationally in their own self-declared interests. 3.4 Intentionality Touching upon ‘intentionality’ makes for a third point. Throughout Nudge Thaler and Sunstein define nudges as non-intentional ‘aspects’ or ‘factors’ of choice architecture. In later writings Sunstein repeats this claim, saying e.g. that cafeterias (Sunstein 2016, p. 22) and even ‘nature nudges’ (Sunstein 2016, p. 37). Yet, all interesting examples of nudges are instances of intentional interventions applied to as well as integrating irrational behaviour. So why insist on a non-intentional concept of ‘nudge’ that might as well be covered by the concept of ‘accidental influence’? One reason is that it allows for the unavoidability argument (Hansen and Jespersen 2013). This argument says, that since any behaviour takes place in the context of choice architecture and any choice architecture activates irrational psychological mechanisms that lead to irrational patterns of behaviour, then you are always being influenced by aspects of choice architecture in ways that should not matter to Econs, but does to Humans. In turn, defining nudges as aspects of choice architecture means that you are always being nudged, whether anybody intends this or not – ‘Cafeterias nudge, even if those who design them do not intend any nudging’ (Sunstein 2016, p. 22) – and hence, that you cannot be against nudging per se. While one may agree that you cannot be against nudging per se, ensuing debate has made clear, that the distinction between accidental choice architecture and the intentional design of such is a crucial one. It serves to emphasize the ontological as well as ethically relevant difference between mere accidental attributes of choice architecture and that of intentional interventions (Hausman and Welch 2010; Hansen and Jespersen 2013). Thus, ignoring the intentional aspect of a nudge, not only breaks with the original semantics of the concept (Hansen and Jespersen 2013). It also fails to reflect the normative responsibility that policy-makers and other choice architects assume when making use of nudges. Recognizing this undermines the unavoidability argument so often found attractive by policy-makers when discussing the ethics of nudge. Ethical responsibility cannot be avoided, why – if not for conceptual precision alone – nudges should be defined as intentional attempts at influencing behaviour so as to emphasize this. Given all of the above we may follow Hansen (2016) in defining a nudge thus: A nudge is a function of (1) any attempt at influencing people’s judgments, choices or behaviours in a predictable way that is (2) motivated because of cognitive boundaries, biases, and habits in individual and social decision-making posing barriers for people to perform rationally in their own self-declared interests, and (3) which works by making use of those boundaries, biases and habits as integral parts of such attempts.

M4716-STRAßHEIM_9781785367847_t.indd 70

25/04/2019 13:32

The concepts of nudge and nudging in behavioural public policy  71 3.5 Implications Provided this definition, it also turns out that the original definition of nudge is an implication, not a characterization of nudges (Hansen 2016). That is, the original definition of a nudge provided by Thaler and Sunstein (2008, p. 6) is actually a consequence of this more fundamental definition above. A nudge conceptually denotes the use of behavioural insights from behavioural science to alter the behavior of Humans, in ways that would be ignored by Econs since they do not constitute reasons for actions in the sense that they are endorsed as features that makes an action desirable. This rules out the rational effects of bans, mandates and fiddling with incentives, which implies that a nudge defined as above performs its behaviour changing function without forbidding any options or significantly changing economic incentives. Thus, the original definition provided by Thaler and Sunstein should rather be regarded as a characterization of what is implied by nudges than the definition that it is not. A further implication is that the provision of information and rational persuasion are not nudges as otherwise proposed by Sunstein when writing the ‘Provision of information is certainly a nudge, but it may or may not qualify as paternalistic…’ (Sunstein 2014, p. 55). While the provision of information and rational persuasion may function as a nudge in the sense that it emphasizes certain aspects of a choice architecture at the expense of others, information and rational persuasion, conceived of in isolation, if anything, besides preferences, constitute rational reasons for action. Another implication is that while (A) information and rational persuasion are surely reconcilable with libertarianism (i.e. when liberty preserving, but intended to influence behaviour in people’s own interest), but do not qualify as nudges, and (B) supermarkets and banks may definitely use nudges without any intention of being libertarian paternalistic understood as ‘… an approach that preserves freedom of choice but authorizes both private and public institutions to steer people in directions that will promote their welfare’ (Thaler and Sunstein 2003, p. 179), then (C) the concept of nudge is not equivalent to that of libertarian paternalism as it is often mistakenly portrayed.

4. PUSH, CURLING AND NUDGE – THREE APPROACHES OF BEHAVIOURAL PUBLIC POLICY Besides these clarifications the structure of the theoretical definition also provides the basis for distinguishing between three basic strategies for applying behavioural insights in public policy of which nudges are only one. These strategies are: push, curling and nudge. 4.1 Push Challenged with behavioural problems, policy-makers may decide to ‘push’ in the sense of strengthening the aspects of the choice architecture that provide rational reasons for action beyond what ought to be required from a purely rational approach. The strategy behind ‘push’ is to trump any irrational psychological mechanisms, which would otherwise give rise to the irrational behaviour observed. Thus, ‘push’ satisfies condition (1) and (2) in the definition above, but takes a different road to solving behavioural problems. The strategy is illustrated by policies substantially increasing tax on alcohol, cigarettes

M4716-STRAßHEIM_9781785367847_t.indd 71

25/04/2019 13:32

72  Handbook of behavioural change and public policy and other products of self-harm and providing severe or public stigmatization or punishment by law for even minor offences beyond what ought to be required from a purely rational approach. The policy approach of push thus puts words in continuation of the rational approach to policy-making, but is behaviourally informed if prescribing – from a rational viewpoint – a self-consciously exaggerated use of traditional regulatory approaches based on the assumption that humans act irrational. 4.2 Curling A second behavioural policy approach when faced with a pattern of irrational behaviour is to attempt to weaken, remove and/or counter the activation of the irrational psychological mechanisms that would otherwise give rise to the pattern in question. That is, condition (1) and (2) in the definition of nudge is satisfied, but rather than approaching the problem with nudges, policy-makers opt for trying to de-bias decisions by removing ‘psychological barriers’ and ‘friction’ or counter illicit nudges by, for example, banning choice architectural features. The curling strategy is illustrated by the presentation of risky decisions in frames that people better understand, EU’s ban of pre-ticked boxes on shopping websites to aid consumers (EC 2011), mandatory cool down periods on payday loans and ‘making it easy’ to fill out online forms by minimizing the number of clicks the user has to perform. 4.3 Nudge Finally, the third behavioural policy approach when faced with a pattern of irrational behaviour is to nudge. As explained above, this strategy actively seeks to exploit the same kind of irrational mechanisms that produce irrational behaviour. Thus, a nudge not only satisfies condition (1) and (2), but also (3). The strategy is illustrated by cases such as pensions savings, where the irrational pattern of non-saving is most likely produced by a combination of loss aversion and hyperbolic discounting, that is, when payment means renouncing on current income in order to gain in the far future. Here programmes such as save-more-tomorrow have successfully nudged people to save for their pension by turning the tables on hyperbolic discounting (agree now to save in the future), using money from a pay raise (thereby ‘curling’ loss aversion), and integrating mental accounting (savings will come from your pay raise) as well as sometimes defaulting people into the pensions programme.

5.  WHAT IS NUDGING? At conferences, in seminars and when teaching one will often be asked whether ‘we haven’t always been nudging?’ While it may be tempting to answer ‘yes’ as it leaves the audience feeling more comfortable, it also leads directly to another question: ‘Then, what’s new about it?’ Instead of answering ‘yes’ one may offer the following answer that relies upon making a conceptual distinction between ‘nudges’ and ‘nudging’ (see also Hansen et al. 2016). A nudge is as defined above and, yes, we have always been using such measures to influence behaviour; but nudging is the systematic and evidence-based development and

M4716-STRAßHEIM_9781785367847_t.indd 72

25/04/2019 13:32

The concepts of nudge and nudging in behavioural public policy  73 implementation of nudges in creating behaviour change. Thus, in this sense it is something new, and today it is this effort that is properly referred to as the field of ‘nudging’ and, as a discipline it is increasing in its influence on public policy and behaviour change strategies across the world. A key institution in this development has been the Behavioural Insights Team (BIT) established 2010 by UK Prime Minister David Cameron. Led by Dr. David Halpern and Managing Director Owain Service it rolled out of government to become partly privatized in 2014 (Halpern 2015). However, BIT is part of a broader trend, which has seen nudge units, initiatives, and networks emerging in the US, Denmark, Singapore, France and Canada, to mention just a few of the many exciting places this is happening. Likewise the OECD, the World Bank and the European Commission have published reports, held meetings and actively supported research to further examine the potential of nudging (see Lunn 2014; World Bank 2015; Lourenço et al. 2016; OECD 2017). Together, all of these efforts have led to the emergence of the field. Yet nudging is only in its infancy. Still, it should be noticed that a common scientific framework of reference unites these efforts. In particular, nudging relies heavily on theories and methodologies from behavioural economics, cognitive and social psychology, using microeconomic decision theory as a baseline. Also, a central focus within the field is the biases and heuristics programme of Kahneman and Tversky, which, as we saw, is rooted in dual-process theories of cognition and information processing (Kahneman and Tversky 1979) and made accessible to the wider public by Kahneman’s dual-system theory presented in his famous book Thinking, Fast and Slow (Kahneman 2011). In addition to the shared psychological underpinnings, the widespread efforts falling under the auspices of nudging are unified by the ambition to advance and apply quantitative experimental approaches to field research. The choice (not the invention) of this methodological approach may be ascribed to the intellectual origins in the laboratory experiments of behavioural economics and experimental psychology. Randomized controlled trials are also explicitly formulated as the ideal in, for example, the BIT’s 2012 methodology report, Test, Learn and Adapt (Haynes et al. 2012). However, it is important to stress that the objective of nudging is just as much about evaluating the efficacy and policy implications of nudge interventions and examining the potential real-world feasibility and applicability of behavioural insights as it is about extending the boundaries of scientific knowledge. Hence, the aspirations of nudging as currently carried out have much in common with what is usually referred to as real-world research and the research relationship ideally pursued by stakeholders might best be characterized by this quote from Hall and Hall (1996, p. 12) (albeit in a different context): ‘The research relationship is between equals, and is not exploitative: the client organization is not being “used” merely to develop academic theory or careers nor is the academic community being “used” (brains being picked). There is genuine exchange. The research is negotiated.’

6.  CONCERNS ABOUT NUDGING IN PUBLIC POLICY Despite various versions of this ideal being adopted by practitioners of nudging, the current efforts have prompted speculation and suspicion (see Hansen and Jespersen 2013). One set of worries pertains to the threat of science being utilized by potentially biased

M4716-STRAßHEIM_9781785367847_t.indd 73

25/04/2019 13:32

74  Handbook of behavioural change and public policy policy-makers to manipulate citizens. Another set of worries pertains to whether nudging is being used as an excuse to roll back traditional regulatory efforts. This latter worry has been most prominent in Europeans’ response to nudging; the US response has been worried more about the former critique pointing to the paternalistic aspects of the approach. Fortunately, most of these concerns turn out to rely on superficial readings of the scientific foundations of nudging or on ignoring the challenges that face any attempt at regulating citizens’ behaviour (see Hansen and Jespersen 2013). First, nudges do not only invoke automatic processes, for example, they may invoke such processes to activate consciousness and choice such as when using prompts. In this sense nudges may not only be limited to being liberty-preserving, but also liberty enhancing. Further, automatic processes are not necessarily ‘subconscious’ (by analogy: when a plane is on autopilot it does not imply that the pilot is flying subconsciously or unaware of what is going on). Hence, nudging is not characterized by psychological manipulation, as some critics tend to believe. Still, some nudges do rely on non-transparent measures that threaten with transferring responsibilities to citizens in ways that should be regarded as manipulative and thus as illegitimate strategies in public policy (Hansen and Jespersen 2013). To this end it does not suffice to say that by principle nudges leave all choice-options from the original status quo available post-intervention since we are dealing with a paradigm that by its nature discards theoretical principles in favour of empirical practice. Instead we need to take the ethics of nudging seriously on a case-by-case basis since nudges comprise such a vast array of different measures that they cannot be evaluated as one (see Sunstein 2016). Still, who should provide such evaluation? This leads to a second point with regards to the ethics of nudging. While policy-makers may indeed be biased themselves the ‘who nudges the nudgers’ critique fails for at least two reasons. First, with regards to the goals pursued, while true that nudging is directed by potentially biased politicians, so is all other regulatory effort. Second, with regards to bias in policy-makers’ interpretation of results, while nudges integrate behavioural insights into our choice architecture, nudging rests on approaches that comprise scientific state-of-the-art methods for trying to detect and avoid cognitive biases in interpretation. This latter point also provides an answer to those who fear that applying nudging in public policy is just an excuse to extend or roll back traditional regulatory efforts. Nudging as well as nudges are fully compatible with and hence should be evaluated relative to standard regulatory measures. However, when it comes to the standards used for evaluating traditional regulatory measures aimed at changing behaviour, nudging with its firm rooting in the behavioural sciences may actually turn out to raise the bar quite substantially due to its strict methodologies. Thus, nudging should be expected to change the way we do public policy-making and delivery due to its introduction of scientific requirements by means of its evidence-based standards. Thus, the potential consequences of nudging should not be characterized as public roll back or overreach. Rather it is part of a shift of regulatory paradigm that may properly be referred to as Behavioural Public Policy.

M4716-STRAßHEIM_9781785367847_t.indd 74

25/04/2019 13:32

The concepts of nudge and nudging in behavioural public policy  75

NOTES 1. According to this distinction computational level theories are concerned with identifying relations between inputs (e.g. stimuli) and outputs (e.g. behavior), algorithmic level theories are concerned with the mechanisms that translate inputs into outputs (e.g. DPTs), and implementational level theories are concerned with the physical systems that implement the mechanisms identified at the algorithmic level, e.g. as in neuroscience. 2. However, it should be observed that, as much of behavioural economics tries to study the deviations of output behaviour from the rational baseline without much eye to the mechanics of the brain, it often comes close to being a computational level theory. 3. Originally introduced in their 2003 essay of the same name (Thaler and Sunstein 2003), the concept of Libertarian Paternalism actually precedes that of nudge and was originally defined in a stricter sense. A policy was said to be ‘paternalistic’ ‘if it is selected with the goal of influencing the choices of affected parties in a way that will make those parties better off’, where ‘better off’ was meant to be ‘measured as objectively as possible’ and not always equating revealed preference with welfare (Thaler and Sunstein 2003, p. 175). In Thaler and Sunstein (2008) this was modified so that ‘a policy is “paternalistic” if it tries to influence choices in a way that will make choosers better off, as judged by themselves’ (Thaler and Sunstein 2008, p. 5).

REFERENCES Adams, T. (2014), ‘Nudge Economics: Has Push Come To Shove For a Fashionable Theory?’, The Guardian, accessed 14 February 2017 at http://www..com/science/2014/jun/01/nudge-economics-freakonomics-danielkahne​man-debunked. Ariely, D. (2010), Predictably Irrational, Revised and Expanded Edition: The Hidden Forces That Shape Our Decisions, New York: Harper Perennial. Bhargava, S. and G. Loewenstein (2015), ‘Behavioral Economics and Public Policy 102: Beyond Nudging’, American Economic Review: Papers and Proceedings, 105(5), 396–401. Bovens, L. (2008), ‘The Ethics of Nudge’, in T. Grüne-Yanoff and S.O. Hansson (eds), Preference Change: Approaches from Philosophy, Economics and Psychology, Berlin: Springer, pp. 207‒219. Burgess, A. (2012), ‘“Nudging” Healthy Lifestyles: The UK Experiments with the Behavioural Alternative to Regulation and the Market’, European Journal of Risk Regulation, 3(1), 3–16. Dennett, D. (1987), The Intentional Stance, Cambridge MA: MIT Press. Dimand, M. and R.W. Dimand (1996), The History Of Game Theory, Volume 1: From the Beginnings to 1945, Routledge Studies in the History of Economics, Abingdon, Oxfordshire: Routledge. Dunt, I. (2014), ‘Nudge Nudge, Say No More. Brits’ Minds Will Be Controlled Without Us Knowing It’, The Guardian, accessed 17 March 2017 at https://www.theguardian.com/commentisfree/2014/feb/05/ nudge-say-no-more-behavioural-insights-team. EC (2011), New EU rules on consumer rights to enter into force, accessed 23 February 2017 at http://europa.eu/ rapid/press-release_MEMO-11-675_en.htm. Evans, J.S.B. (2008), ‘Dual-Processing Accounts of Reasoning, Judgment, and Social Cognition’, Annual Review of Psychology, 59, 255–278. Friedman, M. (1966), Essays in Positive Economics, Chicago: University of Chicago Press. Gawronski, B., J.W. Sherman and Y. Trope (2014), ‘Two of What?: A Conceptual Analysis of Dual-Process Theories’, in J.W. Sherman, B. Gawronski and Y. Trope (eds), Dual-Process Theories of the Social Mind, New York: Guilford, pp. 3–19. Gigerenzer, G. (1991), ‘How to Make Cognitive Illusions Disappear: Beyond “Heuristics and Biases”’, European Review of Social Psychology, 2(1), 83‒115. Gigerenzer, G., P. Todd and the ABC Research Group (1999), Simple Heuristics That Make Us Smart, Oxford: Oxford University Press. Gilovich, T., D. Griffin and D. Kahneman (eds) (2002), Heuristics and Biases: The Psychology of Intuitive Judgment, Cambridge: Cambridge University Press. Hall, D. and I. Hall (1996), Practical Social Research: Project Work in the Community, London: Macmillan. Halpern, D. (2015), Inside the Nudge Unit: How Small Changes Can Make a Big Difference, London: WH Allen. Hansen, P.G. (2016), ‘The Definition of Nudge and Libertarian Paternalism’, European Journal of Risk Regulation, 7(1), 155–174. Hansen, P.G. and A.M. Jespersen (2013), ‘Nudge and the Manipulation of Choice’, European Journal of Risk Regulation, 4(1), 3‒28.

M4716-STRAßHEIM_9781785367847_t.indd 75

25/04/2019 13:32

76  Handbook of behavioural change and public policy Hansen, P.G., K.L. Skov and L.R. Skov (2016), ‘Making Healthy Choices Easier: Regulation vs. Nudging’, The Annual Review of Public Health, 37, 237‒251. Hausman, D. and B. Welch (2010), ’Debate: To Nudge or Not to Nudge’, Journal of Political Philosophy, 18(1), 123‒136. Haynes, L., B. Goldacre and D. Torgerson (2012), Test, Learn, Adapt: Developing Public Policy with Randomised Controlled Trials, London: Cabinet Office – Behavioural Insights Team. Hewstone, M., W. Stroebe and K. Jonas (eds) (2012), An Introduction to Social Psychology (5th edn), West Sussex: John Wiley & Sons. Horton, T. (2009), ‘The Darker Side of Nudging’, The Political Quarterly, 80(2), 295–298. House of Commons (2011), House of Commons, Public Health: Twelfth Report of Session 2010-12 Vol. 1: Report. Together with Formal Minutes, London: Parliament: House of Commons: Health Committee. Kahneman, D. (2011), Thinking, Fast and Slow, New York: Farrar, Straus & Giroux. Kahneman, D. and A. Tversky (1979), ‘Prospect Theory: An Analysis of Decision under Risk’, Econometrica, 47(2), 263‒292. Kahneman, D. and A. Tversky (eds) (2000), Choice, Values, and Frames, New York: Cambridge University Press. Leggett, W. (2014), ‘The Politics of Behaviour Change: Nudge, Neoliberalism and the State’, Policy & Politics, 42(1), 3–19. Lourenço, J.S., E. Ciriolo, S.R. Almeida and X. Troussard (2016), Behavioural Insights Applied to Policy: European Report 2016, accessed 15 January 2019 at http://publications.jrc.ec.europa.eu/repository/bitstream/ JRC100146/kjna27726enn_new.pdf. Lunn, P. (2014), Regulatory Policy and Behavioural Economics, Paris: OECD Publishing. Marr, D. (1982), Vision: A Computational Investigation into the Human Representation and Processing of Visual Information, San Francisco: Freeman. Mongin, P. and M. Cozic (2014), ‘Rethinking Nudges’, HEC Paris Research Paper No. ECO/SCD-2014-1067. Mongin, P. and M. Cozic (2018), ‘Rethinking Nudge: Not One But Three Concepts’, Behavioural Public Policy, 2(1), 107‒124. Mullainathan, S. and E. Shafir (2013), Scarcity: Why Having Too Little Means So Much, New York: Times Books. Neal, D., W. Wood, J. Labrecque and P. Lally (2011), ‘How do Habits Guide Behavior? Perceived and Actual Triggers of Habits in Daily Life’, Journal of Experimental Social Psychology, 48, 492‒498. OECD (2017), Behavioural Insights and Public Policy: Lessons from Around the World, Paris: OECD Publishing. Premack, D. and G. Woodruff (1978), ‘Does the Chimpanzee have a Theory of Mind?’, Behavioral and Brain Sciences, 1(4), 515‒629. Rizzo, M.J. and D.G. Whitman (2009), ‘Little Brother Is Watching You: New Paternalism on the Slippery Slopes’, Arizona Law Review, 2009, 51(3), 685‒739. Shafir, E. (ed.) (2012), The Behavioral Foundations of Public Policy, Princeton: Princeton University Press. Simon, H.A. (1955), ‘A Behavioral Model of Rational Choice’, The Quarterly Journal of Economics, 69(1), 99‒118. Smith, A.C. and T.J. Zywicki (2014), ‘Behavior, Paternalism, and Policy: Evaluating Consumer Financial Protection’, George Mason Law & Economics Research Paper No. 14-05. Stanovich, K.E. (2012), ‘On the Distinction between Rationality and Intelligence: Implications for Understanding Individual Differences in Reasoning’ in K. Holyoak and R. Morrison (eds), The Oxford Handbook of Thinking and Reasoning, New York: Oxford University Press, pp. 343‒365. Sunstein, C.R. (2014), Why Nudge?: The Politics of Libertarian Paternalism, New Haven: Yale University Press. Sunstein, C.R. (2016), The Ethics of Influence: Government in the Age of Behavioral Science, New York: Cambridge University Press. Tetlock, P.E. and B.A. Mellers (2002), ‘The Great Rationality Debate’, Psychological Science, 13(1), 94–99. Thaler, R.H. (1985), ‘Mental Accounting and Consumer Choice’, Marketing Science, 4(3), 199‒214. Thaler, R.H. (1992), The Winner’s Curse: Paradoxes and Anomalies of Economic Life, Princeton: Princeton University Press. Thaler, R.H. (2016), Misbehaving: The Making of Behavioral Economics, New York: W.W. Norton & Company. Thaler, R.H. and C.R. Sunstein (2003), ‘Libertarian Paternalism’, American Economic Review, 93(2), 175‒179. Thaler, R.H. and C.R. Sunstein (2008), Nudge: Improving Decisions about Health, Wealth, and Happiness, New Haven: Yale University Press. Thaler, R.H. and C.R. Sunstein (2009), Nudge: Improving Decisions about Health, Wealth, and Happiness, Revised and Expanded Edition, New York: Penguin Books. Thaler, R.H., C.R. Sunstein and J.P. Balz (2010), ‘Choice Architecture’ in E. Shafir (ed.), The Behavioral Foundations of Public Policy, Princeton: Princeton University Press, Chapter 25, pp. 428‒429. Thomas, M. and V. Morwitz (2005), ‘Penny Wise and Pound Foolish: The Left-Digit Effect in Price Cognition’, Journal of Consumer Research, 32(1), 54‒64.

M4716-STRAßHEIM_9781785367847_t.indd 76

25/04/2019 13:32

The concepts of nudge and nudging in behavioural public policy  77 Tierny, J. (2014), ‘A Nudge (or Is it a Shove?) To the Unwise’, New York Times, accessed 14 February 2017 at http://tierneylab.blogs.nytimes.com/2008/03/24/a-nudge-or-is-it-a-shove-to-the-unwise/. Tversky, A. and D. Kahneman (1974), ‘Judgment Under Uncertainty: Heuristics and Biases’, Science, 185(4157), 1124–1131. Tversky, A. and D. Kahneman (1981), ‘The Framing of Decisions and the Psychology of Choice’, Science, 211(4481), 453‒458. World Bank (2015), World Development Report 2015: Mind, Society, and Behavior, Washington, DC: World Bank.

M4716-STRAßHEIM_9781785367847_t.indd 77

25/04/2019 13:32

6.  Behavioural considerations in public policy: matching policy tools and their targets Michael Howlett

1. INTRODUCTION: THE UTILITARIAN ROOTS OF MUCH POLICY DESIGN THINKING Policy tools have targets (Weaver 2009a, 2009b, 2010). That is, not just ‘targets’ in the sense of policy aims and goals and their measures but ‘targets’ in the sense of individuals and groups whose behaviour is expected to be affected by policy activity.1 Tools and targets are linked in the sense that policy tool use involves implicit or explicit assumptions and expectations about the effect tool deployment has upon those impacted by it. It is critically important for policy-making that the behaviour resulting from tool use in practice actually matches that anticipated in the pre-deployment or formulation period (May 2004; Kaine et al. 2010; Duesberg et al. 2014). Regardless of whether those targets are purely social constructions with few empirical referents (Schneider and Ingram 1993, 2005) or if they reflect more objective assessment of the actual behaviour of relevant groups of policy actors, it is critical for efficient and effective policy-making that tools match targets and vice versa. There is thus a significant behavioural component to policy design, tool use and choice which is critical to policy success and failure (Weaver 2009b; Lynn 1986; Schneider and Ingram 1990; Shafir 2013). However, despite the fact that ‘compliance’ with government intentions has been a significant issue in areas such as regulatory studies for many years (Feeley 1970; Etienne 2011; Meier and Morgan 1982; Rodgers 1975; Mulford and Etzioni 1978), this aspect of policy design has only rarely been systematically examined (Grabosky 1995; Weaver 2009a, 2009b, 2014, 2015; Winter and May 2001; Parker and Nielsen 2012). Moreover, it has only rarely been related to the effective use of particular kinds of policy instruments (Duesberg et al. 2014; Corner and Randall 2011) and to decisions to use one, or more, of some particular type rather than another (Taylor et al. 2013). 1.1  Rethinking Policy Tool Use: Compliance and Its Vicissitudes Weaver (2009b, p. 5) has recently enumerated some of the various ‘compliance problems’ or ‘barriers’ to compliance which targets and governments face. These include: ●● ●● ●● ●●

Incentive and sanction problems where positive and/or negative incentives are insufficient to ensure compliance. Monitoring problems where target compliance may be difficult or costly to monitor. Resource problems where targets lack the resources to comply even if they want to. Autonomy problems where targets do not have the power to make decisions that comply with policy even if they want to. 78

M4716-STRAßHEIM_9781785367847_t.indd 78

25/04/2019 13:32

Behavioural considerations in public policy  79 ●● ●● ●●

Information problems where targets lack information that would make compliance more likely. Attitude and objectives problems where targets are hostile /mistrustful towards providers or programmes. Herding effects, where targets follow their peers without giving due consideration to alternative actions.

These problems should be the subject of investigation and action as what is needed are effective policy designs that overcome these problems. This requires more systematic analysis and understanding of the motivations of policy targets which would allow better matching of tools and targets right at the outset of policy-making (Weaver 2009a, 2009b; Braithwaite 2003; Schneider and Sidney 2009; Chatterton and Wilson 2014; Pierce et al. 2014).

2. BEYOND COMPLIANCE-DETERRENCE LOGICS: CONTEMPORARY CHALLENGES TO THE UTILITARIAN FOUNDATIONS OF PAST POLICY DESIGN THINKING The immediate aim of most public policy is to invoke in the ‘targets’ of government efforts the behavioural change needed for them to comply with government aims. This is done in order to secure better adherence of populations to government aims and ambitions be it in the promotion of public safety and security or in the provision of effective healthcare and social welfare. Compliant target behaviour is expected to be achieved through deployment of governing resources in the form of specific combinations of substantive and procedural policy tools aimed at specific kinds of behaviour (Anderson 1977; Baldwin 1985). Desired changes can be large or small, and the expectation of compliance can be rapid or gradual. But in all cases some changes in behaviour in a direction congruent with government aims is expected from the utilization of state resources. If perfect compliance of targets with government aims existed automatically, of course, there would be little purpose in undertaking state activity beyond providing citizens with information about government goals and their expected behaviour, which would then simply occur. Why such compliance is not always forthcoming is thus a key question in the policy sciences, related to better understanding the conditions of policy success and failure and the kinds of designs and activities more likely to attain success with minimal effort and expenditure (Feeley 1970; Mulford and Etzioni 1978). The subject of instrument deployment and target behaviour, however, is one which has often been examined only in a very cursory fashion and often under the burden of many, mainly economistic, assumptions about the motivations and behaviour of policy targets (Stover and Brown 1975). This way of thinking about public policy compliance and policy implementation is generally congruent with a utilitarian orientation towards policy-making in general, one which has been generally pervasive in the policy sciences from the very founding of the discipline (Tribe 1972; Banfield 1977). This orientation, which assumes individuals act as rational utility maximizers, continues to dominate even more nuanced recent thinking about tool use. This is the case, for example, with notions of policy ‘nudging’ which have led to the

M4716-STRAßHEIM_9781785367847_t.indd 79

25/04/2019 13:32

80  Handbook of behavioural change and public policy questioning of the relevance of many traditional policy-related utilitarian concepts such as perfect information and reciprocal risk and benefit valuations in policy analysis and design activities (Oliver 2015; Leggett 2014; Room 2013; John et al. 2009). Legitimate taxes and rules, however, are still often simply expected to be paid and obeyed by the majority of citizens and targets, with penalties and fines established to punish non-compliance in such a way as to ‘deter’ a minority of non-compliers and the spread of unwanted and illegitimate behaviour to the majority (Doern and Phidd 1983). Utility calculations are extended to the calibration of penalties, with these set at such a level as to punish those who might contest the legitimacy of such actions or seek to free-ride on compliers (Lowi 1966; Balch 1980). In practice, however, the compliance of policy targets with government intentions is, as Weaver noted, recognized as a more complex problem (Meier and Morgan 1982; Barile et al. 2015). In most circumstances governments must determine whether or not a target is likely to comply with government actions and intentions and whether any compliance is reluctant or freely given (Scholz 1991). That is, even the most basic activities of governing such as collecting taxes and ensuring laws and rules are obeyed involves not just individual hedonic – or pleasure and pain seeking – behaviour but also considerations on the part of targets and the public of issues such as the legality and normative ‘appropriateness’ of government’s levying and collecting such taxes or passing and enforcing such rules (March and Olsen 1989). Moreover, different kinds of target groups and individuals exist or are perceived to exist in terms of government expectations of the nature of their compliant or non-compliant behaviour and can be and are treated differently by governments according to those assumptions. And, of course, governments have more tools at their disposal than just authority-based ones and thus have a range of options available both to implement policy and promote compliance, including those linked to education and persuasion of targets in addition to the exercise of coercion (Hawkins and Thomas 1989; Hood 1986). 2.1  Re-conceptualizing Target Behaviour and Policy Design Imperatives This variation in target structure, motivation and compliance behaviour makes policy design a more challenging activity than a simple hedonic utilitarian perspective would have it. It is an even more complex situation than the behavioural economics which lie behind ‘nudges’ – that is, actions such as changing defaults or using visual clues to promote action – would have it as even behaviourally-inspired ‘rules’ of semi-rational economic calculations are not enough to capture all the considerations of cultural and psychological appropriateness cited above (Knetsch 2011; Koh 2011). Understanding whether a proposed action is likely to trigger behaviour linked to ‘affiliation’ or ‘conformity’ with government wishes or results in various kinds of non-compliance from outright disobedience to ‘boomerang’ effects encouraging the action they are aimed at discouraging, or vice versa, thus remains a critical (Cialdini and Goldstein 2004; Cialdini et al. 2006) but not well understood aspect of policy designs and designing. Nevertheless, despite its limitations, consideration and plans for ‘nudging’ and other aspects of the application of the findings of behavioural economics and behavioural psychology to policy-making in recent years have served to undermine the utilitarian paradigm in the field and bring a new focus to contemporary policy behaviour (Thaler and Sunstein 2009; Ariely 2010). The same is true of the recent employment of policy tools

M4716-STRAßHEIM_9781785367847_t.indd 80

25/04/2019 13:32

Behavioural considerations in public policy  81 such as co-production or faith-based public service delivery (Alford 1998; Hula et al. 2007; Kissane 2007; Zehavi 2008) and, to a lesser extent, ‘social marketing’ efforts (Pykett et al. 2014), whose successes in areas such as education and healthcare have also undermined confidence in the ability of utilitarian models to capture critical aspects of target behaviour responsible for participation in, and compliance with, government schemes and intentions. This is the case, for example, with efforts to use information-based tools or moral suasion to try to convince citizens to do their duty and refrain from, for instance, littering (Grasmick et al. 1991; John 2013a); to undertake healthy behaviours like walking and exercising or to ‘do the right thing’, in giving up their seats on public transportation to pregnant women, the disabled, the elderly and others less fortunate than themselves; or even to just pay their taxes on time (Stanbury and Fulton 1984; Bardach 1989; Torgler 2004; Corner and Randall 2011).

3. BETTER UNDERSTANDING COMPLIANCE: LINKING POLICY TARGET BEHAVIOUR AND RESOURCE-BASED TAXONOMIES OF POLICY INSTRUMENTS If target behaviour is not utilitarian, though, then what is it and how can it best be anticipated and linked through policy designs to the efficient and effective attainment of government goals? In his pathbreaking early works on public policy-making, for example, Harold Lasswell (1954), a political psychologist by training, conceived of the main instruments of politics as involving, among other things, the manipulation of symbols, signs and icons which rely on individual’s affections and loyalties to particular ideas and actors in addition to financial and other kinds of incentives and disincentives subject to more utilitarian calculations. Lasswell noted the extent to which governments could affect every aspect of policy-making through such manipulations and argued that a principal task of the policy sciences must be to understand the nuances of these actions and their effects (Lasswell 1954, 1971; Doern and Phidd 1983; Doern and Wilson 1974). By the 1980s, however, Lasswell’s admonition to include a wide-ranging set of factors in the examination of underlying policy tool compliance and effectiveness began to be replaced by a single focus on ‘rational’ or utilitarian rationales for policy choices of all kinds. Many studies of policy instruments, heavily influenced by economists, for example, assumed both decision-makers and policy targets were motivated exclusively by relatively narrow utilitarian self-interest maximization (Stokey and Zeckhauser 1978; Trebilcock and Hartle 1982; Dewees 1983). Other studies often reflected this view in part because they followed the lead of economists in focussing on the use of economic tools such as regulation, public enterprises, or subsidies which more or less directly affected the type, quantity, price or other characteristic of goods and services being produced in industrial and environmental policy spheres, which could in fact be analysed in largely economistic terms (Salamon 1989; Bemelmans-Videc 1998; Peters and van Nispen 1998). Policy designs in new fields such as environmental policy-making, for example, reflected this economistic orientation with policy initiatives in areas from pollution prevention to professional regulation assuming a distinctly utilitarian bent in so doing (Hippes 1988; Trebilcock 1983). An important development during this era, however, was related to the rise of behavioural economics, for example, with the expectation that the insights generated by

M4716-STRAßHEIM_9781785367847_t.indd 81

25/04/2019 13:32

82  Handbook of behavioural change and public policy empirical behavioural research would lead to better outcomes than would reliance on a priori assumptions: that is, to more compliance and behaviour change congruent with government wishes (Selinger and Whyte 2012; Liu et al. 2014). This approach promoted, amongst others, the design and adoption of tools linked to modest behavioural modification through the provision of subliminal and other types of cues and decision frames or ‘choice architectures’ and behavioural ‘nudges’ (Thaler et al. 2010; Liu et al. 2014; Lehner et al. 2016). Around the same time other governments similarly devoted much time and energy to ‘social marketing’ or the use of enhanced appeals to collective identities and social mores (such as altruistic or non-altruistic corporate social responsibility (CSR) voluntary codes) with the same expectation that increased and improved compliance could be achieved at modest or less cost using different instruments from, for example, traditional command and control regulation based on hedonic logics (Perri 6 et al. 2010; John 2013b; Dolan et al. 2012; Tallontire 2007; Steurer 2009; Campbell 2012). Both these efforts, and others, cast doubt on orthodox views of compliance and the motivations of target behaviour. Poor experiences with even these alternative tools, however, has also continued to prompt a re-thinking of the relationship between policy tools and behaviour and the need for better and more evidence-informed design (Howlett and Lejano 2013; Moseley and Stoker 2013).

4. COMPLIANCE REVISITED: DEMAND AND SUPPLY CONSIDERATIONS IN EFFECTIVE POLICY TOOL USE All these efforts begin with the observation that some congruence between tool characteristics and target behaviour must exist in order for there to be any impact at all from the deployment of governing tools. Taxonomies of policy tools generated in earlier eras (Tupper and Doern 1981; Hood 1986; Vedung 1997; Howlett 1991) help shed light on this relationship by clarifying the nature of the governing or ‘statecraft’ resources employed by different types of tools (Hood 1991, 1995; Dunleavy and Hood 1994; Riker 1983, 1986; Dunsire 1993; Salamon 2002). Hood (1986), following Anderson (1977), for example, grouped tools into a small number of categories according to whether they rely upon the use of ‘nodality’ (or information), authority, treasure or the organizational resources of government for their effectiveness (see Table 6.1). The overall range of policy tools available to governments in this view includes both the ‘substantive’ ones traditionally examined by economists as well as a range of less economically-oriented ‘procedural’ ones (Howlett 2000) which can be used to affect interest groups and other actor behaviour. This occurs, for example, in the use of information-based procedural instruments which can both facilitate the provision of information as well as suppress it, and can involve the release of misleading as well as accurate information (Mueller 1973; Saward 1992; Howlett 2019). Hood’s idea was that each basic category of tool relied upon a particular different kind of governing resource for its effects and that one of the main reasons one tool would be chosen over another was supply-oriented: that is, that governments would usually prefer to utilize tools deploying resources it had in ample supply or which could be easily replenished (Hood 1983). This is an important insight. However, ‘demand-side’ considerations are also very

M4716-STRAßHEIM_9781785367847_t.indd 82

25/04/2019 13:32

Behavioural considerations in public policy  83 Table 6.1  A resource-based taxonomy of procedural and substantive policy instruments Governing Resource and Target Need Purpose of Tool

Substantive Procedural

Information

Authority

Treasure

Organization

Public Information Campaign Official Secrets Acts

Independent Regulatory Agencies Administrative Advisory Committees

Subsidies and Grants

Public Enterprises

Interest Group funding

Government Re-organizations

Note:  Cells provide examples of instruments in each category. Source:  Adapted from Howlett (2000).

significant here. That is, in general, each category of tool involves the use of a specific governing resource which is expected to trigger or lever a specific behavioural characteristic or target receptor. Thus the effectiveness of the deployment of government resources is linked both to resource availability – a precondition of their use – and to the existence of different proclivities on the part of policy targets which make them receptive to the use of this resource when deployed. The tool in this case is an ‘activator’, or ‘catalyst’ which utilizes governing resources to trigger mechanisms in target groups which change their behaviour (Capano et al. 2019). In the case of information use, for example, this tool’s effectiveness relies both on the availability of knowledge and the means to distribute it and also on the target’s belief in the accuracy of the messages being purveyed, or their credibility, including the target’s desire to believe a message being delivered. Credibility, in this sense, is a mechanism which is activated by the tool and serves as a prerequisite to behavioural change linked to the message broadcast, and received. Similarly, the effectiveness of the use of authoritative tools, as discussed above, depends on target perceptions of government legitimacy; and the effective use of treasure resources on target group financial need and receptivity to government funding or their cupidity. And the effective use of organizational tools relies upon target groups’ trust government competence and the ability of government personnel to provide services and rules. Table 6.2 presents a list of the behavioural pre-requisites which governing tools rely upon for the effect.

5. BETTER LINKING TOOLS AND TARGETS: DEALING WITH MULTIPLE TARGETS AND SOCIALLY CONSTRUCTED TARGET GROUPS The fundamental design problem for governments then, is not just determining a given governmental resource endowment and calculating the range of prison sentences or the amount of fines and subsidies to levy in some situation based on a utilitarian compliancedeterrence logic, but rather to understand on which basis compliance is likely to occur or not by whom. That is, to understand to what extent a target group trusts, believes, wants and needs government actions and to choose instruments which match those mechanisms.

M4716-STRAßHEIM_9781785367847_t.indd 83

25/04/2019 13:32

84  Handbook of behavioural change and public policy Table 6.2  Behavioural needs for resource effectiveness Tool Type

Statecraft Resource Applied

Target Behavioural Pre-Requisite

Nodality

Information

Authority

Coercive Power/Force

Treasure

Financial

Organization

Organization

Degree of Credibility/Gullibility – willingness to believe and act on information provided by government Degree of Legitimacy – willingness to be manipulated by government invoked penalties and proscriptions Degree of Cupidity/Need – willingness to be manipulated by gain/losses imposed by governments Degree of Trust in Government Competence/ Capability – willingness to receive goods and services from government and enter into partnership arrangements

Source:  Howlett (2011).

This is a design challenge which requires detailed empirical investigation and analysis in each case of tool deployment, and continued monitoring over time to ensure these fundamental conditions have not changed or been undermined by any action undertaken. Governments enjoying a high level of trust, for example, may be able to undertake actions through moral suasion while governments which do not enjoy that credibility will need to employ other tools. But whether or not this high level of trust is being maintained is a key determinant of policy effectiveness and continual monitoring and assessment is required to ensure this remains the case and that existing tools continue to function effectively over time.2

NOTES 1. Note this is different from the ‘policy targets’ invoked by Boswell (2014) and others which relate to measures and indicators of policy goals or objectives rather than their intended audience or ‘public’. See also Ghosh et al. (2016) for a similar use of the term. 2. Two aspects of policy-making add complications to this situation and require additional measures to deal with them. These are the fact that many situations involve multiple tools or policy mixes (Howlett 2014; Howlett and del Rio 2015) and that images of policy targets are often socially constructed in such a way that they contain prejudices and biases about individual and group behaviour which interfere with their true representation (Schneider and Ingram 1990; 1997). Space limitations prevent a detailed examination of either issue here.

REFERENCES Alford, John. “A Public Management Road Less Travelled: Clients as Co-producers of Public Services.” Australian Journal of Public Administration, 57, no. 4 (1 December 1998): 128–137. doi:10.1111/j.1467-8500.1998. tb01568.x.

M4716-STRAßHEIM_9781785367847_t.indd 84

25/04/2019 13:32

Behavioural considerations in public policy  85 Anderson, Charles W. Statecraft: An Introduction to Political Choice and Judgment. New York: John Wiley and Sons, 1977. Ariely, Dan. Predictably Irrational, Revised and Expanded Edition: The Hidden Forces That Shape Our Decisions. New York: Harper Perennial, 2010. Balch, G.I. “The Stick, the Carrot, and Other Strategies: A Theoretical Analysis of Governmental Intervention.” Law and Policy Quarterly 2, no. 1 (1980): 35–60. Baldwin, D.A. Economic Statecraft. Princeton, NJ: Princeton University Press, 1985. Banfield, Edward C. “Policy Science as Metaphysical Madness” in Statesmanship and Bureaucracy, edited by Robert A. Goldwin, 1–35. Washington DC: American Enterprise Institute for Public Policy, 1977. Bardach, E. “Moral Suasion and Taxpayer Compliance.” Law Policy 11, no. 1 (1989): 49–69. Barile, Lory, John Cullis and Philip Jones. “Will One Size Fit All? Incentives Designed to Nurture Prosocial Behaviour.” Journal of Behavioral and Experimental Economics 57 (1 August 2015): 9–16. https://doi. org/10.1016/j.socec.2015.04.004. Bemelmans-Videc, Marie-Louise, Ray C. Rist and Evert Vedung, eds, Carrots, Sticks and Sermons: Policy Instruments and Their Evaluation. New Brunswick: Transaction Publishers, 1998. Boswell, Christina. “The Double Life of Targets in Public Policy: Disciplining and Signaling in UK Asylum Policy.” Public Administration, 1 October 2014, n.p. doi:10.1111/padm.12134. Braithwaite, V.A., ed. Taxing Democracy: Understanding Tax Avoidance and Evasion. Aldershot, UK and Burlington, VT, USA: Ashgate Publishing, 2003. Campbell, Bonnie. “Corporate Social Responsibility and Development in Africa: Redefining the Roles and Responsibilities of Public and Private Actors in the Mining Sector.” Resources Policy, 37, no. 2 (June 2012): 138–143. doi:10.1016/j.resourpol.2011.05.002. Capano, Giliberto, Michael Howlett, M. Ramesh and Altaf Virani, eds. Making Policies Work: First- and Second-Order Mechanisms in Policy Design, Cheltenham: Edward Elgar Publishing, 2019. https://www.­eelgar.com/shop/making-policies-work. Chatterton, Tim and Charlie Wilson. “The ‘Four Dimensions of Behaviour’ Framework: A Tool for Characterizing Behaviours to Help Design Better Interventions.” Transportation Planning and Technology, 37, no. 1 (2 January 2014): 38–61. doi:10.1080/03081060.2013.850257. Cialdini, Robert B. and Noah J. Goldstein. “Social Influence: Compliance and Conformity.” Annual Review of Psychology, 55, no. 1 (2004): 591–621. doi:10.1146/annurev.psych.55.090902.142015. Cialdini, Robert B., Linda J. Demaine, Brad J. Sagarin, Daniel W. Barrett, Kelton Rhoads and Patricia L. Winter. “Managing Social Norms for Persuasive Impact.” Social Influence, 1, no. 1 (1 March 2006): 3–15. doi:10.1080/15534510500181459. Corner, Adam and Alex Randall. “Selling Climate Change? The Limitations of Social Marketing as a Strategy for Climate Change Public Engagement.” Global Environmental Change, 21, no. 3 (August 2011): 1005–1014. doi:16/j.gloenvcha.2011.05.002. Dewees, D.N. “Instrument Choice in Environmental Policy.” Economic Inquiry, 31 (1983): 53–71. Doern, G.B. and Richard W. Phidd, Canadian Public Policy: Ideas, Structure, Process, Toronto: Nelson, 1983. Doern, G.B. and V.S. Wilson, “Conclusions and Observations.” Issues in Canadian Public Policy, 337–45. Toronto: Macmillan, 1974. Dolan, P., M. Hallsworth, D. Halpern, D. King, R. Metcalfe and I. Vlaev. “Influencing Behaviour: The Mindspace Way.” Journal of Economic Psychology, 33, no. 1 (February 2012): 264–277. doi:10.1016/j. joep.2011.10.009. Duesberg, Stefanie, Áine Ní Dhubháin and Deirdre O’Connor. “Assessing Policy Tools for Encouraging Farm Afforestation in Ireland.” Land Use Policy, 38 (May 2014): 194–203. doi:10.1016/j.landusepol.2013.11.001. Dunleavy, Patrick and Christopher Hood, “From Old Public Administration to New Public Management.” Public Money and Management, 14, no. 3 (1994): 9–16. Dunsire, A. “Modes of Governance” in Modern Governance, edited by J. Kooiman, 21–34. London: Sage, 1993. Etienne, Julien. “Compliance Theory: A Goal Framing Approach.” Law & Policy, 33, no. 3 (1 July 2011): 305–333. doi:10.1111/j.1467-9930.2011.00340.x. Feeley, Malcolm. “Coercion and Compliance: A New Look at an Old Problem.” Law & Society Review, 4, no. 4 (1 May 1970): 505–519. doi:10.2307/3052818. Ghosh, Atish R., Jonathan D. Ostry and Marcos Chamon. “Two Targets, Two Instruments: Monetary and Exchange Rate Policies in Emerging Market Economies.” Journal of International Money and Finance 60 (1 February 2016): 172–196. doi:10.1016/j.jimonfin.2015.03.005. Grabosky, P.N. “Using Non-Governmental Resources to Foster Regulatory Compliance.” Governance, 8, no. 4 (1995): 527–550. Grasmick, Harold G., Robert J. Bursik and Karyl A. Kinsey. “Shame and Embarrassment as Deterrents to Noncompliance with the Law the Case of an Antilittering Campaign.” Environment and Behavior, 23, no. 2 (1 March 1991): 233–251. doi:10.1177/0013916591232006.

M4716-STRAßHEIM_9781785367847_t.indd 85

25/04/2019 13:32

86  Handbook of behavioural change and public policy Hawkins, K. and J.M. Thomas. “Making Policy in Regulatory Bureaucracies” in Making Regulatory Policy, edited by K. Hawkins and J.M. Thomas, 3–30. Pittsburgh: University of Pittsburgh Press, 1989. Hippes, Gjaltes “New instruments for Environmental Policy: A Perspective.” International Journal of Social Economics, 15, no. 3/4 (1988): 42–51. Hood, Christopher. “Using Bureaucracy Sparingly.” Public Administration, 61, no. 2 (1983): 197–208. Hood, Christopher. The Tools of Government, Chatham: Chatham House, 1986. Hood, Christopher. “A Public Management for All Seasons?” Public Administration, 69, no. Spring (1991): 3–19. Hood Christopher. “Contemporary Public Management: A New Global Paradigm?” Public Policy and Administration, 10, no. 2 (1995): 104–117. Howlett, Michael “Policy Instruments, Policy Styles, and Policy Implementation: National Approaches to Theories of Instrument Choice.” Policy Studies Journal, 19, no. 2 (1991): 1–21. Howlett, Michael. “Managing the ‘Hollow State’: Procedural Policy Instruments and Modern Governance.” Canadian Public Administration, 43, no. 4 (2000): 412–431. Howlett, Michael. Designing Public Policies: Principles and Instruments. New York: Routledge, 2011. http:// www.routledge.com/books/details/9780415781336/. Howlett, Michael. Designing Public Policies: Principles and Instruments. New York: Routledge, 2019. Second edition. Howlett, Michael. “From the ‘Old’ to the ‘New’ Policy Design: Design Thinking beyond Markets and Collaborative Governance.” Policy Sciences, 47, no. 3 (28 May 2014): 187–207. doi:10.1007/s11077-014-9199-0. Howlett, Michael and Pablo del Rio. “The Parameters of Policy Portfolios: Verticality and Horizontality in Design Spaces and Their Consequences for Policy Mix Formulation.” Environment and Planning C, 33, no. 5 (2015): 1233–1245. Howlett, Michael and Raul Lejano. “Tales from the Crypt: The Rise and Fall (and Re-Birth?) of Policy Design Studies.” Administration & Society, 45, no. 3 (2013): 356–380. Hula, Richard, Cynthia Jackson-Elmoore and Laura Reese. “Mixing God’s Work and the Public Business: A Framework for the Analysis of Faith-Based Service Delivery.” Review of Policy Research, 24, no. 1 (2007): 67–89. John, Peter “Experimentation, Behaviour Change and Public Policy.” The Political Quarterly 84, no. 2 (2013a): 238–246. doi:10.1111/j.1467-923X.2013.12010.x. John, Peter. “All Tools Are Informational Now: How Information and Persuasion Define the Tools of Government.” Policy & Politics, 41, no. 4 (1 October 2013b): 605–620. doi:10.1332/030557312X655729. John, Peter, Graham Smith and Gerry Stoker. “Nudge, Nudge, Think Think: Two Strategies for Changing Civic Behaviour.” The Political Quarterly, 80, no. 3 (1 July 2009): 361–370. doi:10.1111/j.1467-923X.2009.02001.x. Kaine, Geoff, Helen Murdoch, Ruth Lourey and Denise Bewsell. “A Framework for Understanding Individual Response to Regulation.” Food Policy, 35, no. 6 (December 2010): 531–537. doi:10.1016/j.foodpol.2010.06.002. Kissane, Rebecca Joyce. “How Do Faith-Based Organizations Compare to Secular Providers? Nonprofit Directors’ and Poor Women’s Assessments of FBOs.” Journal of Poverty, 11, no. 4 (2007): 91–115. Knetsch, Jack. “Behavioural Economics, Policy Analysis and the Design of Regulatory Reform” in Behavioural Economics and Policy Design, edited by Donald Low, 161–182. Co-Published with Civil Service College Singapore, 2011. http://www.worldscientific.com/doi/abs/10.1142/9789814366014_0010. Koh, Tsin Yen. “Key Ideas in Behavioural Economics – and What They Mean for Policy Design” in Behavioural Economics and Policy Design, edited by Donald Low, 17–34. Co-Published with Civil Service College Singapore, 2011. http://www.worldscientific.com/doi/abs/10.1142/9789814366014_0002. Lasswell, Harold. “Key Symbols, Signs and Icons” in Symbols and Values: An Initial Study, edited by Lymon Bryson et al., 77–94. New York: Abe Books, 1954. Lasswell, Harold D. A Pre-View of Policy Sciences, New York: American Elsevier, 1971. Leggett, Will. “The Politics of Behaviour Change: Nudge, Neoliberalism and the State.” Policy & Politics, 42, no. 1 (29 January 2014): 3–19. doi:10.1332/030557312X655576. Lehner, Matthias, Oksana Mont and Eva Heiskanen. “Nudging – A Promising Tool for Sustainable Consumption Behaviour?” Journal of Cleaner Production, Special Volume: Transitions to Sustainable Consumption and Production in Cities, 134 (15 October 2016): 166–77. https://doi.org/10.1016/j.jclepro.2015.11.086. Liu, Peggy J., Jessica Wisdom, Christina A. Roberto, Linda J. Liu and Peter A. Ubel. “Using Behavioral Economics to Design More Effective Food Policies to Address Obesity.” Applied Economic Perspectives and Policy, 36, no. 1 (1 March 2014): 6–24. doi:10.1093/aepp/ppt027. Lowi, Theodore J. “Distribution, Regulation, Redistribution: The Functions of Government,” in Public Policies and Their Politics: Techniques of Government Control, edited by Randall B. Ripley, 27–40. New York: W.W. Norton, 1966. Lynn, Laurence E., Jr. “The Behavioral Foundations of Public Policy-Making.” The Journal of Business, 59, no. 4 (1 October 1986): 379–384. March, J.G. and J.P. Olsen. Rediscovering Institutions: The Organizational Basis of Politics. New York: The Free Press, 1989.

M4716-STRAßHEIM_9781785367847_t.indd 86

25/04/2019 13:32

Behavioural considerations in public policy  87 May, Peter J. “Compliance Motivations: Affirmative and Negative Bases.” Law & Society Review, 38, no. 1 (2004): 41–68. doi:10.1111/j.0023-9216.2004.03801002.x. Meier, Kenneth J. and David R. Morgan. “Citizen Compliance with Public Policy: The National Maximum Speed Law.” The Western Political Quarterly, 35, no. 2 (1 June 1982): 258–273. doi:10.2307/448019. Moseley, Alice, and Gerry Stoker. “Nudging Citizens? Prospects and Pitfalls Confronting a New Heuristic.” Resources, Conservation and Recycling, SI: Resourceful Behaviours, 79 (October 2013): 4–10. doi:10.1016/j. resconrec.2013.04.008. Mueller, Claus. The Politics of Communication: A Study in the Political Sociology of Language, Socialization and Legitimation, New York: Oxford University Press, 1973. Mulford, C.L. and A. Etzioni. “Why They Don’t Even When They Ought to: Implications of Compliance Theory for Policymakers” in Policy Research, edited by A. Etzioni, 47–62. Leiden: E.J. Brill, 1978. Oliver, Adam. “Nudging, Shoving, and Budging: Behavioural Economic-Informed Policy.” Public Administration, 1 March 2015, n.p. doi:10.1111/padm.12165. Parker, Christine and Vibeke Lehmann Nielsen. Explaining Compliance: Business Responses to Regulation, Cheltenham, UK and Northampton, MA: Edward Elgar Publishing, 2012. Perri 6, Charlotte Fletcher-Morgan and Leyland, Kate. “Making People More Responsible: The Blair Governments’ Programme for Changing Citizens’ Behaviour.” Political Studies, 58, no. 3 (1 June 2010): 427–449. doi:10.1111/j.1467-9248.2009.00817.x. Peters, B. Guy and F.K.M. van Nispen, eds, Public Policy Instruments: Evaluating the Tools of Public Administration, Cheltenham, UK and Northampton, MA: Edward Elgar Publishing, 1998. Pierce, Jonathan J., Saba Siddiki, Michael D. Jones, Kristin Schumacher, Andrew Pattison and Holly Peterson. “Social Construction and Policy Design: A Review of Past Applications.” Policy Studies Journal, 42, no. 1 (1 February 2014): 1–29. doi:10.1111/psj.12040. Pykett, Jessica, Rhys Jones, Marcus Welsh and Mark Whitehead. “The Art of Choosing and the Politics of Social Marketing.” Policy Studies, 35, no. 2 (4 March 2014): 97–114. doi:10.1080/01442872.2013.875141. Riker, William H. “Political Theory and the Art of Heresthetics” in Political Science: The State of the Discipline, edited by Ada W. Finifter, 47‒67. Washington, DC: American Political Science Association, 1983. Riker, William H. The Art of Political Manipulation, New Haven: Yale University Press, 1986. Rodgers, Harrell R. Coercion to Compliance, Or How Great Expectations in Washington Are Actually Realized at the Local Level, This Being the Saga of School Desegregation in the South as Told by Two Sympathetic Observers – Lessons on Getting Things Done, Lexington, MA: Lexington Books, 1975. Room, Graham. “Agile Policy on Complex Terrains – Nudge or Nuzzle?” 2 October 2013. http://www.horizons.gc.ca/eng/content/agile-policy-complex-terrains-%E2%80%93-nudge-or-nuzzle (accessed 10 January 2019). Salamon, Lester M., ed. Beyond Privatization: The Tools of Government Action, Washington, DC: Urban Institute, 1989. Salamon, Lester M., ed. The Tools of Government: A Guide to the New Governance, New York: Oxford University Press, 2002. Saward, Michael. Co-optive Politics and State Legitimacy, Aldershot: Dartmouth, 1992, pp. 27, 150 and 153. Schneider, A.L. and H. Ingram. “Behavioural Assumptions of Policy Tools.” Journal of Politics, 52, no. 2 (1990): 511–529. Schneider, A. and H. Ingram. “Social Construction of Target Populations: Implications for Politics and Policy.” American Political Science Review, 87, no. 2 (1993): 334–347. Schneider, A.L. and H. Ingram. Policy Design for Democracy. Lawrence, KS: University Press of Kansas, 1997. Schneider, Anne L. and Helen M. Ingram, eds. Deserving and Entitled: Social Constructions and Public Policy, SUNY Series in Public Policy. Albany: State University of New York, 2005. Schneider, Anne and Mara Sidney. “What Is Next for Policy Design and Social Construction Theory?” Policy Studies Journal, 37, no. 1 (2009): 103–119. Scholz, J.T. “Cooperative Regulatory Enforcement and the Politics of Administrative Effectiveness.” American Political Science Review, 85, no. 1 (1991): 115–136. Selinger, Evan and Kyle Powys Whyte. “Nudging Cannot Solve Complex Policy Problems.” European Journal of Risk Regulation, 2012 (2012): 26. Shafir, Eldar, ed. The Behavioral Foundations of Public Policy, Princeton: Princeton University Press, 2013. Stanbury, W. T., and J. Fulton. “Suasion as a Governing Instrument” in How Ottawa Spends 1984: The New Agenda, edited by A. Maslove, 282–324. Toronto: Lorimer, 1984. Steurer, Reinhard. “The Role of Governments in Corporate Social Responsibility: Characterizing Public Policies on CSR in Europe.” Policy Sciences 43, no. 1 (May 2009): 49–72. doi:10.1007/s11077-009-9084-4. Stokey, E. and R. Zeckhauser. A Primer for Policy Analysis, New York: W.W. Norton, 1978. Stover, Robert V. and Don W. Brown, “Understanding Compliance and Noncompliance with Law: The Contributions of Utility Theory.” Social Science Quarterly, 56, no. 3 (1 December 1975): 363–375.

M4716-STRAßHEIM_9781785367847_t.indd 87

25/04/2019 13:32

88  Handbook of behavioural change and public policy Tallontire, Anna. “CSR and Regulation: Towards a Framework for Understanding Private Standards Initiatives in the Agri-Food Chain.” Third World Quarterly, 28, no. 4 (2007): 775–791. Taylor, C.M., S.J.T. Pollard, A.J. Angus and S.A. Rocks. “Better by Design: Rethinking Interventions for Better Environmental Regulation.” Science of The Total Environment 447 (1 March 2013): 488–499. doi:10.1016/j. scitotenv.2012.12.073. Thaler, Richard H. and Cass R. Sunstein. Nudge: Improving Decisions About Health, Wealth, and Happiness, revised and expanded edition. New York: Penguin Books, 2009. Thaler, Richard H., Cass R. Sunstein and John P. Balz. “Choice Architecture.” SSRN Scholarly Paper. Rochester, NY: Social Science Research Network, 2 April 2010. http://papers.ssrn.com/abstract=1583509. Torgler, B. “Moral Suasion: An Alternative Tax Policy Strategy? Evidence from a Controlled Field Experiment in Switzerland.” Economics of Governance, 5, no. 1 (2004): 235–253. Trebilcock, M.J. “Regulating Service Quality in Professional Markets” in The Regulation of Quality: Products, Services, Workplaces, and the Environment, edited by D.N. Dewees, 83–108. Toronto: Butterworths, 1983. Trebilcock, M. and D.G. Hartle. “The Choice of Governing Instrument.” International Review of Law and Economics, 2 (1982): 29–46. Tribe, L.H. “Policy Science: Analysis or Ideology?” Philosophy and Public Affairs, 2, no. 1 (1972): 66–110. Tupper A. and G.B. Doern, “Public Corporations and Public Policy in Canada” in Public Corporations and Public Policy in Canada, edited by A. Tupper and G.B. Doern, 1–50. Montreal: Institute for Research on Public Policy, 1981. Vedung, Evert. “Policy Instruments: Typologies and Theories” in Carrots, Sticks and Sermons: Policy Instruments and Their Evaluation, edited by Marie-Louise Bemelmans-Videc, Ray C. Rist and Evert Vedung, 21–58. New Brunswick: Transaction Publishers, 1997. Weaver, Kent. “If You Build It, Will They Come? Overcoming Unforeseen Obstacles to Program Effectiveness.” The Tansley Lecture – University of Saskatchewan, 2009a. Weaver, Kent. “Target Compliance: The Final Frontier of Policy Implementation.” Washington, DC: Brookings Institution, 2009b. http://www.brookings.edu/research/papers/2009/09/30-compliance-weaver (accessed 10 January 2019). Weaver, Kent. “But Will It Work?: Implementation Analysis to Improve Government Performance.” Washington DC: Brookings Institution, 2010. http://www.brookings.edu/research/papers/2010/02/implementation-analy​ sis-weaver (accessed 10 January 2019). Weaver, R. Kent. “Compliance Regimes and Barriers to Behavioral Change.” Governance, 27, no. 2 (2014): 243–265. Weaver, R. Kent. “Getting People to Behave: Research Lessons for Policy Makers.” Public Administration Review, 75, no. 6 (2015): 806–816. Winter, Søren C. and Peter J. May. “Motivation for Compliance with Environmental Regulations.” Journal of Policy Analysis and Management, 20, no. 4 (2001): 675–698. doi:10.1002/pam.1023. Zehavi, Amos. “The Faith-Based Initiative in Comparative Perspective: Making Use of Religious Providers in Britain and the United States.” Comparative Politics, 40, no. 3 (2008): 331–351.

M4716-STRAßHEIM_9781785367847_t.indd 88

25/04/2019 13:32

PART II NUDGING AROUND THE WORLD: PROLIFERATION AND PROFESSIONALIZATION

M4716-STRAßHEIM_9781785367847_t.indd 89

25/04/2019 13:32

7.  Nudging around the world: a critical geography of the behaviour change agenda Mark Whitehead, Rhys Jones and Jessica Pykett

1. INTRODUCTION: GEOGRAPHY AND THE INTERNATIONALIZATION OF BEHAVIOURAL INSIGHTS In May 2014 an interesting partnership emerged between the World Bank, the Guatemalan government and the UK’s Behavioural Insights Team. In an attempt to support the tax raising efforts of the Guatemalan state, the Behavioural Insights Team was called upon to see if subtle nudges could be used in the wording of tax letters. The UK’s Behavioural Insights Team has for a long time been at the forefront of the development and application of the kinds of behaviour change tactics that this volume explores (see Jones et al. 2013). Following the promotion of psychologically informed policy developments in the UK, Australia, the Netherlands, Singapore, and USA (inter alia), the Behavioural Insights Team’s partnership with the Guatemalan tax authorities was its first with a government in the developing world (see Howgego 2014). This partnership was actually forged by the World Bank, who hired the Behavioural Insights Team to work in Guatemala. Over recent years the World Bank has become interested in the ways in which the emerging insights of the psychological and behavioural sciences could be used to shape international development policies (see World Bank 2015). The trial itself involved adapting and testing the wording of tax payment reminders letters in order to invoke conscious and unconscious prompts to pay (these prompts included stating the percentage of Guatemalans who pay their tax on time, forging a link between the paying of tax and national pride, and the suggestion that tax authorities were watching citizens). While these behavioural adaptations only resulted in relatively small percentage increases in the number of people who completed their tax returns, it demonstrated the potential of behavioural insights to generate significant extra revenues for the tax authorities. We recount the story of the Guatemalan tax authority’s behavioural turn here because it reveals something about the international spread of nudge-styles policies. There is a common, if often implicit, hypothesis concerning the geographical spread of behaviour changing policies. This hypothesis suggests that the insights developed within the Western academies of behavioural and psychological sciences, and first tested on public policy by prominent groups such as the UK Behavioural Insights Team, or the USA’s Social and Behavioural Sciences Team, are now, with the help of prominent organizations such as the World Bank (but also including USAid, EUAid, the OECD, and the World Economic Forum), gradually spreading to a wide range of countries around the world. Our research has been seeking to explore this hypothesis in two ways. First, we have explored the extent of the geographical spread of behaviourally-informed public policy-making. Second, we have analysed the nature of the geographical proliferation of nudge-style policies. On the basis of this research in this chapter we claim that there has been a significant geographi90

M4716-STRAßHEIM_9781785367847_t.indd 90

25/04/2019 13:32

Nudging around the world  91 cal spread of nudge-type policies around the world. Furthermore, however, we also claim that the geographical spread of behavioural policy has been more complex and variegated than is often recognized. This chapter begins with an account of the extent of the geographical spread of nudgetype policies around the world. In the next section we consider the uneven geographies that are associated with the internationalization of related behavioural policies. In the final section we consider how a geographical perspective on emerging forms of behavioural government can add critical perspective to analyses of such developments.

2.  THE GLOBALIZATION OF NUDGE In the autumn of 2013 we commenced the first systematic study of international spread of nudge-style policies (see Whitehead et al. 2014; see also Lunn 2014; OECD 2017).1 This project presented a series of challenges. The first concerned how precisely we could assess the presence or absence of nudge-type policies in one hundred and ninety five different countries around the world. The second concerned how we could accurately define and then identify ‘nudge-style policies’. In response to the first challenge we developed an online assessment methodology, which enabled us to effectively search for related policies in all of the independent states around the world (see Whitehead et al. 2014).2 On the basis of this methodology, it is important to recognize that this study did not provide a definitive list of where new behavioural insights are informing the development of public policy. It is highly likely that there are many states that are now deploying nudge-type policies that our study did not identify (this is largely because related policies simply did not show-up in our web-based research, and because of the time that has elapsed since the original study).3 Our survey was thus meant to serve as a minimum baseline indicator of where the new behavioural sciences associated with nudging have had an impact on public policy, and to use this baseline to consider the nature of this impact (for a recent international review of emerging behaviour change policies see OECD 2017). In response to the second challenge we developed a framework of policy tropes and proxies that we used to identify relevant policies. The idea of nudging people into new patterns of behaviour is based on the insight of behavioural economists that people routinely behave in non-rational ways. Our behaviours can plunge us into debt, make us unhealthy, and damage the environment. It is also based on work within behavioural psychology and cognitive design, which suggests that it is possible to reshape the environments within which people make decisions in a way that makes it easier for them to adopt financial, health and environmental behaviours that are in their own best interests. Three things essentially unite these sciences and philosophies of behaviour: (1) they all suggest that human behaviour is not only the product of autonomous (as in deliberative and intentional) individual decision-making; (2) they recognize the role of social/cultural norms (however rational or irrational they are) in influencing human behaviour; (3) they stress the role that our surrounding environments and choice architectures play in shaping conduct. Behavioural Insights policies have thus become popular because they suggest that it may be possible to address a range of social problems at minimal cost while also preserving people’s personal freedoms. In our research we were interested in nudging policies, but we were also concerned with a range of behavioural interventions that while

M4716-STRAßHEIM_9781785367847_t.indd 91

25/04/2019 13:32

92  Handbook of behavioural change and public policy Social-Cultural Community-Based Social Marketing Connected Conversations Anchoring Life-Stage Intervention Salience Co-Design

Social Norming

Segmenting Messenger: Peer-to-Peer Education Entertainment

Social Marketing

Non-verbal priming Non-Linguistic Framing Defaults

Choice Editing Data Streaming Choice Architecture

Design

Linguistic Framing Verbal Priming

Feedback

Communication

Figure 7.1  Nudge-type policy proxies and tropes not being referred to as nudges share many of their characteristics. We like to refer to these forms of behavioural interventions as part of the wider family of nudge-type policies. Some of these interventions (such as social marketing – the use of established marketing techniques of persuasion and suggestion to address social issues) actually predate nudge in their use and application. Figure 7.1 illustrates the broad range of policy techniques that have emerged in response to new understandings of the nature of human behaviour. We find it useful to organize these different policy initiatives around three clusters of activities, which often overlap. The first cluster of initiatives centres on the importance of salient forms of communication. Traditional forms of public policy tended to focus on the presentation of rational arguments to encourage behavioural shifts. Recent policy developments recognize that when pursuing behaviour change the way in which information is framed is as important as the actual information given. The second cluster of policy initiatives concerns socio-cultural norms. Research has shown that in the absence of alternative information humans will often follow the behaviours that they see being practiced by others when making a decision. The third and final set of policy initiatives concerns issues of design. Design plays a crucial role in shaping human behaviour. At one level design can relate to the shape and nature of our physical surroundings. It can also, however, relate to the layout of the forms and software that we are routinely expected to engage with. The policy tropes outlined in Figure 7.1 were the indicators that we used to identify the presence (or absence) of nudge-type policies in our study. In order to be able to be considered a nudge-type initiative in our study policies had to exhibit two key characteristics: (1) be non-regulatory and thus preserve the liberty of those who were subject to them;

M4716-STRAßHEIM_9781785367847_t.indd 92

25/04/2019 13:32

Nudging around the world  93 and (2) draw on behavioural insights (for example human preference for status quo over change; tendencies to following social norms; prioritizing the present over the future; and use of unconscious habits). We recognize that this policy framework meant that we cast our analytical net wide in relation to what we considered to be a nudge-related policy intervention. The policies that we counted included opt-out organ donation and company pension schemes; peer-to-peer public health initiatives; habit-oriented handwashing campaigns; and the creation of social norms around household energy use through socially comparable billing inter alia. We openly acknowledge that this methodology meant that we include in our study policies that would not under many definitions be considered true nudges. The advantage of our chosen methodology, however, was that it enabled us to track the broader impacts of behavioural and psychological sciences upon which nudge policies are based, and to consider a more varied set of geographical policy adaptations. We note also that there are many other policy instruments that are based on behavioural insights (including think strategies, boost techniques, and budge policies). These policies, however, were not included in our survey because we were primarily interested in the application of behavioural policies within non-regulatory circumstances that involved a particular fusion between psychological governance and liberal freedoms. While these policies may have been inspired by new behavioural insights they were either too regulatory (and thus not liberal enough), or too deliberative (and thus not psychologically aligned with irrational prompts), to be counted. In their analysis of behaviourally-informed public policy-making, the European Commission provides two broad taxonomies of related policy forms (see European Union 2016, pp. 15‒17). First, the European Union Report distinguishes between policies that have been explicitly and implicitly informed by behavioural insights. Our survey did not require policy initiatives to explicitly quote behavioural insights work, but merely to display the forms of hallmarks we would expect such policies to adopt. We thus included explicit and implicit nudge-style policies in our research. The European Commission further distinguishes between behaviourally-tested initiatives (those policies that are based on trials of policies directly informed by behavioural insights); behaviourally-informed initiatives (those policies that are based on the results of other behavioural insights studies); and behaviourally-aligned initiatives (policies that are not based on any previous behavioural insights research, but still use behavioural levers within the policy process) (European Union 2016, p. 16; Ciriolo et al., Chapter 8 in this volume). Our survey was explicitly designed to record evidence of behaviourally-tested, behaviourally-informed, and behaviourally-aligned policies. Utilizing these policy tropes, our research revealed that nudge-style policies had travelled around the world to a significant extent. Figure 7.2 shows all of the independent states (and Taiwan) where we have found some evidence of the impact of the new behavioural sciences on the design and/or implementation of public policy. As this figure demonstrates, the behavioural sciences are clearly having a global impact on public policy initiatives. According to our study, 135 independent states (and Taiwan) have seen the behavioural sciences have some impact on aspects of public policy delivery in some part of their territory (Whitehead et al. 2014). This number represents 69 per cent of all states. Given that this study is meant to provide a minimum baseline, it is highly likely that well over 70 per cent of independent states have seen the application of new behavioural insights within their public policy sector. What Figure 7.2 does not reveal is the great diversity of policy areas to which the

M4716-STRAßHEIM_9781785367847_t.indd 93

25/04/2019 13:32

94  Handbook of behavioural change and public policy

Figure 7.2 Independent states (shaded) where evidence was found of nudge-type public policies behavioural sciences are now being applied. They are now being used to shape policy areas as diverse as tax payments, hand washing/personal hygiene, HIV/AIDS prevention, vaccination programmes, charitable giving, malaria prevention, nutrition promotion, healthy pregnancy initiatives, fertilizer use, youth empowerment, breast feeding promotion, pension savings, police force reform, automated bank saving, preventing violence in schools, energy conservation, loan repayments, and organ donation, among many others (Whitehead et al. 2014). Of the 135 countries that have seen public policy development influenced by the new behavioural sciences, our survey revealed that 51 had developed centrally directed policy initiatives bearing the characteristic policy proxies of the behavioural sciences (over one quarter of the world’s independent states) (see Figure 7.3). Examples of centrally orchestrated policies range from systems of opt-out organ donation, to automatic pension enrolment, and pre-filled tax forms. Centrally orchestrated policy systems are significant because they are uniformly applied to resident populations within a state (as opposed to simply being part of more local behavioural interventions led by NGOs, corporations, and/or local governments). As national-level initiatives they have often been subject to much more public debate and scrutiny and their adoption consequently reflects a significant commitment to the insights of the new behavioural sciences on the part of the state in question.

3. THE GEOGRAPHICAL NATURE OF THE INTERNATIONALIZATION OF BEHAVIOURAL INSIGHTS In addition to exploring the extent of the geographical spread of nudge-type policies, our study was also designed to shed light on the nature of the spatial travel of new behavioural

M4716-STRAßHEIM_9781785367847_t.indd 94

25/04/2019 13:32

Nudging around the world  95

Figure 7.3 Independent states (shaded) with centrally orchestrated nudge-type policy initiatives insights. Two things in particular interested us in relation to the nature of this nudge geography. The first was any unevenness in the geographical development of related policy regimes. In this context we were particularly interested in why related policies were evident in some places and not others, and any variations in policy forms that were evident in places that had nudge policy regimes. The second area of interest for us was to ascertain what was supporting, channelling, and constraining the geographical spread of nudge-style policy regimes. In relation to the geographical spread of nudge-type policies, our research revealed certain geopolitical regions where it clearly had had far less impact than others. South (but not central) America, parts of Eastern Europe and large portions of the Middle East were all characterized by an apparent absence of nudge-type public policies (although we do recognize that the languages used in these regions may have resulted in our survey failing to identify relevant policy initiatives). At present we can only speculate about why certain regions appear to be impervious to the impacts of the behavioural sciences. We could actually be looking at a series of false negatives, and the regions in question have adopted nudge-type policies that our methodology has simply not picked up. There could, however, be something more systematic at play, with the political cultures in these regions offering resistance to the uptake of nudge-type behavioural policies. An important question for future research in this area is to ask whether the more authoritarian political traditions in Eastern Europe have acted as bulwarks against the softer forms of paternalism promoted by nudge-type policies (although this would not explain the presence of such policies in China and Russia). To these ends, it is important to consider whether the insights of the behavioural sciences are most applicable to public policies in liberal political systems – particularly given the popularity of such policies in North America, Europe, and Australasia. It is also important to consider the extent to which the new behavioural sciences could contribute to the development of less authoritarian

M4716-STRAßHEIM_9781785367847_t.indd 95

25/04/2019 13:32

96  Handbook of behavioural change and public policy forms of political intervention in certain regions of the world. The emerging popularity of nudge-style policies in Singapore is a particularly interesting case study in this context. As a form of ‘authoritarian democracy’, Singapore is characterized by the much stronger emphasis that it places on economic as opposed to political freedom (see Whitehead et al. 2017). In this context it is important to ask why authoritarian-leaning states would show interest in a form of policy that utilizes techniques that are specifically designed to not infringe on personal liberties. Ultimately, it appears that Singapore’s long interest in design-led solutions to social problems, and a broader move towards more liberal systems of governing have made nudge policies popular (Whitehead et al. 2017). Understanding why certain regions appear to have emerged as heartlands of nudge-type policy is also important. The impact of the behavioural sciences on policy design and implementation is clearly strongest in North America, Western Europe, and Australasia. At a general level, nudge-type policies have clearly taken hold in the English speaking states whose academies first developed the behavioural theories on which they are based. In more specific terms, there have been particular acts of policy transfer that have contributed to the concentration of nudge-type initiatives in certain places. So, Richard Thaler acted as a consultant to the UK government when it was in the process of developing the apparatus of its behavioural state (Thaler 2015). The resultant Behavioural Insights Team has inspired the US government to develop its own ‘nudge squad’ and advised the government of New South Wales in Australia (Halpern 2015). In terms of Europe, it is clear that the European Commission is playing an important role in the promotion of the behavioural sciences within its member states (for example, the European Commission produced a report entitled Applying Behavioural Sciences to EU Policy-making, van Bavel et al. 2013). The prominent role now being played by the behavioural sciences in African public policy can be attributed to a range of processes. It appears that many of the most pressing public policy agendas in Africa (in particular, preventing the spread of HIV/AIDS, promoting improved hygiene practices, and combating the spread of malaria) relate to everyday domestic practices and habits. As we have already noted, the new behavioural sciences are particularly adept at developing effective, and politically acceptable, ways of changing the types of household behaviours that conventional government policies have often failed to reach (Whitehead et al. 2014). Although some nudge-type policies are actually endogenous to Africa, it is important to recognize the important role that international development organizations have played in promoting the new behavioural sciences on the continent. International organizations such as USAID, AusAID, UNICEF, the World Health Organization, and the United Nations Population Fund have all been instrumental in this context. In addition to the aforementioned development agencies that have been advocates of nudge-type policies, our research also highlighted two other types of international organization that have been behind the global spread of the new behavioural sciences. First, are a series of international NGOs, charities, paragovernmental agencies and consultancies, such as the NSMC (National Social Marketing Centre, UK), the WWF and Change Labs. While charities have long been associated with the promotion of positive forms of behaviour change, it is interesting to see that major marketing consultancies are now promoting similar aims. This eclectic group of organizations is not only promoting the use of the new behavioural sciences in local and national governments, but they also offer behaviour change expertise that is used to support the programmes of major development agencies.

M4716-STRAßHEIM_9781785367847_t.indd 96

25/04/2019 13:32

Nudging around the world  97 Second is a series of multinational corporations that are embracing the principles of behaviour change as part of their emerging Corporate Social Responsibility programmes. Corporations have, of course, always been in the business of behaviour change. Convincing consumers of the need to buy their new product or services, or persuading them to shift to their particular brand, is what corporate marketing is all about. It is interesting to note that the behavioural insights of the corporate world (including the importance of segmentation, product placement, narrative, reciprocity, inertia, and habit) have been central to the public sector’s newly emerging behaviour change initiatives. What appears to be changing is that now the private sector is learning how to promote what are routinely referred to as ‘sustainable’ or ‘positive’ behaviours (which essentially involve encouraging/ enabling consumers to lead more healthy, financially responsible, and environmentally benign lives). It is in this context that large corporations such as Unilever are now becoming active in promoting the use of new behavioural insights within the pursuit of public policy goals (see Whitehead et al. 2017, Chapter 7). Unilever, for example, are now utilizing behavioural insights to inform their community-based sustainable development programmes. Given the financial power and global reach of these corporations, it is clear that they are key drivers within the globalization of the behavioural sciences. It is important to reflect upon why corporations are becoming so involved in positive behaviour change interventions and what the implications of this process are. Conventional wisdom suggests that corporations promote positive behaviour change in order to improve their public image and ultimately enhance their bottom line. Our research, however, suggests that the corporate promotion of behaviour change is about establishing a non-commercial relationship with their consumers (of caring non-exploitation), which helps to cement forms of commercial brand loyalty. While the corporate promotion of responsible social behaviours should be encouraged, the emerging role of corporations in the field of public policy does raise some interesting ethical questions. Precisely where the line between positive and commercially-oriented forms of behaviour change (or indeed between social marketing and plain marketing) lies is often difficult to discern, but nonetheless needs be rigorously policed. There are considerable differences in the ways in which central governments have promoted the use of the behavioural sciences. In many of the states represented in Figure 7.3 governments have simply promoted the use of behavioural sciences in one particular policy area. In China, for example, the department of health has been promoting the use of social marketing in its HIV/AIDS prevention programme. In other countries, such as Belgium, we see the application of new behavioural insights across a number of policy areas (in this case organ donation, energy use and tax payment initiatives). A further set of states, including the Netherlands, Singapore, Australia, France, the UK and the USA, have seen more strategic attempts to integrate new behavioural insights across many relevant policy sectors. In the USA, for example, the Obama administration appointed the nudge-advocate Cass Sunstein to head its Office of Information and Regulatory Affairs (OIRA) (see Sunstein 2013). Given that the OIRA has oversight right across the Federal Government, Sunstein was able to apply the insights from behavioural economics and psychology to a range of policy areas. Perhaps unsurprisingly, our study revealed an uneven geography to the spread and up-take of nudge-type policies around the world. An appreciation of this geography raises some interesting questions in and of itself concerning the processes that shape and steer

M4716-STRAßHEIM_9781785367847_t.indd 97

25/04/2019 13:32

98  Handbook of behavioural change and public policy policy transfer and transformation (see Jones et al. 2014). It is our contention, however, that a geographical perspective on nudge-style policies can also facilitate a broader critical analysis of emerging, international systems of behavioural governance. It is to this form of geographical project that the following section turns.

4.  A CRITICAL GEOGRAPHY OF NUDGE Understanding the emerging geography of nudge-style policies has more than merely descriptive value. There already exists a series of attempts to excavate the historical evolution of behaviourally-informed policies (for a review of these studies see Jones et al. 2013). What tends to characterize these genealogies is an attempt to uncover coherent, and often singular points of emergence (often key individuals, such as Daniel Kahneman, or key institutions, like the Behavioural Insights Team). We acknowledge that histories of nudge can be used to uncover the diverse origin points of behavioural policy regimes (see for examples Whitehead et al. 2017). However, we also assert that considering nudge geographically more readily facilitates an account of the significant variations that characterize emerging systems of behavioural policy-making. Geography in this context can be used to expose more than merely the uneven spatial application of nudge-styled policies: it can be employed to consider how, why, and to what effect related policies can themselves take different forms. It is also our contention that spatializing nudge enables a more critical perspective to be developed on related policy initiatives. Over the last decade we have been engaged in the development of a critical geographical study of nudge policies (see Jones et al. 2011; Jones et al. 2013; Whitehead et al. 2017; see also the work of Berndt 2015, Berndt Chapter 17 in this volume; and Strauss 2009). This project has involved an ongoing attempt to fuse the study of emerging systems of behavioural governance with broader meta-physical, political, and sociological concerns with the human condition, the nature of social transformation, and political emancipation. At the heart of this project has been the development of three interconnected critical perspectives on the nudge project (see Jones et al. 2013). The first critical perspective concerns the ethics of nudging, and the extent to which the often-unconscious targets of behavioural policy transcend established norms concerning the right to personal autonomy and self-determination. Ethical concerns have formed an important part of the governmental scrutiny of nudge policies (see Sunstein 2016) even though they still appear to be a relative marginal concern for those involved in the delivery of behavioural policy (OECD 2017). The second critical perspective concerns the efficacy, or long-term effectiveness, of related policies. The idea of the enduring nudge – or nudges that can shift behaviour over longer periods within people’s lifestyle – is becoming something of an elusive goal of nudge strategist. As a goal the quest for enduring nudges draws attention to some of the observed limitations of nudge when it comes to long-term behavioural shifts. Our own work has sought to draw particular attention to the limited appreciation of the nature of socio-technical transformations evident in many systems of behavioural governance (Whitehead et al. 2017). The third critical perspective on nudge policies we have explored is the issue of empowerment. We use the notion of empowerment to denote the extent to which behavioural insights into the cognitive limitations associated with human decisionmaking are utilized within public policy-making as a basis for legitimately bypassing

M4716-STRAßHEIM_9781785367847_t.indd 98

25/04/2019 13:32

Nudging around the world  99 conscious acts of human behaviour change. There are many examples of novel forms of behavioural policy, which shift the responsibility for behaviour change from the deliberative human subject and into the surrounding choice environments. There are also clear attempts to utilize the insights of the behavioural and psychological sciences as a basis for developing enhanced forms of neurological reflectively in and through which people are empowered to nudge their own way out of problematic behavioural patterns (see Rowson 2011; Whitehead et al. 2017). Classic forms of neurological reflectivity include self-nudges, such as the self-exclusion schemes employed by gambling addicts. At the heart of this critical project is an attempt to reveal that there is more at stake within emerging systems of behavioural governance than merely successful or unsuccessful iterations of public policy. The ethics of nudge-style policies raise constitutional issues about the nature of the relation between citizens and the state. The efficacy of emerging behavioural policies expose important divisions between nuanced, positivist accounts of human behaviour and motivations, and more metaphysical concerns with the structuration of socio-technological practices and change (Davies and Doyle 2015; Shove 2010). Furthermore, the questions of empowerment surfaced by nudge policies bring into sharper focus questions concerning our collective understanding of human character and dignity (Furedi 2011). None of these critical perspectives necessarily means the impacts of the behavioural sciences on public policies are pernicious. These emerging impacts do, however, behove us to think about their broader impacts of behavioural sciences on public life and politics. Geography is important to this critical project for two reasons. First because it can reveal spatial differences in the application of nudge style policies, which in turn expose variations in the potential ethical impacts, efficacy, and empowering potential of emergent policies. Second, in exposing these differences a geographical perspective necessarily positions nudge policies within broader political, economic, cultural, and social systems and practices, which are routinely ignored within the behavioural sciences themselves. Being sentient of geographical variations and adaptations of nudge policies provides an important context to consider the forces that may be shaping such differences. Relative levels of scientific expertise, available resources, institutional capacities, political ideologies, cultural norms, and/or technological capacity can all drive variations in nudge policies. Exposing these processes of differentiation represents an important step in understanding the conditions that best support the development of ethically sensitive, effective, and socially empowering variants of behavioural policy, and likewise the processes that inculcate less progressive policy forms.

5. CONCLUSIONS This chapter has explored the extent of the geographical spread of behaviourallyinformed public policy-making and analysed the nature of the geographical proliferation of such policies. It has demonstrated that there has been a significant geographical spread of nudge-type policies around the world (a proliferation process that appears set to only expand in the immediate future). Crucially this chapter has also demonstrated that the geographical spread of behavioural policy has been more complex and variegated than is often recognized.

M4716-STRAßHEIM_9781785367847_t.indd 99

25/04/2019 13:32

100  Handbook of behavioural change and public policy This chapter began with an account of the extent of the geographical spread of nudgetype policies around the world. According to our research the behavioural sciences are clearly having a global impact on public policy initiatives. Our study revealed that 135 independent states (and Taiwan) have seen the new behavioural sciences have some impact on aspects of public policy delivery in some part of their territory (Whitehead et al. 2014). This number represents 69 per cent of all states. Given that this study is meant to provide a minimum baseline, and was completed in 2013, it is highly likely that well over 70 per cent of independent states have seen the application of new behavioural insights within their public policy sector. Our survey also revealed that 51 states had developed centrally directed policy initiatives bearing the characteristic policy proxies of the behavioural sciences. In the second section of this chapter we considered the uneven geographies that are associated with the global spread of nudge-styled policies. Our research revealed certain geopolitical regions where related policies had clearly had far less impact than others. South (but not central) America, parts of Eastern Europe and large portions of the Middle East were all characterized by an apparent absence of nudge-type public policies. Our research also revealed key heartlands of nudge-type policies. The impact of the behavioural sciences on policy design and implementation is clearly strongest in North America, Western Europe, Australasia, part of South East Asia and Africa. Within these heartland areas our research revealed differences in the extent to which nudge policies have been adopted by different states. Certain governments utilized behavioural policies in isolated policies domains, while others had rolled-out nudge policies across a series of major policy areas. In the final section of this chapter we considered how a geographical perspective on emerging forms of behavioural government could add a critical perspective to analyses of such developments. In this context we argued that geography could be used to expose more than merely the uneven spatial application of nudge-styled policies – it can, we argued, be employed to consider how, why, and to what effect related policies take different forms. As part of our ongoing research into the impact of the behavioural and psychological sciences on public policy, we have utilized a geographical perspective to consider variations in the ethics, efficacy, and empowering potential of emerging nudge style policy regimes. Ultimately, we claim that an awareness of geographical variations (or indeed similarities) in behaviourally-informed public policy provides a framework within which it becomes possible to understand more effectively the forces that shape these policies and, offers a basis to promote more effective, empowering and ethical practices of nudging.

NOTES 1. We acknowledge that related policy databases do exist. Particularly helpful resources in this context are Mark Egaat’s Nudge Database: http://economicspsychologypolicy.blogspot.ie/2013/03/nudge-database_3441. html; the iNudgeyou portal: http://www.inudgeyou.com/; and the Nudge Blog: http://nudges.org/. But none of these resources offers an internationally comprehensive account of nudge policies. 2. This methodology involved us conducting a systematic online search for evidence of nudge-style policy proxies in 195 independent states (and Taiwan). In conducting this research we ran selected searches on Google that combined the independent state’s name with two search strings: the first including the phrase ‘nudge’ the second ‘behaviour change’. Utilizing these search strings, we identified related policy initiatives in the selected states. We limited our search to the first two pages of Google (approximately the first 20

M4716-STRAßHEIM_9781785367847_t.indd 100

25/04/2019 13:32

Nudging around the world  101 entries). The available online documentation associated with these policy initiatives was then searched to see if it contained evidence of the application of our identified policy proxies. While evidence of just one of our policy proxies ensured that a state was flagged as showing evidence of the impact of the new behavioural sciences, we collected evidence of multiple related policy initiatives if they were present. 3. We also recognize that our inability to be able to recognize/translate the policy proxies used in this report into all of the languages used in all the states studied limits the scope of our report.

REFERENCES Berndt, C. (2015) ‘Behavioural economics, experimentalism and the marketization of development’ Economy and Society 44: 567‒591. Davies, A. and Doyle, R. (2015) ‘Transforming household consumption: from backcasting to homelabs experiment’ Annals of the Association of American Geographers 205: 425‒436. European Union (2016) Behavioural Insights Applied to Policy – European Report European Union accessed 16 March 2016 at http://publications.jrc.ec.europa.eu/repository/bitstream/JRC100146/kjna27726enn_new.pdf. Furedi, F. (2011) On Tolerance: A Defence of Moral Independence London: Continuum. Halpern, D. (2015) Inside the Nudge Unit: How Small Changes and Make a Big Difference Virgin Digital. Howgego, J. (2014) ‘How behavioural science could revamp development’ SciDevNet accessed 1 March 2017 at http://www.scidev.net/global/policy/feature/behavioural-science-development-policy-innovation.html. Jones, R., Pykett, J. and Whitehead, M. (2011) ‘Governing temptation: changing behaviour in an age of libertarian paternalism’ Progress in Human Geography 35: 483‒501. Jones, R., Pykett, J. and Whitehead, M. (2013) Changing Behaviours: On the Rise of Psychological State Cheltenham, UK and Northampton, MA: Edward Elgar Publishing. Jones, R., Pykett, J. Whitehead, M. (2014) ‘The geographies of policy translation: how nudge became the default policy option’ Environment and Planning C: Government and Policy 32: 54‒69. Lunn, P. (2014) Regulatory Policy and Behavioural Economics Paris: OECD Publishing. OECD (2017) Behavioural Insights and Public Policy: Lesson from Around the World Paris: OECD Publishing. Rowson, J. (2011) Transforming Behaviour Change: Beyond Nudge and Neuromania London: Royal Society for the Encouragement of Arts, Manufactures and Commerce. Shove, E. (2010) ‘Beyond the ABC: climate change policy and theories of social change’ Environment and Planning A 42: 1273‒1285. Strauss, K. (2009) ‘Cognition, context, and multimethod approaches to economic decision making’ Environment and Planning A 41: 302‒317. Sunstein, C. (2013) Simpler: The Future of Government New York: Simon & Schuster. Sunstein, C. (2016) The Ethics of Influence: Government in the Age of Behavioural Science Cambridge: Cambridge University Press. Thaler, R. (2015) Misbehaving: The Making of Behavioural Economics London: Penguin. Van Bavel, R. Herrmann, B. Esposito, G. and Proestakis, A. (2013) Applying Behavioural Sciences to EU Policymaking Brussels: European Commission. Whitehead, M. Howell, R. Jones, R. Lilley, R. and Pykett, J. (2014) Nudging All Over the World: Assessing the Impacts of the Behavioural Sciences on Public Policy Aberystwyth: ESRC. Whitehead, M. Jones, R. Lilley, R. Pykett, J. and Howell, R. (2017) Neuroliberalism: Behavioural Government in the 21st Century Abingdon: Routledge. World Bank (2015) Mind, Body and Society – World Development Report Washington, DC: World Bank.

M4716-STRAßHEIM_9781785367847_t.indd 101

25/04/2019 13:32

8.  The application of behavioural insights to policy in Europe Emanuele Ciriolo, Joana Sousa Lourenço and Sara Rafael Almeida

1. INTRODUCTION The effectiveness of public policies often depends on how people react to it, and on the extent to which people’s real behaviour is taken into account during the initial stage of policy design. For instance, behavioural elements – such as default bias, present bias, loss aversion, or overconfidence – should be taken into account in policy design so that the proposed solutions explicitly tackle them. Behavioural insights (BIs, hereafter) are becoming valuable input to policy-making in international organizations (such as the European Commission, OECD, World Bank), and are increasingly influencing policy-making at the national level. This trend may well continue as policy-makers are becoming more and more interested and familiar with experimentation. More recently, large existing datasets of behavioural data have started to be used to extrapolate insights for policy-making (see e.g. Goldstone and Lupyan 2016). 1.1  Behavioural Insights at the European Commission The European Commission’s (EC) first explicit attempt to inform policy-making with BIs dates back to 2008. By acknowledging the existing scientific evidence on the impact of default options, the EC proposed a Directive on Consumer Rights including a clause limiting the use of default options in consumer contracts. Following this forerunner policy case, the EC applied or explored the application of BIs in a number of policy fields (including taxation, financial services, energy labels and online gambling) (Lourenço et al. 2016). In 2014, the EC created a Foresight and Behavioural Insights Unit, within its Joint Research Centre (JRC). In the same year, the OECD published an influential report reviewing applications of behavioural economics to regulatory policy across the world, in 2015 the World Bank published its yearly World Development Report, where it stressed the need for an expanded understanding of human behaviour for economic development (Lourenço et al. 2016), and in March 2017 the OECD published Behavioural Insights and Public Policy, Lessons from Around the World (OECD 2017). With a similar viewpoint, in 2016 we had published Behavioural Insights Applied to Policy: European Report 2016 (BIAP 2016, hereafter), in which we put forward a ­taxonomy of behavioural policy interventions, encompassed a wealth of policy applications either implicitly or explicitly informed by BIs across 32 European countries and reviewed institutional developments across Europe. This chapter draws on that research project.1

102

M4716-STRAßHEIM_9781785367847_t.indd 102

25/04/2019 13:32

The application of behavioural insights to policy in Europe  103 1.2  Application of Behavioural Insights across EU Member States There have been a number of developments in the application of BIs to policy at national level. Namely, since the creation of the UK government Behavioural Insights Team (UK BIT) in 2010, several EU countries – Germany, the Netherlands, France and Denmark – established behavioural teams. Early developments were also observed in Finland and Austria (Lourenço et al. 2016). In line with the growing interest and visibility of behavioural insights, new developments have already taken place. In Belgium, the Department of Public Governance and Chancellery from the Government of Flanders, has dedicated staff working on BIs (one full- and three part-time, in October 2016) and, in April 2016, a Framework Contract focusing on BIs applied to policy and communication (Department of Public Governance and Chancellery 2016). In Italy, the city of Milan and the region of Lazio paid serious attention and expressed interest in applying BIs to policy-making. In May 2016 the Presidency of the Council of Ministers appointed Matteo Motterlini, a philosopher and neuroeconomist, as Adviser for Social and Behavioural Sciences (NA 2016). In October 2017, the region of Lazio created a ‘Nudging working group’ mandated with putting forward a concrete strategy to be executed in the months to come, including specific issues susceptible of being tackled behaviourally. In Portugal, on October 2016, the Ministry of the Presidency and Administrative Modernisation launched the Public Sector Experimentation Laboratory (LabX). LabX aims to create a new innovation culture in public administration, and will use a user-centric approach and an innovation by design methodology. Their team is multidisciplinary, with a core group of project managers, service designers and social scientists, at times working with external experts in pertinent areas (behavioural sciences, sociology, anthropology, psychology, design, engineering, etc.). Finally, as noted in the recent annual report of the UK Team (UK BIT), BIs have become mainstream among UK policy-makers. Currently, 15 UK government departments or agencies have behavioural capacity through different modalities: a fully dedicated BIs unit; individuals appointed to manage networks of practitioners and coordinate BIs activities; or commissioned BIs projects (The Behavioural Insights Team 2016).

2. INSIGHTS FROM BEHAVIOURAL SCIENCES IN POLICY-MAKING Behavioural sciences involve the systematic analysis of the processes underlying human behaviour, through observation and experimentation. In doing so, they typically combine knowledge and research methods from the fields of psychology, economics, sociology and neuroscience (Lourenço et al. 2016). The integration of psychology research into economic science has helped challenge the assumptions on human behaviour underpinning the theory of rational choice. This development made the dismal science more in tune with reality, filling the gap between the formal elegance of economic models and their relative lack of realism. The neoclassical theory of rational choice assumes that individuals have well-defined preferences and unbiased beliefs, always take into account the available information and make self-interested

M4716-STRAßHEIM_9781785367847_t.indd 103

25/04/2019 13:32

104  Handbook of behavioural change and public policy and temporally consistent decisions. In other words, it portrays humans (also referred as homo economicus) as being consistently rational agents who maximize their self-interest. Notwithstanding the assumptions of rational economic theory, the role of psychology in economics has been defended by economists for a long time. For example, in 1896, Pareto wrote: ‘The foundation of political economy, and, in general of every social science, is evidently psychology. A day may come when we shall be able to decide the laws of social science from the principles of psychology’ (Pareto 2014, as cited by Thaler 2016a, preface xi). More recently, in 1979, Daniel Kahneman and Amos Tversky developed the prospect theory, based on the theory of human cognition (Kahneman and Tversky 1973). The prospect theory – describing the way people choose between probabilistic alternatives that involve risk when the probabilities of outcomes are known – represents the first concrete development as it uses psychological insights to explain economic decisions. From a policy perspective, relying on unrealistic assumptions about people’s behaviour may have severe consequences and lead to ineffective policies. The evidence above shows that – besides the increasing consensus on the value added of applying BIs to tackle social policy issues, where behavioural biases prevent individuals from taking the best decisions – the same view started to gain acceptance in more conservative fields, as competition policy, that had since then represented the stronghold of neoclassical economics. In competition policy, public intervention in the economy is warranted not only when there is case of market failure (that is externalities, natural monopoly, public goods), but also in cases where behavioural market failures are observed (OFT 2010). The prevalence of a biased demand, generated by imperfectly rational consumers, could be seen as a ‘fourth’ market failure, a behavioural market failure. This is true to such an extent that BIs have become a key part of every regulator’s toolkit (e.g. see Lunn 2014). In the cases mentioned above, for example, mandated disclosure – tailored to imperfectly rational consumers, or to sophisticated intermediaries that advise imperfectly rational consumers – can be of help. Also, the relationship between complexity and market competition has been under more intense scrutiny during the last years. A study based on country and distributor panel data for 15 countries over the period 2002‒2010 found that complexity of financial products increases when product market competition is high (Célérier and Vallée 2013). Ultimately, this evidence demonstrates the fundamental role of financial regulation and consumer protection in open and highly sophisticated markets. There are several examples of the value of inserting a behavioural dimension in policymaking and policy analysis. In this section we briefly mention cases from the field of energy (1), education (2), health (3), and taxation (4): 1. Public policy has so far mainly used price-based approaches, such as subsidies, to increase the uptake of more energy-efficient appliances. Nonetheless, non-price interventions, such as the use of social norms, have induced people to conserve energy. Social norms are rules of behaviour that affect the way individuals interact with others by signalling the appropriate behaviour. A notable non-price energy conservation programme ran by a company called Opower consisted in mailing personalized Home Energy Report letters (HERs) which compared a household’s energy use to that of similar neighbours and provided energy conservation tips.

M4716-STRAßHEIM_9781785367847_t.indd 104

25/04/2019 13:32

The application of behavioural insights to policy in Europe  105 Using a randomized-controlled trial (RCT)2 researchers found that the HERs reduced household energy consumption, both in the short and long term (Allcott 2011). 2. In the field of education, research on framing and loss aversion has shown that framing teacher incentives as losses relative to a higher income, rather than rewarding good performance with a bonus, increases the impact of these incentives on student performance (Fryer Jr. et al. 2012). 3. Health policy has used BIs to increase organ donations. The well-known study ‘Do Defaults Save Lives?’ revealed the mismatch between public attitudes and public action regarding organ donation and showed the positive effect of defaults (that is opt-out systems) on donation agreement rates (Johnson and Goldstein 2003). Optout systems for organ donation are in place in several countries in Europe (Lourenço et al. 2016). 4. Leveraging on our individual tendency to overweight small probabilities in the decision-making process, several countries in Europe have introduced tax lotteries (e.g. Malta, Poland, Portugal, Romania, Slovakia) to increase tax compliance (Lourenço et al. 2016). These examples highlight the influence of context and information framing in the decision-making process, as well as the importance of targeting the right audience when planning an intervention. Furthermore, testing and experimentation of possible solutions, as part of the design of public policy, are key to understand what works, how and where (Haynes et al. 2012). The behavioural approach offers the opportunity to develop better models of economic behaviour by incorporating insights from other social science disciplines (Thaler 2016b). By focusing on how people actually make choices, BIs contribute to the delivery of more effective policy solutions. In this sense, policy-makers have started to take a broader view and to look at how policy-making can be improved by the use of behavioural evidence.

3. BEHAVIOURAL INSIGHTS ARE MUCH MORE THAN JUST NUDGING Behavioural economics, behavioural insights and nudging are sometimes used as if they have the same meaning and reach. It should be noticed, instead, that although connected, they are fundamentally different (Lourenço et al. 2016). Behavioural economics has already been defined in the previous section. Behavioural insights result from multidisciplinary research in fields such as economics, psychology, sociology and neuroscience, to understand how humans behave and make decisions in everyday life. The concept of nudging was originally used by Richard Thaler and Cass Sunstein to indicate ’any aspect of the choice architecture that alters people’s behaviour in a predictable way without forbidding any options or significantly changing their economic incentives’ (Sunstein and Thaler 2008, p. 6). Before new concepts such as behavioural insights, behavioural biases and even nudges became popular, economists used to speak of anomalies, as opposed to consistencies. Richard Thaler had a regular homonymous column, Anomalies, on the

M4716-STRAßHEIM_9781785367847_t.indd 105

25/04/2019 13:32

106  Handbook of behavioural change and public policy Table 8.1  Behavioural insights versus nudges Features

Behavioural Insights

Nudges

Occurrence in the policy  process Approach

Input

Output

Broader repertoire of policy tools

Focus on choice architecture

American Economic Association’s Journal of Economic Perspectives, in which he took under review cases of economic behaviour that seemed to violate traditional economic theory. Such anomalies were nothing else than what later came to be known as behavioural biases: behavioural features at odds with the homo economicus’ artefact. Of course, such acknowledgement had great normative consequences since, if people are not fully rational and independent, policies should take this into account to pursue the desired outcome. For example, if people are influenced by what others do (as in a bank run context), social norms can be used to bring about positive changes. Over time, the idea that BIs could inform policies, for these to be more effective, gained increasing traction. However, after the publication of Nudge (Sunstein and Thaler 2008), BIs started to be directly associated with nudges. This was and still is so much of a tendency that the UK BIT came to be unofficially called the ‘Nudge Unit’. However, BIs go well beyond nudges. The latter is an easy and often low-cost intervention (that is an output of the policy process) that modifies the choice architecture, altering people’s behaviour in a predictable way, while preserving the same range of choice options (see Table 8.1). By contrast, BIs represent an input to the policy process, and can be fully integrated with and inform other traditional forms of intervention (that is regulations, incentives, information requirements). In this sense, BIs may support a broader range of policy instruments. This reinforces the idea that, contrarily to nudges, BIs – being an input to the policy process – may help inform traditional forms of policy interventions, as presenting information to the public, designing incentives, or regulating markets. In this sense, BIs do not warrant a specific type of policy action and indeed sometimes suggest that no intervention or a conventional one is the best solution. If policy-makers were to only consider nudges, as a form of behavioural intervention, they would de facto relinquish a wealth of information that BIs could provide. They would forgo useful insights at the beginning and at the end of the policy process. At the beginning of the policy process, BIs help identify the root causes of the problem matter, looking beyond the mere utilitarian approach whereby people take decisions maximizing their welfare, but complementing this view with a rather systematic analysis of co-determinants of individual behaviour. At the end of the policy process, ignoring BIs would entail restricting the range of alternative and feasible policy options. The case of seatbelts is an example of how BIs could have informed and improved policy-making and saved a significant number of lives (see also Graf in this volume). The first patent for seatbelts dates back to 1885. By 1950 some automotive companies started offering seatbelts as either an optional or standard feature. However, over the next five decades, few car drivers or passengers wore seatbelts in Europe. Although the death

M4716-STRAßHEIM_9781785367847_t.indd 106

25/04/2019 13:32

The application of behavioural insights to policy in Europe  107 toll from road accidents became more significant – and visible through media’s frequent accounts of tragic road accidents – people resisted behavioural change. Such resistance to change may be explained from a behavioural perspective. People may have perceived the immediate cost of wearing seatbelts as larger than its future (probabilistic) benefit, a feature called myopia or short-sightedness in the behavioural literature. Or people may be overconfident that road accidents would only happen to others. After all, when asked whether people thought of themselves as good drivers – and, in particular, whether people would rank themselves above the median driver – 80 per cent of respondents rate themselves in the top 30 per cent of drivers (Svenson 1981), a clear contradiction in terms. Or it could simply be a status-quo bias discouraging adopting new habits in a space, our car, which is all too often presented as a symbol of freedom and adventure.3 It is not clear whether the European legislator took these BIs implicitly into account when, in 2006, wearing seatbelts became compulsory for all vehicles throughout the EU. This example is very illustrative of the possibility to use BIs to inform traditional forms of intervention, be these based on regulation, on incentives or information disclosure. It should be noticed that no nudge, no liberty-preserving intervention, could have achieved the same result, and could have saved hundreds of thousands of lives by now (European Commission 2006). Easy access to seatbelts did not encourage people to wear them, until an explicit regulation made it compulsory. Had the policy-maker incorporated BIs from the start, the obligation to wear seatbelts would not have taken ages to manifest itself, and could have been consciously taken much earlier, with huge benefits for society.

4. IMPACT OF THE MAIN BEHAVIOURAL POLICYINTERVENTIONS AT EU LEVEL Starting from 2008, BIs have been integrated in EU policy-making. They have first been applied in consumer and competition policy – spanning areas including terms in consumer contracts through to designing remedies in competition cases (Ciriolo 2011). They were then used to inform policy actions in a number of other fields, including financial regulation, energy policy, online gambling, tobacco and communication policy. This section gives account of the initial policy interventions informed by BIs at EU level, and touches upon on-going developments. The European Commission was a precursor in using BIs in a competition case, to tackle a case of abuse of dominant position, linked to Internet Explorer (IE) being tied to Windows. Contrary to its traditional approach, DG Competition did not impose a fine for infringement of competition law, but rather tackled the supply-side issue by leveraging the demand side. Users of Windows-based PCs were provided with the option to choose an alternative browser, via an on-screen ballot box. This remedy – at work between 2010 and 2014 – pushed consumers to make an active choice as to their preferred browser, and implicitly removed the impact of the default option (Ciriolo 2016). The available evidence suggests that the remedy was more effective than the traditional ones adopted in the past: among the users who viewed the ballot box, one in four downloaded an alternative browser. To this respect, it is illustrative to run a comparative analysis of the web browsers’ market shares, in Europe and North America, since in the

M4716-STRAßHEIM_9781785367847_t.indd 107

25/04/2019 13:32

108  Handbook of behavioural change and public policy latter region no ballot box was in place during the same period. Excluding tablets and smartphones, between March 2010 and November 2014 (the period when the ballot box was active), the market share of IE in Europe dropped from 47 per cent to 17 per cent, whereas it decreased by 23 percentage points (from 55 per cent to 32 per cent) in North America (Ciriolo 2016). Instead of a monetary sanction by the antitrust authority, the imposition of a simple device – which entailed small programming cost for Microsoft – de facto translated into an EU market for web browsers that is substantially more competitive (Coyle 2016). The recent increased rate of innovation in this sector further strengthens this finding. The case of the Internet Explorer Ballot Box epitomizes the importance of considering the demand-side behaviour that co-determines the outcomes in a market, besides the supply-side features. Beyond this very case, BIs provide explanations of why poor consumer outcomes may arise even in markets where there are several competing firms (Kay 2011). This is often the situation in newly-liberalized markets – for example, the markets for energy, fixed and mobile telephony and financial services – where superfluous complexity and the specific features of the service being subscribed de facto contribute to disempower consumers. For this very reason, in 2010 the European Commission carried out the first behavioural study about consumers’ decision-making on retail investment services. At EU level, the novel aspect of this study consisted of a series of on-line and face-to-face experiments conducted with 6 000 consumers in eight EU countries, which produced a number of striking results. For example, it showed that people struggle to make optimal investment choices, even in very simplified investment tasks. Only 56 per cent of funds were invested optimally, and only 1.4 per cent of subjects managed to make all five investment choices optimally. It also confirmed that investment decisions are prone to biases and framing effects. Subjects made worse investment decisions when the optimal choice was harder to understand (fees framed as percentages, annual returns not compounded over the duration of the investment), and they were disproportionately averse to uncertainty, ambiguity and product complexity (Ciriolo 2011). In terms of policy recommendations, the results of the study suggested that standardizing and simplifying product information can improve consumer choices, and such evidence was explicitly used in the review of the Packaged Retail Investment Products (PRIPs) legislation. The impact of the study goes beyond its direct application in the sector of retail investment services. Indeed, its results confirm that disclosure of information alone will often be insufficient to provide consumers with what is needed to optimize their understanding and decision-making, and the resulting outcomes. In this sense, the study breaches the limit of conventional regulation, still largely stuck in two competing models – product restrictions and disclosure. This study was not only about product restrictions because the vast majority of financial products on the market are not inherently unreasonable. And it was also not just about disclosure – the regulation of which is largely based on the artefact of the homo economicus – because the European Commission found compelling evidence of sub-optimal decision-making. Instead, the study produced innovative considerations on the design and implementation of regulation, including features such as the framing of information or the provision of warnings. More recently, the European Commission collected behavioural evidence – through a

M4716-STRAßHEIM_9781785367847_t.indd 108

25/04/2019 13:32

The application of behavioural insights to policy in Europe  109 large-scale web experiment – in view of the revision of the Tobacco Products Directive (2014/40/EU). Although an increase in taxation was not expected to lead to a substantial drop in demand – as demand for tobacco products is knowingly inelastic, with estimates of price elasticity ranging from -0.2 and -0.6 (International Agency for Research on Cancer 2011) – information disclosure was also seen as a blunt solution, since past labelling rules were considered to be insufficiently effective. The labelling of tobacco products in Europe has been regulated since 1989, when health warnings covered at least 4 per cent of the surface of the packets of cigarettes. Larger warnings were introduced in 2001, to raise consumers’ awareness about the risk of smoking. Since then, cigarette packets sold in Europe had to contain a general health warning covering 30 to 35 per cent of the front surface (‘Smoking Kills/Smoking can kill’ or ‘Smoking seriously harms you and others around you‘) as well as one additional health warning, linked to specific health consequences of smoking, covering 40 to 50 per cent of the back of the packets of cigarettes. In 2013, before the review of the relevant legislation, the European Commission carried out a behavioural study to collect useful information. The behavioural nature of the study allowed the collection of precious evidence both on the salience of alternative health warnings, and the smoking intentions for both smokers and non-smokers. The study was expected to help identify suitable pictures to be associated with 14 smoking-related text warnings. Since May 2016, these combined text-plus-image health warnings, covering 65 per cent of the front and back of packs of cigarettes, are compulsorily displayed on all cigarette packets sold in European Member States. Besides the applications on consumer and competition policy, on financial services and tobacco, the European Commission recently collected behavioural evidence to inform policy-making in a number of other policy areas, including energy labelling, online gambling and food information. Moreover, BIs are currently being used to improve procurement procedures and increase the expected impact of fieldwork on gender inequality and communication campaigns around issues related to gender violence. While such developments are fully in line with evidence-based policy-making, it is yet too early to assess the impact of such work.

5. A TAXONOMY OF BEHAVIOURAL POLICY INTERVENTIONS Despite the recent academic rise in the application of BIs to policy-making, explicit policy applications are still relatively rare. However, as we claimed in a more extensive report (Lourenço et al. 2016), awareness about the behavioural dimension of current  policy interventions is perhaps hidden even to the very promoters of these interventions. Indeed, if researchers were to give account of behavioural policy applications strictu sensu only, they could not provide but a conservative picture. We adopt a less strict or more articulated definition of behavioural insights, and we classify behavioural policy initiatives in three groups. To avoid confusion between various types of behavioural policy initiatives, we propose a classification based on the degree to which behavioural considerations have helped shape the various initiatives (see Table 8.2).

M4716-STRAßHEIM_9781785367847_t.indd 109

25/04/2019 13:32

110  Handbook of behavioural change and public policy Table 8.2  Behavioural insights versus nudges Behaviourally-tested initiatives:

Initiatives being explicitly tested, or scaled out after an initial ad hoc experiment. Examples of these may include self-commitment strategies to address gamblers’ overconfidence and myopia and the resulting irresponsible online gambling; or social norm feedback in simplified and attractive letters to encourage timely tax compliance.

Behaviourally-informed initiatives:

Initiatives designed explicitly on previously existing behavioural evidence. An example could be the ban of pre-checked boxes for online purchases, introduced by a recent EU Directive (EU 2011).

Behaviourally-aligned initiatives:

Initiatives that, at least retrospectively, can be found to be aligned to behavioural evidence. Examples of these include the first national laws changing the default option for organ donations, or the penalty point systems for driving licences.

Our taxonomy includes three types of policy initiatives: ●●

●●

●●

Behaviourally-tested initiatives: these are initiatives being explicitly tested, or scaled out after an initial ad hoc experiment. At EU level, the EC Recommendation on Online Gambling (July 2014) – which advocates Member States to help players set self-commitment strategies – explicitly incorporates the results of a dedicated behavioural study (European Commission 2014). Behaviourally-informed initiatives: these are initiatives designed after an explicit review of previously existing behavioural evidence, although not benefiting from any specific prior ad hoc experiment. This was the case of the inclusion of a ban on pre-checked boxes in the Consumer Rights Directive (European Union 2011). Since the available evidence was considered compelling enough to support the policy initiative, the European Commission carried out no ad hoc trial to justify the inclusion of the ban. Behaviourally-aligned initiatives: these are initiatives where BIs can be identified retrospectively, although these initiatives do not explicitly rely on behavioural evidence, be it available literature or evidence coming from an ad hoc test. In this type of initiative a behavioural lever is usually used to tackle a behavioural driver, often complementing traditional forms of intervention (e.g. information provision, taxation). At the national level, the decremental penalty point system for driving licences (adopted in a number of European countries) is designed to leverage drivers’ loss aversion so to encourage the respect of the Highway Code, and increase road safety.

The classification above is instrumental to describing differences and stressing similarities between the various behavioural policy interventions under review. Behaviourally-tested initiatives, with clear exceptions, are still regrettably rare among behavioural policy initiatives. This calls for the adoption of a more systematic and transparent approach to the application of BIs to policy-making.

M4716-STRAßHEIM_9781785367847_t.indd 110

25/04/2019 13:32

The application of behavioural insights to policy in Europe  111

6. AN APPLICATION OF THE PROPOSED TAXONOMY TO RECENT POLICY INTERVENTIONS In the present section we outline a number of recent policy initiatives, which were not covered in the Behavioural Insights Applied to Policy: European Report 2016 (Lourenço et al. 2016). We grouped initiatives by policy area, and classified them using the proposed taxonomy of behavioural policy initiatives. Note that a particular focus is given to behaviourally-tested initiatives, and that a comprehensive collection of behavioural policy initiatives was beyond the scope of this chapter. Finally, we gathered information on specific initiatives mostly through desk research of published material (such as academic and policy papers) or, in the few not referenced cases, through personal exchanges. 6.1  Consumer Protection and Competition Behaviourally-tested initiative – Increasing the number of consumer disputes resolved without a court. The Danish Competition and Consumer Authority applied BIs in the implementation of the Alternative Dispute Resolution (ADR) Directive, which entered into force in July 2015. A RCT was run from April to June 2015 with a sample of 400 consumers to assess the effect of a behaviourally-informed form to increase the number of consumer disputes resolved without a court ruling. The form provided consumers with a structured overview of the complaint as well as information and guidance on the steps to take, thus using simplification to make it easier for consumers to reach an agreement. Furthermore, the form increased the salience of the complexity of legal procedures (through a realistic overview of the time and effort involved) and used framing to promote the use of ADR (‘most ADR cases are solved by the consumer without any need for the involvement of a court’). Results showed that the behaviourally-informed form reduced the number of cases resolved with litigation by almost 50 per cent (that is, from 37 per cent of the cases in the control group, to 19 per cent in the behaviourally-informed condition). Behaviourally-tested initiative – Investigating consumers’ capabilities with complex products. In Ireland, a series of lab experiments were carried out to examine how complex a product has to become before the consumers’ ability to accurately choose between good and the bad products is reduced. The results show that accuracy of consumer choices can be greatly reduced already for products which entail taking into account simply more than one or two attributes. This suggests that consumers may be incurring significant losses (and opportunity costs) due to products’ complexity. The study offers guidance regarding the number of key attributes that consumers are able to simultaneously trade off, and has thus relevance for the conception of standardized disclosures. In terms of potential remedies, the authors point to the importance of ‘independent and accurate consumer advice, which can be designed to highlight the key attributes of complex products that cannot be ignored without risking negative consequences’. They also highlight the potential value of independent price comparison sites and other ‘choice engines’ aimed at simplifying the information and helping consumers make informed choices, while also pointing to some of the limitations of the current tools (Lunn et al. 2016b).4

M4716-STRAßHEIM_9781785367847_t.indd 111

25/04/2019 13:32

112  Handbook of behavioural change and public policy 6.2  Energy and Environment Behaviourally-tested initiative – Increasing purchasing of more energy efficient appliances. A field experiment in Norway showed that using labels displaying the cost of the products’ lifetime energy operating cost (thus addressing present bias), together with relevant training of sales staff, increased the purchase of more energy efficient tumble driers by 4.9 per cent (no similar effect was found for fridge freezers though) (Kallbekken et al. 2013). In Switzerland, lifetime energy operating cost information is also made available to consumers through the TopTen.ch platform (Lourenço et al. 2016). Behaviourally-tested initiative – Decreasing energy consumption. In a recent field experiment, researchers examined the effect of feedback about students’ electricity and heating usage at a hall of residence in London, UK. Similar to the energy reports produced by Opower (Allcott 2011), social norms were used to encourage energy savings. Students were emailed weekly energy reports (from 3 June 2013 for 10 weeks) containing a comparison of their energy use to that of their neighbours (all/20 per cent most efficient), information on how they ranked in comparison to the 88 participants, and simple and easy to implement energy conservation tips. The results showed a reduction of over 20 per cent in overall energy consumption in the treatment group receiving the energy reports, relative to the control group (that is participants received a single email at the start of the experiment containing energy-saving tips). As pointed out by the authors, the effect is not driven by cost-saving incentives, as residents do not pay for their energy consumption. Interestingly, the experiment used a second treatment group where, in addition to weekly energy reports, participants were informed about a prize competition to reward the most energy-saving resident. This external reward seemed to have reduced or removed the intrinsic motivation to save energy, as the positive effect relative to the control group was lost after two weeks. The authors point to the importance of considering these findings in the context of energy management and efficiency policies, including smart meter interfaces (Alberts et al. 2016). 6.3  Finance and Taxation Behaviourally-tested initiative – Encouraging consumers to switch or negotiate insurance policies at renewal. In December 2015, the UK Financial Conduct Authority (FCA) published proposals to require general insurance providers to publish details of last year’s premium on renewal notices.5 The proposals aim to stimulate consumers to shop around for the best insurance product, thus maximizing the chances of getting better deals, and build on behavioural evidence showing low levels of consumer engagement, switching and a lack of competition at the time of renewal of home or motor insurance policies. Specifically, results from a RCT with over 300 000 consumers across the UK had showed that providing information about the last year’s premium on renewal notices – thus making this information salient – led to between 11 per cent and 18 per cent more costumers switching or negotiating lower premiums (Financial Conduct Authority 2015). Behaviourally-tested initiative – Improving response rates to revenue surveys. In Ireland, a RCT tested the effect of personalized notes on the responses rates to a regular revenue survey of small and medium-sized enterprises (SMEs). A sample of 2 000 SMEs taxpayers received the request to complete the survey either with a personalized handwritten post-it note attached to it (treatment group) or with no personalized note (control group).

M4716-STRAßHEIM_9781785367847_t.indd 112

25/04/2019 13:32

The application of behavioural insights to policy in Europe  113 This behavioural intervention led to higher response rates (36 per cent versus 19.2 per cent in the treatment and control groups after 15 days) and quicker responses (Central Expenditure Evaluation Unit 2014).6 Behaviourally-tested initiative – Increasing tax compliance. The Belgian Federal Ministry of Finance is currently testing, together with the University of Oxford, different ways of simplifying tax notices, sent to citizens, to increase fast tax compliance (Vermeersch and Benyaich 2015). The project also involves testing different nudges, such as including information about the social norm or penalties (De Neve 2015). Prior to this initiative explicitly using BIs, the Belgian Federal Ministry of Finance introduced pre-populated online income declarations as a means of simplification. Finally, since 2012, the Ministry went a step further by proposing simplified declarations with an opt-out scheme: unless citizens do not agree with the Ministry’s proposal of income declaration, the proposal is considered as accepted as such (NA 2014). This opt-out scheme is targeted to specific segments of taxpayers (2 million in 2016), mainly the retired and people with no or low income. The initiative was successful, as 96 per cent of taxpayers receiving this proposal did not opt out in 2015 (Mathieu 2016). 6.4 Health Behaviourally-tested initiative – Increasing physical activity of school children. In collaboration with Italian schools and Cambridge University, the Joint Research Centre carried out an experiment in 13 schools (350 participants aged between 10 and 11, over a period of seven weeks) to test the effect of social incentives in promoting physical activity. The aim was to help find ways of tackling child obesity. Findings showed that social incentives were on average 10 per cent more effective than individual incentives, suggesting that features of group behaviour like cooperation and altruism can be harnessed to improve overall group health (Proestakis et al., n.d.). Behaviourally-tested initiative – Reducing unnecessary prescriptions of antibiotics by general practitioners. The UK BIT, together with the UK Department of Health, ran an RCT to assess the effect of providing social norm feedback to general practitioners on the amount of prescribed antibiotics. General practitioners (GPs) were randomly assigned to the treatment group (GPs who received the social norm feedback showing their prescription rate were in the top quintile, or the control group (GPs who received no feedback). The social norm feedback had a significant effect in decreasing antibiotic prescriptions by approximately 3 per cent (Hallsworth et al. 2016). Behaviourally-tested initiative – Promoting medication adherence. Using a large-scale RCT, delivered through a UK pharmacy chain and encompassing 10 739 patients, researchers examined the effect of pre-commitments on adherence rates. Participants were assigned to either the control or one of three treatment groups: (a) signing a commitment sticker: ‘I commit to taking this medication exactly as prescribed’; (b) signing a commitment sticker paired with an external cost of low adherence: ‘The National Health Service loses £300 million per year from wasted medication. I want to do my bit to support the NHS, so I commit to taking this medication exactly as prescribed’; and (c) signing a commitment sticker paired with an internal cost: ‘Not taking my medication as prescribed could risk my health. I want to do all I can to improve my health, so I commit to taking this medication exactly as prescribed’. The experiment ran between July 2015 to March 2016

M4716-STRAßHEIM_9781785367847_t.indd 113

25/04/2019 13:32

114  Handbook of behavioural change and public policy and results showed that while signing a commitment sticker paired with an internal cost enhanced adherence by 4.17 per cent, pairing with an external cost decreased adherence by 5.32 per cent (NA 2015). 6.5  Social Welfare and Employment Behaviourally-tested initiative – Examining the effect of different social assistance approaches. The city of Utrecht (Netherlands), together with researchers from Utrecht University, is currently running an experiment to examine the effect of changes to the current social assistance rules. The experiment – designated ‘What Works’ (‘Weten wat werkt’ in Dutch) – will be carried out from June 2018 to October 2019. It will encompass a control group (no changes to the welfare benefits scheme) and three treatment groups, including: (a) ‘Getting in action by yourself’ (participants free to decide whether or not to look for paid or other work); (b) ‘Getting in action with extra help’ (receive extra help and guidance from the municipality); (c) ‘Work pays off’ participants who work while getting benefits can keep 50 per cent of this income up to a maximum of €202 per month. The aim is to examine whether less complex and/or strict benefits can lead to better results. The study will examine the effect on several outcomes, including whether welfare recipients find work more quickly and/or become more active in society, among others (City of Utrecht, n.d.). Behaviourally-tested initiative – Encouraging retirement savings by mailing letters: Assessing impact through analysis of a large dataset. Since 2004, the German pension authority sends out annual letters with detailed and comprehensible information about the pension system and individual expected pension payments. The letters also inform about the growing pension gap (as future pensions should grow at a lower rate than wages), stresses the importance of increasing savings through private retirement accounts and points to the need of accounting for the loss in purchasing power at an older age. Using a 5 per cent stratified random sample of the German Taxpayer Panel – an administrative dataset collected by German tax authorities in the period 2001 to 2010, and managed by the German Federal Statistical Office – researchers at the Center for European Economic Research recently examined the effect of the letters on retirement savings. Results showed that the letters increased private retirement savings persistently over several years, though partly reducing charitable donations. The analysis of the results revealed that the effect could be attributed both to increased knowledge and heightened salience of expected pension payments. Moreover, it is worth noting the use of personalization as, besides the general information, individuals were also clearly provided with their own future pension entitlements. Finland, France, Sweden and the US seem to have in place similar letters (Dolls et al. 2016). Behaviourally-aligned initiative – Improving matching of skills and jobs. A project in Austria is aimed at improving job matching through a skill-matching engine, rather than through job titles. The objective is to better respond to the demands of an increasingly diversified and specialized job market. A job vacancy app is also available (downloaded by 40 000 people by winter 2015) to simplify the process and make it easier for job seekers to find vacancies (e.g. anyone can save searches and get updates about new job vacancies in one’s region) (Keegan 2015). Behaviourally-tested initiative – Helping unemployed getting back into work. Results

M4716-STRAßHEIM_9781785367847_t.indd 114

25/04/2019 13:32

The application of behavioural insights to policy in Europe  115 from a Randomized Controlled Trial at a Jobcentre in Essex (UK BIT, in collaboration with the Department for Work and Pensions and Jobcentre Plus), showed that when advisors asked individuals to make specific commitments to job-seeking activities in the following week (making the process more personalized and social, by getting individuals to make public pre-commitments) there was a five percentage point increase in getting back to work for the group receiving the intervention relative to the control; the intervention has been scaled up to all Jobcentres in the UK (Lourenço et al. 2016).

7. THE FUTURE OF BEHAVIOURAL POLICY-MAKING IN EUROPE There is growing evidence on the impact of behavioural sciences in policy-making. This is particularly the case at country level, with some evidence of successful behavioural initiatives at supra-national and at local level. In the future, there is a need to look at the interventions that work at the micro-level, so scale them up at a country level, and possibly even across countries. More information is also needed on the long-term effects of policy interventions that have a behavioural approach. Such knowledge will contribute to increasing transparency in the use of BIs and will reveal which behavioural interventions are more effective in achieving long-lasting results. In the future, BIs can also be applied to a wider range of domains, such as risky consumption behaviours (for example, alcohol consumption, smoking) and addictive behaviours (for example, social media, sugar intake), to explore whether nudging could help consumers take decisions more in line with their underlying preferences, contextually avoiding societal costs, too. This would lead to better understanding how consumers deal with risk, uncertainty and time delay. Furthermore, behavioural sciences can also be used to tackle more complex issues, such as social isolation, migration, fairness, stereotypes, racism – these problems have proved to be very difficult to solve with a business as usual approach. Beyond the different methodological approaches one may use, the incorporation of BIs in policy-making implies a paradigm shift in the way the public administration works. Encouraging a culture of experimentation in policy-making will avoid an approach where everything is fixed from the outset. More, experimentation will bring about evidencebased policy and improve economic models of decision-making. As recommended by the late professor Roger Jowell, what is needed in policy-making is ‘a spirit of experimentation, unburdened by promises of success’ (Jowell 2003, p. 10). Even with the right amount of political buy-in, tackling more complicated issues, scaling-up interventions and having more knowledge about long-term effects of behavioural interventions require capacity building, cross-fertilization between academia and policy-making and access to data. In what concerns capacity building, there is still limited awareness about which policy interventions would benefit from the incorporation of a behavioural approach (Lourenço et al. 2016). Training public officials, including those in management positions, to apply BIs will hopefully help them make better decisions, and identify specific resources and tools that can be used to support any behavioural intervention. Furthermore, policymakers themselves might be subject to biases in their own decision-making (Jowell 2003). In this regard, unveiling potential biases is an important first step (Brest 2013). In this

M4716-STRAßHEIM_9781785367847_t.indd 115

25/04/2019 13:32

116  Handbook of behavioural change and public policy respect, the World Bank (World Bank 2015) carried out trials showing its own staff is subject to sunk cost bias (that is, the tendency of individuals to continue a project once an initial investment of resources has been made). The World Bank’s trials, run with public officials, follow precursory cases of experiments designed to raise awareness of policy-makers’ susceptibility to biases, similar to the general population. Indeed, in 2008 the European Commission ran a series of experiments on framing, choice overload and constructed preferences (Ciriolo and Oullier, n.d.). Then, in 2012 the OECD tested the effects of defaults on thermostat settings with OECD employees (Brown et al. 2012). Finally, in 2014, the European Commission carried out a before-and-after study on the impact of social norms and choice architecture in encouraging officials to use the stairs instead of the lifts (Biard et al. n.d.). As previously mentioned, such trials contribute to show that policy-makers are not exempted from biases, and indirectly make them more prone to take biases into account when designing policy solutions. Besides awareness-raising, however, there is obviously the need to build capacity and knowledge around a relatively new approach to policymaking. For this reason, since May 2017, the European Commission has been training public officials in behavioural sciences. Cross-fertilization between academic research and public administration can bring mutually beneficial results. First, academics have the expertise in behavioural sciences that is usually lacking in the public administration. This expertise can be used to enhance evidence-based policy-making by taking advantage of the methodologies and behavioural theories used in academic research to solve real life issues. Second, academics struggle to have access to large and high-quality datasets. The analysis and merging of existing large datasets and the design of initiatives with embedded outcome measures can bring valuable insights for policy-makers.

8.  DISCUSSION AND CONCLUDING REMARKS In the previous sections we reviewed major contributions of behavioural sciences to policy-making and gave account of a number of recent developments in the application of BIs to policy. Even though some constraints remain (e.g. need for capacity building, more systematic use of the behavioural approach, evaluation of policy impacts, etc.), it is clear that this is a fast-developing area. In the last few years we have witnessed increasing interest and significant dynamism in the application of BIs to policy-making in Europe, of which institutional developments – seen most notably through the creation of a number of governmental teams – are an example. Equally interesting, however, is the fact that BIs are increasingly spreading from academia to policy-making. Namely, academic experts are increasingly contributing to shaping behavioural policies through participation in advisory panels of behavioural teams at government level, through collaborations with governmental ministries/­departments in specific projects, or through applied research tackling real policy problems (several examples were presented above). Collaborations with behavioural scientists in universities/ research institutions should continue to be strengthened as they can support governments in applying new thinking and rolling-out policies which are informed by greater evidence. In particular, academics can bring in-depth knowledge of behavioural sciences (as well as

M4716-STRAßHEIM_9781785367847_t.indd 116

25/04/2019 13:32

The application of behavioural insights to policy in Europe  117 fresh knowledge about the most recent theoretical advances in the field), methodological expertise, the technical skills needed to analyse large existing datasets of systematically collected behaviourally-relevant data and immediate access to research facilities. Even when one of these elements is present in public administration, it is rare that all of them are met. Furthermore, while expertise may exist within a given governmental department, restrictions in terms of human resources are often observed. Additionally, the novelty of the behavioural approach and the relative lack of capacity at the government level in most European countries should not be overlooked. Thus, perhaps not surprisingly, the ability to recognize which policy interventions would benefit from the incorporation of a behavioural approach is not generalized across public officials (Lourenço et al. 2016). Providing training in behavioural sciences to public officials, including those in management positions, will contribute to more systematic application of BIs and support more efficient decision-making (starting already with the understanding of how policy-makers themselves can be subject to biases). For instance, capacity building across different policy departments should improve policy-makers’ understanding of how BIs can complement traditional tools by helping to fine-tune policy measures – ­including through testing – and how behavioural biases might impact the potential success of a policy. Furthermore, training should improve their ability to pro-actively/independently identify entry points for BIs at the early stages of policy development. Relatedly, there are instances where regulatory policy is designed in a way that does not anticipate implementation and/or enforcement issues. BIs and testing can help to overcome these hurdles which risk jeopardizing the effectiveness of policy initiatives. Critically, this requires that BIs are taken into account early on in the policy process and not simply regarded as ‘add-ons‘ to be taken into account at later stages. At the beginning of the policy process, BIs help identify the root causes of the problem matter, looking beyond the mere utilitarian approach whereby people take decisions maximizing their welfare, but complementing this view with a rather systematic analysis of co-determinants of individual behaviour. In the BIAP 2016 Report, we advocated for expanding the ‘Test, Learn, Adapt’ approach (UK BIT), by adding a fourth sequential principle, ‘Share’. Sharing can lead to more robust behavioural policy initiatives, built with a greater understanding of what works, and under which conditions (e.g. cultural, geographic, of specific cohorts) (Lourenço et al. 2016). Both the present chapter and BIAP 2016 aim to directly contribute to this. They present a taxonomy of behavioural policy initiatives together with a collection and analysis of a large number of behavioural policy initiatives across Europe. This collection will hopefully provide inspiration and inform future (and more robust) policy applications. Interestingly, we see already examples of this taking place. For instance, as reported in section 6, the project targeted at job centres in Portugal was inspired by the earlier RCT using commitment devices and ran by the UK BIT at a Jobcentre in Essex. Critically, the project in Portugal did not jump straight to implementing commitment devices (or, in other words, simply copy the UK intervention), but started instead by carrying out an assessment of the reality on the ground. This is linked with a well-established feature. That is, context matters, such that interventions that can be effective with one target population may not be with the other – hence, one needs to map the reality on the ground as well as pre-test initiatives in order to evaluate their effectiveness before deploying them.

M4716-STRAßHEIM_9781785367847_t.indd 117

25/04/2019 13:32

118  Handbook of behavioural change and public policy Sharing should contribute to avoiding pitfalls and support the conception of increasingly sound behavioural research for policy-making. One can even aim to be more ambitious, and strive to make good practices more systematic. The pre-registration of experimental protocols in The American Economic Association’s registry for randomized controlled trials is an excellent example. In addition, there is much to be gained from additional research into the long-term impacts of behaviourally-informed policy interventions, since this type of research is still residual. Opower’s energy reports tapping on social norms to reduce household energy consumption and showing how the effects are fairly persistent even when reports are discontinued, represent an excellent example (Allcott 2011; Allcott and Rogers 2014). Notwithstanding a possible increasing political buy-in, this aspect is expected to become even more of a concern in the future. Furthermore, one needs to be aware that there may be behavioural spill-overs (e.g. doing exercise to lose weight, but ‘permitting’ oneself to eat less healthy food after each run) (Dolan and Galizzi 2015). Thus, as much as possible, practitioners should adopt a broader perspective, trying to determine whether these occur and the extent to which they enhance, or rather jeopardize, the effects of the main policy intervention. Finally, the analysis of large existing datasets of systematically collected behaviourallyrelevant data holds great potential in terms of uncovering novel and useful insights for policy. Despite the massive amount of physical data currently recorded, and the current large-scale data-processing technologies, behavioural experts often struggle to gain access to such wealth of information. Policy-makers can have a role to play here, not only by facilitating access to usable data, but also, as remarked by Costa and King (2016), ‘regulators should use existing forms of aggregated data, such as complaints and consumer reviews, to develop a more complete understanding of markets, identify where behavioural (and traditional) market failures may be occurring, and address problem areas before they escalate’. Before ‘big data’, there is a preliminary ‘data revolution’ that has never taken place, the one that creates maximum value from publicly-available datasets, and from merged versions of the latter. Overall, we observed substantial developments to prove that there is substance behind the would-be hype of a trendy new scientific discipline and policy-making approach. If impact assessments of current and forthcoming behavioural policy initiatives will be carried out in a robust way, time will confirm whether BIs lead to better policy-making, better regulation and better spending. Preliminary evidence seems to suggest so.

NOTES 1. This chapter revisits the taxonomy of behavioural interventions that was put forward in BIAP 2016 and applies it to a number of recent European policies. These initiatives, not covered by BIAP, were gathered mostly through desk research. Only a subset of this is presented due to space restrictions, but the broader number of initiatives collected will be included in a future update of the BIAP Country Overviews. It is worth pointing out that BIAP 2016 drew on data collected via a mix of desk research, survey reaching out to around 900 potential respondents (27 per cent of whom replied) and personal exchanges, including interviews. Desk research was based on online resources (that is available reports and scientific papers) as well as web searches using a series of keywords (Lourenço et al. 2016). BIAP 2016 was accompanied by a set of Country Overviews, which can be downloaded online. Country Overviews give account of relevant behavioural policy initiatives, of pertinent institutional developments, and provide useful evidence on the

M4716-STRAßHEIM_9781785367847_t.indd 118

25/04/2019 13:32

The application of behavioural insights to policy in Europe  119

2.

3. 4. 5. 6.

existence of national resources and capacity in this field, individually, for each of the 32 countries covered by BIAP. A randomized controlled trial (RCT) is an experiment that randomly allocates participants to a control group or to one or more treatment groups. It allows the identification of a cause-effect relationship between a specific feature (e.g. mailing a personalized Home Energy Report letter) and its impact, while controlling other features by design. This is so much the case that, once the law entered into force, drivers in some countries started to circumventing local seatbelt laws by purchasing and wearing special self-deceptive T-shirts – featuring a stripe that runs from shoulder to hip – to try and cheat on policy controls. Relatedly, for research investigating consumer choice in the complex market for personal loans see Lunn et al. (2016a). In the UK, the FCA has been at the forefront of using BIs to protect financial consumers and inform regulation. The FCA has conducted pioneering behavioural work and has published a number of behavioural papers (in particular, Occasional Papers 1, 2, 3, 7, 9, 10 and 12). Other projects implemented by the Revenue Commissioners using BIs include the use of behaviourallyinformed letters to increase tax compliance and applying framing to examine the effect of ‘soft’ and ‘hard’ wording in letters about outstanding tax debt on payments. In both cases, the use of BIs led to improved outcomes (for details please see Central Expenditure Evaluation Unit 2014).

REFERENCES Alberts, G., Z. Gurguc, P. Koutroumpis, R. Martin, M. Muûls and T. Napp (2016), ‘Competition and norms: a self-defeating combination?’, Energy Policy, 96, 504–523. Allcott, H. (2011), ‘Social norms and energy conservation’, Journal of Public Economics, 95(9), 1082–1095. Allcott, H. and T. Rogers (2014), ‘The short-run and long-run effects of behavioral interventions: experimental evidence from energy conservation’, The American Economic Review, 104 (10), 3003–3037. Biard, A., E. Ciriolo, E. and C. Salinier (n.d.), Lifts vs. stairs: a before-and-after behavioural study. Manuscript in preparation. Brest, P. (2013), ‘Quis custodiet ipsos custodes? Debiasing the policy makers themselves’, in The Behavioural Foundations of Public Policy, Eldar Shafir, ed., Princeton: Princeton University Press, 481‒493. Brown, Z., N. Johnstone, I. Haščič, L. Vong and F. Barascud (2012), ‘Testing the effect of defaults on the thermostat settings of OECD employees’, OECD Environment Working Paper, N° 51. Célérier, C. and B. Vallée (2013, July), ‘What drives financial complexity? A look into the retail market for structured products’, in A Look into the Retail Market for Structured Products, Paris December 2012 Finance Meeting EUROFIDAI-AFFI Paper. Central Expenditure Evaluation Unit (2014), ‘Behavioural Economics’, accessed 12 January 2018 at http://igees. gov.ie/wp-content/uploads/2013/10/Behavioural-Economics-1.pdf. Ciriolo, E. (2011), ‘Behavioural economics in the European Commission: past, present and future’, Oxera Agenda, January 2011. Ciriolo, E. (2016), ‘Do behavioural insights matter for competition policy’, CPI Europe Column, 15 July 2016. Ciriolo, E. and O. Oullier (n.d.), ‘Neither cats nor dogs: and if consumers are rather “gulls”’?, Unpublished paper. City of Utrecht (n.d.), ‘What works: studying the effects of fewer rules in social assistance’, City of Utrecht, accessed 8 January 2019 at https://www.utrecht.nl/city-of-utrecht/study-on-rules-in-social-assistance/. Costa E. and K. King, ‘Applying behavioural insights to regulated markets’, 26 May 2016, The Behavioural Insights Team, accessed: 8 January 2019 at https://www.behaviouralinsights.co.uk/consumer-affairs/applyingbehavioural-in​sights-to-regulated-markets/. Coyle, D. (2016), ‘Digital newcomers puzzle competition authorities’, Financial Times, 9 March 2016. De Neve, J.-E. (2015), ‘Behavioral insights and tax compliance: experimental evidence from Belgium’, accessed 30 September 2016 at https://www.socialscienceregistry.org/trials/827. Department of Public Governance and Chancellery (2016), ‘Uitvoeren van onderzoeken om inzicht te verkrijgen in drempels en hefbomen om effectiviteit van beleid en communicatie te versterken’, accessed 3 October 2016 at https://overheid.vlaanderen.be/raamcontract-onderzoek-gedragsinzichten. Dolan, P. and M.M. Galizzi (2015), Like ripples on a pond: behavioral spillovers and their implications for research and policy’, Journal of Economic Psychology, 47(January), 1–16. Dolls, M., P. Doerrenberg, A.Peichl and H. Stichnoth (2016), ‘Do savings increase in response to salient information about retirement and expected pensions?’, ZEW Discussion Paper No 16-059.

M4716-STRAßHEIM_9781785367847_t.indd 119

25/04/2019 13:32

120  Handbook of behavioural change and public policy European Commission (2006), ‘Road safety: new European rules on seat belt use’, accessed 30 September 2016 at http://europa.eu/rapid/press-release_IP-06-583_en.htm. European Commission (2014), Commission Recommendation of 14 July 2014 on principles for the protection of consumers and players of online gambling services and for the prevention of minors from gambling online, 2014/478/EU. European Union (EU) (2011), Directive on Consumer Rights, N. 2011/83/EC. Financial Conduct Authority (2015), FCA to require insurance firms to publish details of last year’s premium, accessed 30 September 2016 at https://www.fca.org.uk/news/press-releases/fca-require-insurance-firms-publishdetai​ls-last-year%E2%80%99s-premium. Fryer Jr, R.G., S.D. Levitt, J. List and S. Sadoff (2012), Enhancing the efficacy of teacher incentives through loss aversion: A field experiment (No. w18237), National Bureau of Economic Research. Goldstone, R. L. and Lupyan, G. (2016), Discovering psychological principles by mining naturally occurring data sets. Topics in Cognitive Science, 8, 548–568. doi:10.1111/tops.12212. Hallsworth, M., T. Chadborn, A. Sallis, M. Sanders, D. Berry, F. Greaves, F. and S.C. Davies (2016), ‘Provision of social norm feedback to high prescribers of antibiotics in general practice: a pragmatic national randomised controlled trial’, The Lancet, 387(10029), 1743–1752. Haynes, L., B. Goldacre and D. Torgerson (2012), ‘Test, learn, adapt: developing public policy with randomised controlled trials’, Cabinet Office-Behavioural Insights Team. International Agency for Research on Cancer (2011), Tax, price and aggregate demand for tobacco products. Effectiveness of tax and price policies for tobacco control, Lyon, France, accessed 30 September 2016 at http:// www.iarc.fr/en/publications/list/handbooks/. Johnson, E.J. and D. Goldstein (2003), ‘Do defaults save lives?’, Science, 302(5649), 1338‒1339. Jowell, R. (2003), ‘Trying it out: the role of “pilots” in policy-making: report of a review of government pilots’, Cabinet Office, Strategy Unit, accessed 10 January 2019 at https://assets.publishing.service.gov.uk/govern​ ment/uploads/system/uploads/attachment_data/file/498256/Trying_it_out_the_role_of_pilots_in_policy.pdf. Kahneman, D. and A. Tversky (1973), ‘On the psychology of prediction’, Psychological Review, 80(4), 237. Kallbekken, S., H. Sælen and E.A. Hermansen (2013), ‘Bridging the energy efficiency gap: a field experiment on lifetime energy costs and household appliances’, Journal of Consumer Policy, 36(1), 1–16. Kay, J. (2011), ‘The $10 minibar beer is no basis for capitalism’, Financial Times, 11 July 2011. Keegan, M.J. (2015), ‘Preventing and reducing unemployment: insights from Dr. Johannes Kopf, Managing Director, Austrian Public Employment Service’, IBM Center for The Business of Government, accessed 30 September 2016 at http://bizgovmagazine.org/articles/preventing-and-reducing-unemployment-insightsfrom-dr-johannes-kopf-managing-director-austrian-public-employment-service/. Lourenço, J.S., E. Ciriolo, S. Rafael Almeida and X. Troussard (2016), Behavioural Insights Applied to Policy: European Report 2016, European Union. Lunn, P. (2014), ‘Regulatory policy and behavioural economics’, OECD Report. Lunn, P., M. Bohacek and A. Rybicki (2016a), An Experimental Investigation of Personal Loan Choices, accessed 10 January 2019 at https://www.esri.ie/pubs/BKMNEXT314.pdf. Lunn, P., M. Bohacek, J. Somerville, A. Ni Choisdealbha and F. McGowan (2016b), PRICE Lab: An Investigation of Consumers’ Capabilities with Complex Products, accessed 30 September 2016 at https://www.esri.ie/pubs/ BKMNEXT306.pdf. Mathieu, F. (2016), ‘Voici les nouveautés de la déclaration d’impôts 2016’, Le Soir, 27 April 2016. NA. (2014), Rapport annuel 2014, accessed 30 September 2016 at http://www.jaarverslag.financien.belgium.be/ fr/115-impots-sur-les-revenus-propositions-de-declaration-simplifiee-pds-lipp. NA. (2015), 15th TIBER Symposium on Psychology and Economics, accessed 30 September 2016 at https://www. tilburguniversity.edu/upload/89ff0c22-e3a7-4834-80da-9cb201c1e84a_Tiber 15 Program.pdf. NA. (2016), CRESA, accessed 5 October 2016 at http://www.cresa.eu/chi-siamo/matteo-motterlini/. OECD (2017), Behavioural Insights and Public Policy, Lessons from Around the World, Paris: OECD. OFT (2010), What does behavioural economics mean for competition policy?, OFT 1224. Oltermann, P. (2016), ‘State handouts for all? Europe set to pilot universal basic incomes’, The Guardian, 2 June 2016. Pareto, V. (2014), Manual of Political Economy: A Critical And Variorum Edition, reprint edited by Aldo Montesano, Alberto Zanni, Luigino Bruni, John S. Chipman and Michael McLure. Oxford, UK: Oxford University Press (Orig. pub. 1906). Proestakis, A., E. Polizzi di Sorrentino, H. Brown, E. van Sluijs, A. Mani, S. Caldeira and B. Herrmann (n.d.), Peer-Active Social Based Incentives for Increasing Physical Activity: A School Based Field Experiment, Manuscript in preparation. Sunstein, C. and R. Thaler (2008), Improving Decisions About Health, Wealth, and Happiness, Yale: Yale University Press. Svenson, O. (1981), ‘Are we all less risky and more skilful than our fellow drivers?’, Acta Psychologica, 47, 143‒148.

M4716-STRAßHEIM_9781785367847_t.indd 120

25/04/2019 13:32

The application of behavioural insights to policy in Europe  121 Thaler, R.H. (2016a), Misbehaving, The Making of Behavioural Economics, New York: W.W. Norton & Company. Thaler, R.H. (2016b), ‘Behavioral economics: past, present and future’ Present and Future, 27 May 2016. The Behavioural Insights Team (2016), The Behavioural Insights Team Update Report 2015–16, accessed 30 November 2016 at http://38r8om2xjhhl25mw24492dir.wpengine.netdna-cdn.com/wp-content/uploads/2016/09/ BIT_Update_Report_2015-16-.pdf. Vermeersch, W. and B.Benyaich (2015), Creër een federale en Vlaams Nudge Unit. Interview met Jan-Emmanuel De Neve (Gedragseconoom Oxford), Stichting Gerrit Kreveld, 22 January 2015. World Bank (2015), World Development report 2015: Mind, Society, and Behavior, Washington, DC: World Bank.

M4716-STRAßHEIM_9781785367847_t.indd 121

25/04/2019 13:32

9.  Behaviour experts in government: from newcomers to professionals? Joram Feitsma and Thomas Schillemans

1. INTRODUCTION Bureaucrats have always been in the business of behaviour change. Nevertheless, now that behavioural science has become more popular through bestsellers like Nudge (Thaler and Sunstein 2008), that business is becoming more and more elaborate, explicit, and scientific. Somewhat of a ‘behavioural turn’ has been invoked within a number of governments in the last decade with bureaucrats turning to insights from the behavioural sciences, behavioural economics in particular, to upgrade their policies. Put briefly, these ‘behavioural insights’ (BIs) challenge the still dominant model of homo economicus: the self-interested utility maximizer with unlimited information and self-discipline. Instead, a different model of human rationality is put forward, which is fundamentally bounded (Simon 1957) and shaped by fast decision-making heuristics (Tversky and Kahneman 1974). When governments acknowledge and tap into the bounded rationality of humans, their policies are likely to become more effective, efficient, and better underpinned. There are many different ways to study the behavioural turn. Scholars have studied, inter alia, the effects (e.g. Sunstein 2014), ethics (e.g. Bovens 2008), and legal aspects (e.g. Alemanno and Sibony 2015) of behavioural policies. This chapter, however, takes a different angle. It starts from the observation that the recent behavioural turn has not only introduced new ideas (e.g. heuristics and biases), new language (e.g. ‘choice architecture’) and new tools (e.g. ‘nudges’); it has also changed the occupational landscape of government. The chapter studies these changes, and explores the kinds of practices that have emerged to institutionalize BIs. (see Lunn 2014; John et al. 2009; World Bank 2015). It examines the behavioural turn in terms of a new profession entering the policy process: the ‘Behavioural Insights Teams’, ‘Behaviour Changers’, ‘Nudge Experts’, ‘Behavioural Insights Officers’, ‘Choice Architects’, and so forth, which are dubbed ‘behaviour experts’. Their emergence raises basic questions – who are they?, what do they do?, how are they developing? – but also deeper ones – what do they add?, how established are they?, are they becoming a fully-fledged profession? This chapter answers the basic questions and further explores the deeper questions within the context of Dutch government. It describes how Dutch behaviour experts are developing their new practice. The key issue revolves around the uniformity or diversity of this new occupation. Are behaviour experts developing into a distinctive, cohesive and exclusive profession in the policy process or is the development of this new occupation more scattered and diverse? The chapter will show that, although the profession is still in its infancy and to some extent homogeneous, it is currently more fragmented than cohesive. Behaviour experts share a small common core, but beyond that they rely on different theories, epistemologies, and tools.

122

M4716-STRAßHEIM_9781785367847_t.indd 122

25/04/2019 13:32

Behaviour experts in government  123 1.1 Professionalization To describe how the occupation of the behaviour experts is developing, we draw upon theories of professionalism (Abbott 1988; Freidson 2001), and in particular of professionalization (Larson 1977). With professionalization we refer to the process in which occupations manage to mobilize themselves, become established, form an exclusive and cohesive identity, and protect themselves against external pressures. The further they go in this process, that is, the more powerful, viable and sustainable they get, the more they become professions. Professionalization has traditionally only been relevant for pure professionals, such as doctors, engineers, and lawyers. However, as part of a new kind of professionalism – one that is more hybrid and less institutionalized and content-focused (Noordegraaf 2007) – many other occupations have since positioned themselves on the road towards professionalization; for instance, police agents (van der Meulen 2009), activation professionals (Schonewille 2015), and strategists (Noordegraaf et al. 2014). In this chapter, we interpret the development of behaviour experts as a case of early professionalization. However, we do not claim that behaviour experts are on the road towards full professionalization comparable to the pure professions. Neither do we make the claim that they should professionalize to that degree, that the process is irrevocable, nor that they understand their own development in those terms. Our aim is not to follow up on Wilensky’s (1964) question about whether we are experiencing ’the professionalization of everyone’. Rather, and more modestly, we take professionalization as a theoretical lens that is useful because it gives structure to the ways in which occupations establish themselves, organize themselves internally and liaise externally. It allows us to describe and understand how the field of behavioural expertise evolves. Professionalization is concerned with how occupations arise and manage to persevere. There are particular professionalization mechanisms which are at the bottom of this process. Building upon the categorization of Noordegraaf et al. (2014; see Thorstendahl 1990), we identify four of those mechanisms: knowledge, standards, positions, and identities. The knowledge mechanism refers to how professionals regulate their work through building cognition (Thorstendahl 1990), content (Noordegraaf 2007), or a body of knowledge and skills (Freidson 2001). This mechanism concerns how they build and demarcate an exclusive body of knowledge, supported by specific schools, training courses, professional journals and knowledge platforms. The standards mechanism refers to how professionals regulate their work through creating and regulating standards about competency, ethics, quality and instrumentation (Thorstendahl 1990). The positions mechanism refers to how professionals regulate their work on a political level (Faulconbridge and Muzio 2012) through claiming autonomy and legitimacy (Freidson 2001). It involves professionals developing strategies to deepen their links to the organization and justify their exclusive position. The identities mechanism refers to how professionals regulate their work through building a shared social identity (Thorstendahl 1990). This mechanism addresses how professionals perceive their own roles in relation to the wider context. It points to the particular heroes, model practitioners, and classic reads that help to define a profession’s collective identity. These mechanisms point at both an internal logic, geared towards finding internal closure and coherence, and an external logic, geared towards positioning oneself vis-à-vis others and other occupations. To professionalize, according to our framework, means to build and legitimize occupational boundaries, within which one develops one’s own

M4716-STRAßHEIM_9781785367847_t.indd 123

25/04/2019 13:32

124  Handbook of behavioural change and public policy Table 9.1  A framework to qualify professionalization projects Knowledge

Standards

Positions

Identities

Education and training Body of knowledge Knowledge platforms

Required competencies Quality standards Ethical codes Standardized instruments

Legitimacy Autonomy

Heroes and role models Role identities

Source:  Based on Noordegraaf et al. (2014) and Freidson (2001).

knowledge, standards, positions, and identities. The pure professions have advanced these mechanisms most consistently and elaborately, yet any existing occupation needs at least some level of shared knowledge, standards, positions, and identities. Together, these four mechanisms form the analytical lens – displayed in Table 9.1 – with which we observe salient aspects in the upcoming field of the behaviour experts.

2. METHODS Our study on the development of behaviour experts is based on desk research, a series of ten preliminary interviews (with behaviour experts and academic experts), and then a series of 24 half-open interviews with 35 behaviour experts in Dutch government over the course of 11 months (April 2015 until February 2016). Interviewing behaviour experts helped us to acquire ‘thick descriptions’ (Geertz 1973) of their work and professionalization activities. During the half-open interviews, we were guided by a set of sensitizing topics (including professional background, organization, practices, successes, challenges, and developmental processes). The interviews were recorded, selectively transcribed and turned into individual field reports, structured along our professionalization framework (i.e. knowledge, standards, positions, and identities). In selecting our respondents, we looked for self-proclaimed behaviour experts who use BIs on a structural and explicit basis. The latter requirement of explicitness means that BIs directly inform our interviewees’ practices and not just implicitly or in retrospect – which Lourenço et. al. (2016) describe as ‘behaviourally-informed’ and not ‘behaviourally-aligned’ (see also Ciriolo et al. Chapter 8 in this volume). To map the presence of these behaviour experts in Dutch government, we started with desk research. A few studies were useful here (e.g. Dorren 2015). We also scanned online content (e.g. reports, LinkedIn profiles, websites) searching for a broad range of relevant keywords such as ‘nudge’, ‘Behavioural Insights Teams’, and ‘applying behavioural insights’. Being flexible about these terminologies was important as the jargon of behaviour experts tends to differ and change, even though their practices are alike. Besides doing desk research, we asked respondents about their knowledge of other behaviour experts in the field. This snowballing technique provided us with more relevant sources than simple searches could accomplish, as many pioneering behaviour experts do not have clear public profiles. We also kept a list of active behavioural units that we verified with both our respondents and an online community within government that shared knowledge about the use of BIs.

M4716-STRAßHEIM_9781785367847_t.indd 124

25/04/2019 13:32

Behaviour experts in government  125 All of the behaviour experts we interviewed work in Dutch central government. We found most BI-related activity in this locus, with behaviour experts coming from nearly all Dutch ministries and many aligned agencies. The 35 behaviour experts we interviewed are already part of twenty different governmental organizations. Our locus of research contrasts with most studies on behavioural policy-making, which tend to focus on the frontrunners in the field: resourceful units, often part of an Anglo-Saxon policy culture, like ‘BIT UK’ (e.g. John 2014). However, beyond these frontrunners there is a broader field of emerging behavioural practices ‘all over the world’ (Whitehead et al. 2014; see also Whitehead et al., Chapter 7 in this volume). If we are interested in how this field is developing, in all of its variety, it is important to study these less outspoken cases as well. The Dutch government is such an underexplored case (for exceptions, see Lourenço et al. 2016; Stinesen and Renes 2014), featuring a broad palette of explorative behavioural policy practices.

3. FINDINGS 3.1  Background: the Rise of the Dutch Behaviour Expert At face value, BIs seem to have sparked the interest of the Dutch government. More and more initiatives are being launched to apply behavioural expertise to the policy process. These are emerging across agencies, at different stages of policy-making, as the 35 interviewed behaviour experts work both in central government (ministerial departments) as well as in agencies further removed from the centre (responsible for knowledge distribution, implementation, regulation, and enforcement). Most behavioural practices have been established in recent years, rendering behavioural expertise still in an explorative and dynamic phase of development. The publication of Thaler and Sunstein’s Nudge in 2008 was an important take-off point. When ‘Nudge’ entered the Dutch policy discourse, officials were increasingly being informed about the value of BIs for policy-making. This learning process has been fuelled by reports of the influential scientific council for government policy (e.g. WRR 2014). These reports were followed by a string of reports from other official think tanks and advisory bodies, culminating in a memorandum to Parliament (Ministry of Economic Affairs 2014). About thirty interviewees have started their practice only after ‘Nudge’ was introduced in Dutch government, somewhere between 2009 and 2015. During that period, seven interviewees were specifically hired by agencies because of their behavioural scientific background, which confirms the increased demand for behavioural expertise at that time. Only six interviewees worked with BIs before ‘Nudge’, about ten to fifteen years ago, when behavioural science was already blossoming but remained largely outside the awareness of policy-makers. Most behavioural policy practices have not been deeply institutionalized. They are organized ad hoc, bottom-up, with limited resources and feeble links to existing policymaking processes and institutional actors. Their development depends on the ambition of individual enthusiasts and efforts of low-key group collaborations, using BIs only sporadically or in spare time. Therefore, the positions of behaviour experts inside their organization tend to be fragile. They need to ‘survive’ in their organizations, challenged

M4716-STRAßHEIM_9781785367847_t.indd 125

25/04/2019 13:32

126  Handbook of behavioural change and public policy by cutbacks, reorganizations, disbelievers, and the burden of having to prove their added value. The most typical organizational role model of behavioural policy-making is the Behavioural Insights Team (BIT): a specialized government unit that applies BIs fulltime, inspired by the original ‘BIT UK’ which was launched in 2010 in British government. At the point of writing and researching, based on a snapshot of the field from early 2016, we discovered five BITs in Dutch central government. These are the ‘Team Gedragsverandering’ [Team Behaviour Change] in the Dutch Tax and Customs Administration, a BIT in the Ministry of Economic Affairs and Climate Policy, a BIT in the Ministry of Infrastructure and Water Management, a BIT in the Dutch Healthcare Authority (NZa) and a BIT in The Netherlands Authority for Consumers and Markets (ACM). These BITs are all relatively small, with up to five members, and new, being founded after ‘Nudge’ entered Dutch policy discourse. The ‘Team Gedragsverandering’, which started in 2009, is the oldest Dutch BIT; the others were all founded after 2012. At the same time, there are many other behavioural policy initiatives beyond the spotlights of BITs. Only focusing on BITs would be to fall for a ‘Behavioural Insights Team Bias’ (Sunstein 2014). Avoiding this bias is critical in the Dutch case too, as the great majority of our interviewees do not work in settings that fit the BIT prototype. Twenty interviewees operate in other types of collaborations that tend to be ‘looser’ and less intensive, such as informal networks, knowledge exchange groups, working groups and strategic projects. Also, seven interviewees work primarily individually (e.g. as ‘Gedragsbeinvloeders’ [Behaviour Influencers]). 3.2 Knowledge In this section we turn to the knowledge mechanism of professionalization. This mechanism refers to how behaviour experts develop and regulate their own knowledge content. We will shed light on the educational background of behaviour experts, the nature of their knowledge, and the presence of overarching knowledge platforms. 3.2.1  Education and training The professional background of behaviour experts in Dutch central government is mixed. While there is no typical schooling required to become a behaviour expert, some types of backgrounds seem more fitting than others. Lengthy behavioural scientific schooling programmes seem useful. Of our 35 interviewees, 17 people had such an official background, with academic degrees in disciplines such as social psychology, behavioural economics, communication sciences and criminology. Interestingly, six interviewees followed the same social psychology master in ‘Behaviour Change’ at Radboud University Nijmegen. These professionally educated behaviour experts mostly work for the formalized BITs. For instance, the BIT at the Ministry of Economic Affairs and Climate Policy consists of three members that all have a doctorate in social science (experimental economics, econometrics and social psychology). Furthermore, the members of these BITs tend to have similar professional backgrounds. For example, the ‘Team Behaviour Change’ at the Dutch Tax and Customs Administration includes four social psychologists, and the Dutch Healthcare Authority consists in a collaboration between five behaviour experts, all specialized in criminology.

M4716-STRAßHEIM_9781785367847_t.indd 126

25/04/2019 13:32

Behaviour experts in government  127 Yet, those who engage in behavioural policy-making do, at present, for a substantial part (half of our interviewees) not possess formal degrees in behavioural studies. Some interviewees have backgrounds that still loosely relate to their current practice, for instance having studied political science, law or public health sciences. But for others, such as those who studied chemistry, human geography or chemical technology, the gap between their original background and their behavioural practice is much wider. They became behaviour experts in less straightforward ways, like in the case of the following interviewee, who, when asked about his background, states: ‘Dutch literature scholar – and by accident I then spent the rest of my life trying to get people to change their behaviour.’ The threshold to become a behaviour expert, in terms of one’s academic background and level of expertise, is thus relatively low at present. While one must possess a basic understanding of BIs, one does not have to be academically trained in one of the behavioural sciences. To acquire the knowledge needed, there are other ways than formal academic training, for example, self-study, on-the-job learning and post-secondary education. Some interviewees also note that they have developed their behavioural expertise more intuitively. 3.2.2  Body of knowledge Behaviour experts tend to draw on a shared set of foundational texts and intellectual heroes. We identified a few books, articles and reports that are particularly popular amongst the interviewees. These were Thaler and Sunstein’s (2008) Nudge, Cialdini’s (1984) Influence, Tversky and Kahneman’s (1974) theories about heuristics and biases in decision-making, Kahneman’s (2011) classification of System I (automatic) versus System II (deliberative) thinking, and models like ‘MINDSPACE’ (BIT UK 2010) and ‘EAST’ (BIT UK 2014). Reports from the Dutch Scientific Council for Government Policy (WRR) (and also some other Dutch policy advisory bodies) are regarded to be important as well (Dorren 2015). Next to this, behaviour experts also actively search for role models, guidelines and best practices. ‘BIT UK’ is such a role model, having published about its theoretical frameworks, methods and concrete projects. Its Randomized Controlled Trials (RCTs) to test new policy measures have inspired several interviewees to conduct similar experiments. For example, the Dutch Tax and Customs Administration’s BIT ran various letter experiments, testing the effects of slight changes in the wordings of tax return letters, which were highly similar to previous trials by ‘BIT UK’. Beyond these recurring core ideas and texts, however, interviewees individually draw upon a mishmash of other theories and models. They tend to operate with a ‘cloud’ of theories in mind and look beyond particular scientific domains. They stipulate the importance of diversity in their accumulated ‘body of knowledge’, like this interviewee does: ‘I mainly use social-psychological literature, but sometimes also political psychology, you know, it comes from all sorts of nooks and crannies, and demarcation doesn’t seem necessary to me.’ Another interviewee, similarly, notes how he draws upon various streams of knowledge in his practice, from behavioural economics to social psychology to neurology. At the same, some interviewees cherish specific models. One respondent, for example, sticks with a model from Poiesz (1999) that explains behaviour change as a three-sided matter of motivating people, strengthening their abilities, and empowering them. These kind of models help them to simplify and give structure to the often complex behaviour change cases that lie before them.

M4716-STRAßHEIM_9781785367847_t.indd 127

25/04/2019 13:32

128  Handbook of behavioural change and public policy 3.2.3  Knowledge platforms While some particular theories seem more popular amongst behaviour experts and closer to the core of the behavioural practice, no specific knowledge content appears to be required in the field. This is not a great surprise given that the field knows no formal knowledge authorities. We identified neither overarching platforms nor journals that represented the particular field of behaviour experts. Behavioural expertise seems to be built in a splintered manner, with behaviour experts drawing upon their own preferred sets of theories, based on their own unique background, without much contact with experts in other agencies or regulation from external authorities that exercise control over what they (should) know. At the same time, there are modest attempts to streamline the behaviour expert’s collectively shared content. Some organizations have written public reports on how to apply BIs. Examples are a booklet called ‘Gereedschap voor Gedragsverandering’ [Tools for Behaviour Change], which translates BIs into practical tools for communication campaign specialists, and a brochure ‘Rare jongen, die Homo Sapiens!’ [Weird guy, this Homo Sapiens!], which links BIs to the work context of social affairs. Furthermore, some organizations have conjoined in platforms to exchange knowledge, like the ‘Behavioural Insights Network Netherlands’ (BIN NL), in which delegates from nearly all ministerial departments participate. ‘BIN NL’ organizes general introductions into BIs, hosts internal meetings to share knowledge, and facilitates a digital library and community of behaviour experts. A similar inter-organizational platform is the programme ‘Handhaving en gedrag’ [Enforcement and behaviour], in which behaviour experts from different regulatory agencies sit together to build and exchange knowledge, for instance during a yearly symposium. Thus, between the different Dutch governmental organizations there is a limited number of non-directive attempts to coordinate the use of BIs. This is often the case in a governance system in which many organizations are relatively independent (Schillemans 2012). Within organizations, however, much greater initiative in building behavioural expertise exists. Disseminating behavioural knowledge internally (through educational programmes, talks, articles, advice, etc.) tends to be a core business of the interviewees. 3.2.4  Conclusion: a cohesive knowledge core with divergent peripheries Behaviour experts in Dutch government draw on a cohesive core of ideas, books, reports, and role models. Beyond this shared core, the knowledge base, expertise, and ideas are broad and diverse, and to some extent uncoupled. New knowledge and experience is mostly built and shared within organizations. There are a few platforms to exchange and streamline knowledge, yet they tend to be optional and organized in a bottom-up fashion. There is also little control from overarching bodies over the knowledge that behaviour experts (should) use. Behaviour experts seem little restrained by requirements in terms of background, experience, and knowledge. Overall, the degree of occupational closure in terms of knowledge is limited. From a strict professionalization perspective, this could be seen as disadvantageous. However, some of our interviewees highlight the pros in this, as it enables them to pragmatically develop their own mixed set of skills and specialisms. Furthermore, as most behaviour experts still draw on a small shared core of ideas and texts, this still leads to some de facto alignment and convergence. As one interviewee stated: ‘The surprise is that we’re all working in different contexts, and have been develop-

M4716-STRAßHEIM_9781785367847_t.indd 128

25/04/2019 13:32

Behaviour experts in government  129 ing our knowledge and expertise fairly separate from each other from the start, but still we’re all arriving at a sort of similar approach.’ 3.3 Standards This section addresses the standards mechanism of professionalization. We particularly look at how behaviour experts standardize required competencies, quality standards, ethical codes and instruments. 3.3.1  Required competencies The previous observation that half of the interviewees are not behaviourally schooled, clearly suggests that there are no formal requirements for behaviour experts. On an informal level, however, looking at the competencies deemed necessary in the field, it becomes apparent that almost all interviewees emphasize the sophisticated nature of their practice. It is ‘not a gimmick, but a craft’. It is ‘hard work’ that requires advanced competencies to understand and control behaviour in particular areas. It appears that behaviour experts rely on two different sets of competencies. The first set underlines their scientific competencies in applying behavioural science. Behaviour experts must know what drives human decision-making, and how that knowledge can be translated and made applicable for policy-purposes. They must also have practical experience with doing research (e.g. regarding literature study, desk research, interviewing, observation, and field experiments). The second set emphasizes the political and administrative competencies. That includes administrative experience, organizational sensitivity, know-how of policy processes and change and project management skills. Behaviour experts ideally are savvy ‘change managers’ who know how to build networks and gather support from key actors. This suggests that interviewees believe that they should be able to combine the best from two worlds: the rigor of behavioural science with the savviness of public policy work. 3.3.2  Quality standards Shared quality standards are crucial for pure professionals. In that sense, the level of professionalization of behaviour experts seems to be in its infancy because our interviewees subscribe to strongly different views on what guarantees the quality of their practice. They especially seem divided in terms of what kind of knowledge counts as valid evidence. The division is between those who favour soft professional knowledge versus those who favour hard experimental knowledge, i.e. those who favour a behaviourally-informed approach to policy-making versus those who favour a behaviourally-tested approach (Lourenço et al. 2016). Or, in Aristotelian terms: those who favour a phronesis type of knowledge versus those who favour episteme types of knowledge (Parsons 2002). The group of behaviour experts that favour experimental knowledge emphasize the evidence-based nature of their practice, and the need of ‘experimentally demonstrating that it works’. Following their modern-empiricist scientific role model, ‘BIT UK’, they advocate that policy interventions should be ex ante evaluated through small-scale controlled scientific experiments. Doing these kind of trials, like RCTs and A/B testing with websites, is the most reliable way to predict human responses to policies, as it helps to isolate the actual effects of single interventions.

M4716-STRAßHEIM_9781785367847_t.indd 129

25/04/2019 13:32

130  Handbook of behavioural change and public policy What you mainly want with behaviour change is examining and demonstrating that it works … And because we’ve got this increase in data, there are much more behaviours … that we can map out. So, then you can start running experiments and effect measurements with A/B testing. And then you’re able to see: this intervention works better than that one.

The group of other experts, favouring professional knowledge, however dismiss RCTs as ‘the holy grail’. They note that RCTs are relatively expensive, time-consuming, and also difficult to set up (challenges involve formulating quantifiable outcomes, keeping control over the treatment groups in a dynamic field setting, and isolating the effects). On top of that, RCTs can introduce ethical questions and political difficulties (e.g. institutional resistance to effect measurement out of fear for potential negative outcomes). Experts in this group thus rely on a different, in their eyes more feasible, source of knowledge: their own professional judgment, based on a mix of common sense, logical reasoning, creative thinking, theoretical knowledge, and field experience. These behaviour experts deviate from the scientific model, stressing that ‘this here is not a university’ and that their job is ‘not an exact science, but an iterative model’ that involves making ‘educated guesses’ and learning through trial-and-error. They often have pragmatic reasons for their approach: limits to organizational resources force them to make do with what they have, and generally this means trading the rigor of experimental knowledge for the relevance of professional ‘indications’. They find themselves in a balancing act, having to make tradeoffs between certainty, rigor, and thoroughness on one hand, and feasibility, speed, and efficiency on the other. They have to make compromises: gathering just enough certainty to make valid assessments without needlessly suspending the flow of work. A result of this balancing act is that these behaviour experts must cope with uncertainty. They lack hard experimental evidence to prove their effectiveness. ‘It’s not rocket science’, one interviewee claims, ‘that’s the difficulty in behaviour change, you can’t promise that it’s going to work’. Even though behaviour experts de facto adhere to different quality standards, they are nonetheless united in their support of evidence-based policy as an ideal. Evidence-based policy-making thus serves as an informal shared frame of reference (or loose standard) for the behaviour experts. 3.3.3  Ethical codes A particular area of interest for standardization is the regulation of ethical norms, given the extensive ethical debates about behavioural policy-making in the administrative (e.g. House of Lords Science and Technology Select Committee 2011) as well as in the academic world (e.g. Bovens 2008). Ethical objections, inter alia, state that nudging lacks transparency, does not promote what people themselves want and diminishes their autonomy through bypassing their reflective decision-making (e.g. Hansen and Jespersen 2013). Although these concerns have traditionally been raised in the political context of the United States, where state intervention meets relatively strong political resistance, they seem to have been transferred to interviewees’ European context too. The popularization of ‘Nudge’ in the Dutch policy sector has not only introduced potentially innovating ways of policy-making, but also sensitive underlying ethical issues (e.g. WRR 2014). Hitherto, ethical debates have not led to shared ethical standards for behavioural policy-making. Instead, interviewees deal with ethical issues on the basis of their own moral compasses. Some disregard the ethical discussion and claim that ethics is not their area of expertise. Others reject the ethical concerns that nudging would be inherently

M4716-STRAßHEIM_9781785367847_t.indd 130

25/04/2019 13:32

Behaviour experts in government  131 ‘dirty’ or ‘wrong’. They emphasize that behaviourally-inspired interventions are by their nature not more or less legitimate than other policy tools. Some interviewees do give more credit to ethical concerns, particularly in relation to the questioned legitimacy of targeting unconscious behaviour. One way to overcome these concerns is by setting prior conditions for legitimate practice. Transparency seems to be such an important condition. It has led the Dutch Tax and Customs Administration’s BIT to publish openly about its existence, working methods, and nudges in a major Dutch newspaper (De Jong and Rusman 2 March 2015). 3.3.4  Standardized instruments The one area where we see clear examples of standardization is in the area of instrumentation. Although most behaviour experts are still setting up shop, we identified some standardized techniques and methods that are popular in the field. These are mostly policy analytical tools. ‘CASI 3.0’ is such an example, a tool which helps communication specialists to develop a behaviourally-informed communication campaign strategy. Similar tools are the ‘Gedragstoets 2.0’ [Behaviour Test 2.0] and the model ‘EAST’ (which exists both as a general guideline and also as a deck of inspiration cards) developed by ‘BIT UK’ (2014). What unites these tools is that they all act as action plans for designing behaviourally-informed policy and communication strategies. In several steps, in series of questions, these tools help to explicate and examine underlying behavioural assumptions in policy and campaign strategies. Underneath, these diagnostic tools subscribe to a rational-comprehensive (Lindblom 1959) perspective on policy-making in which policy problems are (and can be) solved through rational and systematic analysis. To illustrate, one interviewee developed his own systematic ‘action plan for behaviour change’ based on six steps, moving from ‘describe the problematic behaviour’, to ‘describe the desired behaviour’, to ‘analyse the behaviours’, to ‘select the behavioural measures that match the analysed situation’, to ‘implement the measures’, to ‘test the measures with pre- and posttests and control situation’ (Gemeente Schoon 2014). Behaviour experts have also to some extent standardized the use of nudges. Nudging involves the translation of behavioural science into minor readjustments – ‘psychological gimmicks’ – of people’s daily surroundings. Due to their small and informal nature, nudges appear relatively easy to standardize. To illustrate, one interviewee mentioned having some ‘basic’ nudges that she applies throughout her work. A nudge that she standardly suggests to communication colleagues is that when penalizing companies they should always immediately offer a future solution too. Such an action perspective encourages behaviour change, she claims. It is a quick win that cannot go wrong. Standardization can however also be more challenging, especially when behaviour experts deal with complex behaviours in dynamic choice environments. 3.3.5  Conclusion: some shared instruments but little standards beyond that Our research suggests that there are some uniform standards but very little beyond that. There are most notably no formal qualifications needed to become a behaviour expert and there are no clear ethical or quality guidelines. Organizations mostly set their own professional standards. The only area where we see more advanced standardization is in the area of instrumentation. At the same time, some informally shared standards seem to exist, for instance in relation to required competencies and quality of work. Another less

M4716-STRAßHEIM_9781785367847_t.indd 131

25/04/2019 13:32

132  Handbook of behavioural change and public policy formalized but nonetheless important feature of cohesion is the diagnostic nature of the job. Most behaviour experts adhere to a rational-analytical view of policy-making where they are all, some more outspoken than others, behavioural science appliers to policies. 3.4 Positions This section deals with the positions mechanism of professionalization, which concerns how behaviour experts position themselves vis-à-vis others. We will mainly address their search for legitimacy and autonomy. 3.4.1 Legitimacy The search for organizational legitimacy is a particularly thorny issue for the interviewees as they need to carve out a distinctive position in well-established policy-arenas that feature various and competing values and actors. As a result, making themselves visible, giving pitches and internal consultations and building networks, consumes a substantial amount of their time and energy. Many interviewees are, as one put it, ‘struggling’ to win the support from others. An interviewee states: ‘Most of our time is lost by needing to convince our internal clients to do things different externally.’ Part of the struggle lies in general challenges faced by any innovation (e.g. lack of time and resources, start-up problems, not being prioritized, etc.) and in the fact that behaviour experts – and their associated theories and working methods – are not always understood by their colleagues with different backgrounds. Underneath that, a more fundamental mismatch between the experts and the wider policy-making culture seems to exist. Their introduction of a new school of knowledge (behavioural economics) and increased attention for policy experiments and in-depth policy analysis, not only challenges the prevailing knowledge paradigms (law and economics) in Dutch government, it also slows down the policy cycle and fuels it with uncertainty. This does not cohere well with a speedy and solutionoriented policy culture with little space for uncertainty and ‘second order’ activities like analysis and evaluation (Lindblom 1959). At the same time, the fact that behaviour experts are struggling does not wholly prevent them from making progress in finding a legitimate role, generating visibility, and overcoming resistance. These two BIT members, for example, note that their position has gained strength with the years: Our direct colleagues come to us quite a lot … I think in that respect we’re further developed than many other regulation agencies who are still very much concerned with putting themselves on the map internally, building support in order to get colleagues to work with behavioural insights, or getting at the table somewhere at all… Getting involved in things… Well, that’s something we really don’t have to do. Over here, colleagues know what our role is, what our contribution is, and what can be expected of us.

3.4.2 Autonomy As they strive towards more control over their practice, autonomy is another major theme for the behaviour experts. The formation of BITs and other specialized behavioural functions may be of importance in getting more occupational control, as it gives concrete institutional shape to the use of BIs. These BITs do not per se need formal legitimacy to be able to exercise autonomy. One BIT was even purposefully organized unofficially and kept ‘under the radar’ of the organization, with one of its members noting: ‘As a BIT we’re not

M4716-STRAßHEIM_9781785367847_t.indd 132

25/04/2019 13:32

Behaviour experts in government  133 an officially founded team… rather we did it on our own initiative. With the goal to keep as much freedom in what we do and not having to operate within the boundaries of the organization.’ At the same time, the autonomy that comes with not being institutionalized can be compensated by other limitations. Informal BITs have a harder time in acquiring resources and gaining moral support, and they cannot fall back on formal structures, powers and rules. Hence, in the end most behaviour experts do strive for more autonomy through further formal institutionalization. Despite their strivings for autonomy, at the same time the reality of most behaviour experts is that they are highly dependent on many other policy actors. They mostly work with or for others. These dependencies take shape in different forms. Sometimes behaviour experts play advisory roles in existing policy projects. It also happens that they step in as trainers, teaching colleagues how to make use of BIs. Or they become project managers, building broad networks of expertise (including universities, consultancy bureaus, trade organizations, and citizens). The BIT at the Ministry of Infrastructure and Water Management, for example, adopts a ‘network-approach’ which builds a flexible team of experts around each of its projects. Through collaboration, this unit seeks to acquire the necessary knowledge and experience to engage in complex behavioural policy initiatives. For many behaviour experts, an awareness of their dependency within the broader policy system is key. They need to blend into the system, and collaborate with others, in order to work effectively. As one interviewee explains, ‘Behaviour change is about aligning the organization from A to Z so that it carries out the same message and makes the same movements.’ 3.4.3  Conclusion: shared struggles for support in the face of dependency As newcomers in a well-established policy system, behaviour experts face the issue of having to build legitimacy. While some have more success than others, this issue generally tends to be a struggle. The behaviour experts seek more autonomy to increase occupational control. At the same time, their dependencies on others force them to further integrate into the existing system, trying to connect to and merge with existing practices. 3.5 Identities This section turns to the identities mechanism of professionalization, which refers to the ways in which behaviour experts construct and regulate their own identity. We look at their heroes, role models, and typical organizational roles. 3.5.1  Heroes and role models The identity of the behaviour experts partly stems from the role models they follow, such as Thaler and Sunstein, Kahneman, Cialdini and ‘BIT UK’. The fact that the Dutch field also has installed ‘BITs’, mirroring ‘BIT UK’ in name and organizational model, reflects the exemplary role of ‘BIT UK’ well. Yet, many other role models are mentioned, including academic scholars, think tanks, and fellow experts. Similar to how interviewees tend to draw on their own particular ‘cloud’ of ideas, they also have their own particular ‘cloud’ of heroes and role models.

M4716-STRAßHEIM_9781785367847_t.indd 133

25/04/2019 13:32

134  Handbook of behavioural change and public policy 3.5.2  Role identities In a broad sense, our interviewees share the same quest: they are all revolutionists (or phrased more modestly: innovators) who cast doubt upon the, as one put it, ‘bizarre’ conventional ways of policy-making. They all believe that harnessing the power of BIs results in better policies. At the same time, beyond this common quest, interviewees take on different organizational roles that express different identities. Some of these roles seem more prevalent than others. One prevalent role, taken on predominantly by about half of the interviewees, is when behaviour experts directly engage in choice architecture. They actually intervene in particular contexts, for instance in school canteens to stimulate healthy eating or in local neighbourhoods to avoid littering. Sometimes this direct type of choice architecture consists in being creative, offering ‘fresh looks’ on existing practices and loosely suggesting small interventions. Other times it consists in doing rigorously experimental research, setting up labour-intensive RCT projects. Yet, behaviour experts are not always direct choice architects. In (at least) three other roles, their contributions towards behaviour change are more indirect. A first more indirect role consists in making policy analyses. In this role, behaviour experts purposefully play the ‘advocate of the devil’, trying to challenge and (if needed) correct the deeper behavioural assumptions that underlie specific policy theories. They aim for a more thorough policy analysis through dialogue, ‘constantly asking difficult questions… Whyquestions, like a little kid: “Yes, but why? Why? Why?”’ This role resembles Socrates’ role as the social gadfly in Ancient Greece. A second more indirect role consists in managing behavioural policy projects and networks. In this role, behaviour experts set aside matters of content, and instead focus on the process of behavioural policy-making and who gets involved. They ‘assemble’ and coordinate networks of expertise, while managing administrational and political issues. A third more indirect role consists in spreading knowledge about behaviourally-informed policy-making. In this role, behaviour experts mainly inspire other actors to use BIs. They act as intermediaries, positioned between science and policy practice, ensuring that policy-makers and practitioners learn about BIs in a language they understand. 3.5.3  Conclusion: circuitous roads on a common path Behaviour experts implicitly share a small common core of role models, with much fragmentation beyond that. While they are united as BI revolutionists, they also play (and combine) different organizational roles. Table 9.2 situates these (neither all-comprehensive nor mutually exclusive) four roles described above along the lines of how directly they contribute to direct choice architecture. Table 9.2  Diverging roles of behaviour experts Indirect choice architecture

Direct choice architecture

Messenger

Network node

Social gadfly

Choice architect

Spreading knowledge

Managing networks and projects

Making policy analyses

Running RCTs and realizing behaviour change

M4716-STRAßHEIM_9781785367847_t.indd 134

25/04/2019 13:32

Behaviour experts in government  135

4.  CONCLUSIONS AND DISCUSSION In this chapter, we have studied behaviour experts in Dutch central government. More specifically, we have studied how their occupation is developing, interpreting it as a case of early professionalization. Our study shows that behaviour experts are slowly finding their way into Dutch central government, and are being attracted by many agencies across the board. The experts are still at an explorative stage, in the process of setting up shop, making themselves seen, and establishing effective practices, with minimal resources, powers, and ties to their organization. While the behaviour experts are far from identical, they do share a few core characteristics. To begin with, they all more or less rely on a number of iconic authors (e.g. Thaler and Sunstein, Kahneman and Cialdini), role models (e.g. ‘BIT UK’), advisory reports (e.g. WRR 2014), tools and instruments (e.g. ‘C.A.S.I. 3.0’), and loosely coordinated professional networks (e.g. ‘BIN NL’). Moreover, behaviour experts all tend to view policy-making as a highly rational-scientific affair. Though some are more outspoken and rigorous than others, they all apply behavioural science to policies. Nevertheless, behaviour experts form an occupation that is currently more fragmented than cohesive. They give shape to their work in different ways, based on their own experiences, beliefs, and preferences. They rely on their own, widely diverging ‘clouds’ of ideas and role models. The observed fragmentation can be linked to at least five underlying forces, suggesting that behaviour experts will remain relatively fragmented in the near feature. First, government agencies in the Netherlands are traditionally highly autonomous (Schillemans 2012), which may explain part of the fragmented development of behavioural expertise. Second, behaviour experts are relatively unhampered by the kind of restrictions and obligations that the more mature occupations face. This gives them more leeway in shaping their own practice, resulting in fragmentation. Third, behaviour experts are part of government agencies with widely differing tasks as well as interactions with citizens and companies. The fragmentation may be a result of behaviour experts customizing their practice to their particular policy context. Fourth, government agencies are increasingly withdrawing from direct interaction with citizens. This new role puts pressure on the behaviour expert’s most typical role as a choice architect, directly shaping the environment in which people make choices. With the current political emphasis on co-production, agencies are forced to explore different, more indirect routes towards choice architecture, strengthening further fragmentation. Fifth, behaviour experts draw from a rather broad universe of insights from various behavioural scientific disciplines (including contradictory findings and disputed theories). This broadness allows some behaviour experts to cherish their own hobby horses, and some others to merge all sorts of insights into a particularistic hotchpotch, strengthening further fragmentation in the field. Our study has shown salient ways in which behaviour experts in the Netherlands are professionalizing. Given their current status as a loosely coupled group of unfledged outsiders – with innovating ideas but little resources, promising methods but still small portfolios – the question remains how strong their influence on policy-making will become.

M4716-STRAßHEIM_9781785367847_t.indd 135

25/04/2019 13:32

136  Handbook of behavioural change and public policy

REFERENCES Abbott, A. (1988), The Systems of Professions: An Essay on the Division of Expert Labor, Chicago: University of Chicago Press. Alemanno, A. and A. Sibony (2015), Nudge and the Law: A European Perspective, Oxford and Portland: Hart Publishing. Behavioural Insights Team UK (2010), MINDSPACE, Institute for Government. Behavioural Insights Team UK (2014), EAST. Four simple ways to apply behavioural insights, Institute for Government. Bovens, L. (2008), ‘The ethics of nudge’, in Grüne-Yanoff, T. and S.O. Hansson (eds), Preference Change: Approaches from Philosophy, Berlin and New York: Springer, pp. 207‒219. Cialdini, R.B. (1984), Influence: The Psychology of Persuasion, New York: HarperBusiness. De Jong, S. and F. Rusman (2015), Denkt u aan uw aangifte? Dank! Liza en Joyce. 2 March. NRC Dorren, L. (2015), ‘Principes in uitvoering’, Master’s Thesis, Utrecht School of Governance. Faulconbridge, J.R. and D. Muzio (2012), ‘Professions in a globalizing world: towards a transnational sociology of the professions’, International Sociology, 27(1), 136‒152. Freidson, E. (2001), Professionalism, the Third Logic. On the Practice of Knowledge, Cambridge: Polity. Geertz, C. (1973), ‘Thick description: toward an interpretive theory of culture’, in Martin, M. and L.C. McIntyre (eds), Readings in the Philosophy of Social Science, Cambridge, MA: MIT Press, pp. 213‒232. Gemeente Schoon (2014), Stappenplan voor gedragsverandering. Accessed 12 January 2019 at http://www.ken​ niswijzerzwerfafval.nl/document/stappenplan-voor-gedragsverandering. Hansen, P.G. and A.M. Jespersen (2013), ‘Nudge and the manipulation of choice’, The European Journal of Risk Regulation, 1, 3‒28. House of Lords Science and Technology Select Committee (2011), ‘Behaviour Change. Science and Technology Select Committee. 2nd report of session 2010–12’, London: Authority of the House of Lords. John, P. (2014), ‘Policy entrepreneurship in UK central government’, Public Policy and Administration, 29(3), 257–267. John, P., Smith, G. and G. Stoker (2009), ‘Nudge nudge, think think: two strategies for changing civic behaviour’, The Political Quarterly, 80(3), 350‒361. Kahneman, D. (2011), Thinking, Fast and Slow, New York: Macmillan. Larson, M.S. (1977), The Rise of Professionalism. A Sociological Analysis, Berkeley: University of California Press. Lindblom, C.E. (1959), ‘The science of “muddling through”’, Public Administration Review, 19(2), 79‒88. Lourenço, J.S., Ciriolo, E., Almeida, S.R. and X. Troussard (2016), Behavioural Insights Applied to Policy, European Report 2016, Joint Research Centre, European Commission. Lunn, P. (2014), Regulatory Policy and Behavioural Economics, Paris: OECD Publishing. Meulen, M. van der (2009), Achter de schermen. Vakontwikkeling en professionalisering van publieke managers in de zorg en bij de politie, Delft: Eburon. Ministry of Economic Affairs (2014), Reaction of the Cabinet to the advisory reports of RLI, RMO and WRR about the use of behavioural scientific insights in policy. Accessed 7 January 2019 at https://www.eerstekamer. nl/behandeling/20141204/brief_regering_kabinetsreactie_op/info. Noordegraaf, M. (2007), ‘From ‘pure’ to ‘hybrid’ professionalism’, Administration & Society, 39(6), 761‒785. Noordegraaf, M., Steen, M. van der and M. van Twist (2014), ‘Fragmented or connective professionalism?’, Public Administration, 92(1), 21‒38. Parsons, W. (2002), ‘From muddling through to muddling up’, Public Policy and Administration, 17(3), 43‒60. Poiesz, T. (1999), Gedragsmanagement, Utrecht: AnkhHermes. Schillemans, T. (2012), Mediatization of Public Services. How Organizations Adapt to the News Media, Frankfurt: Peter Lang. Schonewille, A. (2015), What activation practitioners do. Dissertation Vrije Universiteit Amsterdam, Ridderprint BV. Simon, H.A. (1957), Models of Man: Social and Rational, New York: John Wiley and Sons. Stinesen, B.B. and R.J. Renes (2014), ‘Overheidsbeleid gericht op gedragsverandering: meer dan een mooi streven?‘ Den Haag: Raad voor Maatschappelijke Ontwikkeling. Sunstein, C.R. (2014), ‘Nudging: a very short guide’, Journal of Consumer Policy, 37(4), 583‒588. Thaler, R.H. and C.R. Sunstein (2008), Nudge: Improving Decisions about Health, Wealth and Happiness, London: Penguin Books. Thorstendahl, R. (1990), Essential Properties, Strategic Aims and Historical Development: Three Approaches to Theories of Professionalism, London: Sage. Tversky, A. and D. Kahneman (1974), ‘Judgment under uncertainty: heuristics and biases’, Science, 185(4157), 1124‒1131.

M4716-STRAßHEIM_9781785367847_t.indd 136

25/04/2019 13:32

Behaviour experts in government  137 Wetenschappelijke Raad voor het Regeringsbeleid (WRR) (2014), Met kennis van gedrag beleid maken, Amsterdam: Amsterdam University Press. Whitehead, M.J., Jones, R., Howell, R., Lilley, R. and J. Pykett (2014), ‘Nudging all over the world’, Economic and Social Research Council (UK). Wilensky, H.L. (1964), ‘The professionalization of everyone?’, American Journal of Sociology, 70(2), 135‒158. World Bank (2015), Mind, Society and Behavior, Washington: The World Bank.

M4716-STRAßHEIM_9781785367847_t.indd 137

25/04/2019 13:32

10.  Nudge and the European Union Alberto Alemanno

1. INTRODUCTION Europe has largely been absent from the US-dominated debate surrounding the introduction of nudge-type interventions in policy-making. Yet the European Union (EU) and some of its Member States are exploring the possibility of informing their policy action with behavioural insights. While a great deal of academic attention is currently been paid to the philosophical, ethical and other abstract implications of behavioural-informed regulation (see, e.g. Conly 2013; Rebonato 2012; White 2013; Sunstein 2013; Hansen and Jespaersen 2013; Selinger and Whyte 2011; Bovens 2012; Sunstein 2015) , such as those concerning autonomy (Wright and Ginsburg 2012), dignity (Waldron 2014; McCrudden 2015) and moral development (Bovens 2008), this chapter charts and systematizes the incipient European nudge discourse. Besides a few isolated initiatives displaying some behavioural considerations (e.g. consumer rights, revised tobacco products directive, sporadic behavioural remedies in competition law), the EU – similarly to its own Member States – has not yet shown a general commitment to systematically integrate behavioural insights into policy-making. Given the potential of this innovative regulatory approach to attain effective, low-cost and choicepreserving policies, such a stance seems surprising, especially when measured against growing citizen mistrust towards EU policy action. At a time in which some EU countries are calling for a repatriation of powers and the European Commission promises to redefine – in the framework of its Better Regulation agenda (Alemanno 2015) – the relationships between the Union and its citizens, nudging might provide a promising way forward. In the aftermath of the Brexit vote, this promise has not only been shared by the 27 remaining Member State but also represents one of their major priorities.1 Yet with promises come challenges too. The chapter proceeds as follows. The first section sets the scene by discussing the growing appeal of nudging among policy-makers within and across Europe. The second section introduces the notion of behavioural policy-making and contrasts it with that of nudging. The third section describes the early and rather timid attempts at integrating behavioural insights into EU policy-making and identifies some domestic experiences. The fourth section discusses the institutional and methodological efforts undertaken by the EU and some of its Member States to embrace behavioural policy-making. In turn, the fifth section discusses the major difficulties of integrating behavioural insights into EU policy-making and offers some concluding remarks.

2.  THE APPEAL OF NUDGING Whilst the integration of behavioural sciences into policy-making is not an entirely new phenomenon (see, e.g. Berns 1963),2 the idea is spreading that regulatory actions cannot 138

M4716-STRAßHEIM_9781785367847_t.indd 138

25/04/2019 13:32

Nudge and the European Union  139 work effectively or efficiently if policy-makers do not consider how targeted people respond (see, e.g. Adler 2008; Hayden and Ellis 2007: pp. 660‒667; Jolls et al. 1998).3 By showing that individuals deviate in predictable ways from neoclassical assumptions of rationality, behavioural research is set to equip policy-makers with new promising approaches to complement more traditional policy-making. Thus, since people – due to inertia and procrastination – tend not to make affirmative choices, default rules, such as opting citizens into insurance schemes by default, have achieved a larger social impact than incentive-based regulation. ‘Framing’ and the presentation of information are also strategic interventions to influence choices: the European Commission is, for example, helping retail investors understand financial products by simplifying complex procedures into Key Information Documents. Behaviour patterns are also heavily influenced by the emergence of social norms, as people are constrained by reputational forces and care about the perceptions of others: an experiment showed that peer pressure increased individual savings by 3.5 times. Finally, evidence suggests that salient and vivid warnings are more effective than statistical and abstract information sets, with serious implications for policy communication and delivery.4 In a wide range of policy fields, such as energy, health, financial services or transport, experimental findings in behavioural research can be used by public authorities (see, e.g. Yeung 2011) in connection with traditional regulatory tools to produce behavioural change (Lobel and Amir 2009). Traditional regulatory tools include command-andcontrol mechanisms, such as coercion (e.g. using threats to ensure compliance), bans (e.g. prohibiting smoking in restaurants) or authorizations (e.g. ensuring that products meet certain requirements of trustworthiness or safety), as well as market-based mechanisms, used to adjust financial incentives (e.g. paying students to get good grades or to follow a healthy diet) or to address economic externalities.5

3. THE EMERGENCE OF NUDGING OR, MORE BROADLY, BEHAVIOURAL POLICY-MAKING This instrumental use of behavioural insights in policy-making is indifferently referred to – by observers, pundits, and scholars – as nudging, behavioural law and economics (see, e.g. Jolls et al. 1998; Jolls 2010; Bubb and Pildes 2014), behaviourally-informed regulation or empirically-informed regulation (Sunstein 2011). However, behind this semantic variance, the phenomenon covered by this prolific terminology is essentially one, namely the application of behavioural insights to policy-making.6 Yet the universal appeal of behaviourally-informed intervention has largely to do with its being presented as a ‘nudge’: any measure intended to affect ‘any aspect of the choice architecture that alters people’s behaviour in a predictable way without forbidding any options or significantly changing their economic incentives’ (Thaler and Sunstein 2008, p. 6). Nudging is generally presented as a cheap and smart alternative to expensive traditional regulatory measures (Sunstein 2011; Yeung 2012). Indeed, the fact that behavioural-informed intervention does not always require legislation is perceived as one of its attractive features for governments (Marteau et al. 2011). Even when it is not used as an alternative to but in combination with traditional legal tools, such as command-and-control and in particular fines, behaviourally informed intervention can still represent a cheap alternative to costly

M4716-STRAßHEIM_9781785367847_t.indd 139

25/04/2019 13:32

140  Handbook of behavioural change and public policy enforcement mechanisms (Alemanno and Spina 2014). Second, nudging promises to be choice-preserving, by always enabling the addressee to opt out of the preferred policy option.7 Third, in specific areas, regulation needs to become behaviourally informed, not to ‘nudge’ citizen, but to offer a ‘counter-nudging’ force against the exploitative use of behavioural insights by market actors. For example, businesses, in particular the new actors of the digital economy, are using behaviourally informed strategies to steer consumer choices.8 It is therefore useful to distinguish between two typologies of behavioural-informed intervention: public nudging and regulation of private nudging (also more aptly named ‘counter-nudging’).9 While public nudging is characterized by the intention to help people correct errors they may be subject to regardless of their exploitative use by market forces, the regulation of private nudging primarily aims at countering active corporate strategies to nudge consumers. In such situations, public intervention does not only seek to correct a bias that (some) people may have; it seeks to counter active corporate exploitation of such biases. Regardless of whether it qualifies as public nudging or counter-nudging, the emerging model of behavioural-informed intervention is based upon the premise that any sensible regulation system must consider how the findings of cognitive science might alter our understanding of the behaviour of citizens. In particular, its inclusion into the regulatory process should prevent policy-makers from making irrational decisions, either because of their own misperceptions, unforeseen reactions from the public or exploitative uses of behavioural biases by the market. Under this emerging approach, behavioural analysis is perceived as an opportunity to improve the efficacy as well as the efficiency of regulatory intervention, especially when – as is often the case – it aims at behavioural change, as it has been demonstrated by several studies over the last decades (Jolls et al. 1998; Jones et al. 2013, pp. 163 et seq.).

4.  THE TIMID EU’S EMBRACE OF NUDGING While behavioural sciences have not been formally integrated into EU policy-making, some of their insights have been integrated into several of its policies, especially consumer protection, health and safety. This is likely accounted for by the fact that these policy areas derive their success from their ability to influence individual behaviour (as opposed to corporate behaviour). Thus, one of the early examples includes the cooling-off period, which can be found in much of EU consumer protection legislation.10 This remedy, by enabling consumers to unconditionally cancel a contract within a limited time span, aims at allowing them to counter their myopia or impulse buying. Similarly, the Directive on Consumer Rights limits the use of pre-checked boxes in order to limit the power of inertia in consumer contracts.11 As a result, if the retailer offers the consumer additional extras – for example, purchasing insurance with a flight ticket – these cannot already be pre-selected on the page. The consumer must positively opt in or tick the box in order to select the relevant products. In the high-profile Microsoft case relating to the bundling of Microsoft Internet Explorer web browser with Windows, the EU Commission services relied on behavioural insights when designing the relevant remedy. Under this remedy, users of Windows-PCs were provided with the option to choose an alternative browser via an on-screen ballot. The idea was to nudge consumers to make an active choice as

M4716-STRAßHEIM_9781785367847_t.indd 140

25/04/2019 13:32

Nudge and the European Union  141 to their preferred browser, thus neutralizing the impact of the default option.12 EU data privacy law has also been a field in which behavioural considerations have informed the policy debate (Borgesius 2013), for example with regard to the cookie law in the E-privacy Directive as amended by the Directive 2009/136/EC.13 Despite these isolated initiatives displaying some behavioural considerations, the EU has not yet shown a general commitment to systematically integrate behavioural insights into policy-making. This is not entirely surprising given the challenges pertaining to the lack of a cognitive theory, the absence of a framework for deciding in face of heterogeneity at the population level and the difficulties to extrapolate from extant behavioural studies. More critically, if behavioural sciences demonstrate the limits of rational action and provide a better understanding of human behaviour, there is no ready-made framework for incorporating their insights into policy-making.14 While policy-makers should have regard to the scientific validity underlying behavioural findings, this aspiration to scientific rigour does not lead to a unique model of behavioural policy-making and the different cultural and social settings will play a role.

5. THE INSTITUTIONAL DESIGN OF NUDGING ACROSS EUROPE: TOWARDS A EU NUDGE UNIT? The question of how to best organize the integration of behavioural policy-making within current EU and Member States constitutional and institutional settings is central to today’s debate surrounding nudge. The UK, being the first mover, seems to have set up the golden standard for behavioural institutional design: a dedicated unit, the Behavioural Insights Team (BIT), initially placed within the Cabinet Office, and made of few experts, specialized in several behavioural disciplines, who work in direct contact with the different government departments. When advising public administrations or charities on the integration of behavioural insights, the BIT relies on a wide range of practical measures to change citizens’ behaviour, spanning from re-wording the content of letters sent by public authorities to tax-payers15 to using small ‘thank you’ gifts to reward charitable donations (cf. Cabinet Office and BIT 2013). An institutional experience whose development appears almost antithetical to that of the UK can be observed in Denmark, where a bottom-up organization, called iNudgeU, animated by academics, civic advocates and behavioural professionals, has created a self-proclaimed Danish Nudge Network.16 Interestingly enough, while the BIT has acquired its autonomy from the cabinet office and been granted a private sector status,17 the Danish Nudge Network has progressively been incorporated into the Danish government. If they differ in the chronology and direction of change between public and private sector, both experiences however have in common that a dedicated unit has been created. An alternative or complementary model would be to educate policy-makers in government departments on a wider scale (Sunstein 2014, p. 587). Across Europe, there is high level of variation in Nudge awareness:18 in many countries, the initial excitement about the novelty of behavioural sciences and its innovation potential has not made inroads into the policy-making circles, and remained confined to academia. Anne-Lise Sibony and I called this the pre-Nudge stage, meaning that the eponymous book has not yet been widely read (Alemanno and Sibony 2015). A second circle is made up of countries in incipient Nudge stage, where awareness of behavioural regulation is present

M4716-STRAßHEIM_9781785367847_t.indd 141

25/04/2019 13:32

142  Handbook of behavioural change and public policy but restricted to very restricted circles. In this category, Germany19 and France20 have shown some sign of interest for behavioural policy-making, but, at the time of writing, it is not clear whether they will go as far as to set up dedicated units. In the Netherlands, the Scientific Council for Government Policy (WRR) recently published a report entitled ‘Policymaking with knowledge of behaviour’.21 The UK and partly Denmark represent a more mature Nudge stage, where cognitive based interventions are tested, used and publicly debated. No country seems to have reached a post-Nudge stage yet. In addition, also the degree of maturity varies within the EU. The development of Member State expertise about behavioural regulation progresses in parallel with developments at EU level. This is set to raise questions regarding the federalism of behavioural regulation. In particular, it will be interesting to observe whether and how behavioural arguments may play in the framework of subsidiarity review, for example, should Member States establish that, in some policy areas, different behavioural patterns justify regulation at national rather than at EU level. In the meantime, the European Union established a ‘Foresight and Behavioural Insights Unit’, which is located within the EU Commission Joint Research Centre. The unit’s raison d’être is to centralize the efforts currently undertaken by some Directorates General of the EU Commission, such as DG SANTE and DG CONNECT, to integrate behavioural insights into EU policy-making. By overcoming the current institutional fragmentation, the unit is expected to develop a robust methodology and to foster a behavioural mindset among the EU Commission civil servants. It remains to be seen what influence this institutional effort will have into the EU administrative culture and whether it will affect the domestic and local level. For the time being, the EU ‘nudge unit’ has not yet decided which of the various institutional design models to embrace – if any. Given the EU constitutional and institutional specificities, one cannot rule out that the EU will develop a new, autonomous model. While the use of behavioural insights does not appear central to newly published Better Regulation (BR) Package,22 several ideas herewith presented indicate a gradual awareness by the EU of the relevance of accounting for people’s behaviour when adopting new policies. Thus, for instance, the Better Regulation toolbox23 concedes for the first time that not only market failures but also ‘behavioural biases’ may call for public action (Better Regulation toolbox, p. 72). In particular, it is said that when market forces do not achieve an efficient outcome due to behavioural biases, ‘a public intervention may be justified which better reflects individuals’ actual behaviour’ (Better Regulation toolbox, p. 72). It is also suggested that behavioural biases must be taken into due account not only in the definition of the policy problem but also in the design of the various policy options capable of attaining the declared objective.24 Although the actual added value of incorporating behavioural insights along the policy cycle is still open to scrutiny, it appears undisputable at this stage that policy-makers, including the EU itself, are expected to increasingly take them into account.

6.  THE CHALLENGES AHEAD FOR A EU NUDGING STATE There exist significant legal as well as practical limitations constraining the ability of the EU to nudge its citizens through choice architecture. This should not come as a surprise.

M4716-STRAßHEIM_9781785367847_t.indd 142

25/04/2019 13:32

Nudge and the European Union  143 If nudging raises a number of significant concerns – be they about its legitimacy, legality or effectiveness – when it is performed by a nation state, it appears intuitively more difficult when it is a supranational entity, like the EU, doing the nudging. One must first observe that, largely due to the composite nature of its administration, the EU – in contrast to any other jurisdiction – has virtually no direct contact with its citizens. This is true for at least three reasons. First, most of the competences requiring states to interact with their citizens, such as taxation, welfare, defence, public health or education, have not been transferred to the Union, but remain in the hands of its Member States. Second, even when it comes to policy areas of EU competence, in the absence of a specific delegation of enforcement power to the Union, their implementation is primarily the prerogative of the Member States. As a result, the latter rely on their administrations, rules of procedure and – in the case of directives – their own solutions when ensuring the implementation and enforcement of EU actions. Third, as a result of the decentralized nature of the EU as an administration, the provision of administrative services – even those governed by the EU – to citizens is generally provided by the Member States, rather than by the EU itself. This is reflected in the tiny EU budget (1 per cent) when confronted to its GDP. The EU’s only direct yet patchy relations with its citizens are largely confined to the management of some of its funding programmes. More critically, nudging supposes a shared, political understanding of what is ‘good’ in society, or at least of what is permissible and what is not. That’s exactly what, in principle, any sovereign state does: to decide what an ‘ideal’ or ‘worthy’ citizen is through the definition of good and bad habits. Yet when applied to the EU, such an assumption does not hold. The EU shares its citizenry with the Member States and as a result several visions of what a ‘good citizen’ is – or ought to be – tend to compete. Despite the limited opportunities for the EU to directly affect its citizens, the Union has over time been increasingly shaping their lives through its legislative, regulatory and judicial action. In particular, the paired instruments of the internal market and EU citizenship have allowed the European Union to redefine at the supranational level what is ‘good’ for its citizens, regardless of the country they come from within Europe (Kochenov 2013). In the implementation of internal market provisions, the EU has often elevated certain moral or ethical questions beyond the national level so as to identify a European ‘correct’ answer to those.25 Moreover, in conjunction with early legislative efforts, the EU has – through its negative integration provisions – also contributed to shaping the emergence of both national and EU lifestyle policies in relation to tobacco, alcohol and diets (Alemanno and Garde 2013, 2014). At the same time, while interpreting EU citizenship provisions, the EU has progressively marginalized the role of Member States as providers of rights in the EU.26 Given this trend of progressive Europeanization of the ‘good life’, the question is therefore whether an EU nudging state might slow down or accelerate this process. The goal-oriented nature of EU law and its wide reliance on purposive legal reasoning make it particularly permeable to behaviourally informed approaches. Also the technocratic character of EU law-making further contributes to its embrace of nudging-type approaches, as those typically do not require the participation of citizens. Yet given the dual democratic legitimacy of the EU – which lies in both representative and participatory democracy – this top-down, technocratic-like intervention might prompt resistance to nudging.

M4716-STRAßHEIM_9781785367847_t.indd 143

25/04/2019 13:32

144  Handbook of behavioural change and public policy

CONCLUSIONS Taken as a whole, our analysis of the rising use of behavioural insights across the EU reveals a growing interest by EU – and some Member States – authorities towards this approach. Despite the limitations highlighted above, several constitutional traits of the EU suggest that its policy-making could not only accommodate but also benefit from the integration of behavioural sciences. First, the goal-oriented drive behind EU policy-making seems particularly prone to empirically driven approaches to law. Second, the fact that EU legislation is initiated by the Commission, a technocratic, non-elected body endowed with sizeable resources and largely protected from immediate political pressure, may be conducive to evidence-based experimentation. Third, the tensions existing between the competence of the EU and that of its Member States, which are typical of any federal system, might benefit from the integration of a body of knowledge capable of injecting fresh empirical guidance on how to draw, interpret and operationalize that polarizing demarcation line. The appeal of behavioural-informed approaches should not lead to underestimate the significance of their effects on the EU legal system. While behavioural considerations may allow policy-makers to consider a broader set of regulatory options and test their effectiveness, their use should be subject to public as well as constitutional scrutiny so as to increase the accountability of the regulatory outcome (Alemanno and Spina 2014). There are indications that the adoption of behavioural insights is proceeding at a speedier rate among policy-makers in some of the countries counting on a strong economy in the continent, such as Germany, the UK, France, the Netherlands or Denmark. Yet wealth distribution does not explain it all. Indeed, some European countries with high GDP per capita, like Luxembourg or Austria, are absent from behavioural policy-making. This potential causal explanation would therefore need to be further studied. Cultural, social, customary and political reasons might be other factors affecting the readiness of policy-makers to take advantage of behavioural insights. Unlike in the United States, the major source of political sensitivity in the European Union does not seem to be the paternalistic nature of the intervention (Alemanno and Sibony 2015). Rather it is the source of that intervention (EU or national). Should behaviourally informed approaches be capable of contributing towards the downsizing of EU action – by limiting its reach or perceived impact on citizens’ daily life – to the benefit of national action, this may pave the way to their success in the European Union. In other words, the fate of nudging in the EU appears as intertwined with its political dimension as it is with its social-scientific one.

NOTES   1. See, the Bratislava Declaration, Informal meeting of the 27 heads of state or government, 16 September 2016.   2. As an illustration, one may think of administrative mechanisms such as the positive silence rule. This amounts to using a default rule by which the inertia of the public administration is presumptively considered indicative that the administration approves a certain behaviour.   3. Adler is identifying the need for policy-makers to account for bounded rationality while drafting legal prescriptions.   4. For a complete and detailed analysis of the several findings of behavioural sciences relevant for regulatory policy, see e.g. Sunstein (2011).

M4716-STRAßHEIM_9781785367847_t.indd 144

25/04/2019 13:32

Nudge and the European Union  145   5. For an overview of regulatory techniques: Baldwin et al. (2001) and Hood et al. Baldwin (2001).   6. The popular literature on behavioural science and public policy is abundant, see, e.g. Thaler and Sunstein (2008); Ariely (2008); Levitt and Dubner (2005); Kahneman (2011); Bazerman and Tenbrunsel (2011) and, lastly, Thaler (2015).   7. This claim is found for example in Jolls and Sunstein (2006, p. 202). Yet this ‘pre-commitment to regulatory tools that preserve choice’, has been pinpointed as its major weakness by Bubb and Pildes (2014). Baldwin, for his part, points out that not all nudges do in fact preserve meaningful choices (Baldwin 2014, p. 836).   8. For an insightful perspective on ‘digital nudging’, see Calo (2014).   9. This distinction is further conceptualized by Sibony and Alemanno (2015). 10. See, e.g. Directive 97/7/EC of the European Parliament and of the Council on the protection of consumers in respect of distance contracts OJ L144, p. 19 (1997). 11. See Article 22 (Additional Payments) of Directive of the European Parliament and of the Council on consumer rights, amending Council Directive 93/13/EEC and Directive 1999/44/EC of the European Parliament and of the Council and repealing Council Directive 85/577/EEC and Directive 97/7/EC of the European Parliament and of the Council, OJ L304 p. 64 (2011). 12. Commission Decision of 16 December 2009 relating to a proceeding under Article 102 of the Treaty on the Functioning of the European Union and Article 54 of the EEA Agreement (Case COMP/C-3/39.530 – Microsoft (tying)). 13. Reference is here made to the E-Privacy Directive 2002/58/EC, the issue relates to whether implied or explicit consent is required for the storing of cookies. Under the previous rules, there was a right to opt-out of the use of cookies, whilst the new rules mandate the need of explicit consent of users in order to use cookies. It is submitted that a third-party’s processing of cookies requires clear legal rules with regard to transparency of the processing and users’ consent. This is more cogent in case of unusual data collection mechanisms as the example of the ‘smart bins’ shows: Miller (2013). 14. See, e.g. the Report on Behaviour Change published by the Science and Technology Select Committee of the UK House of Lords, July 2011; A Practitioner’s Guide to Nudging. Rotman Management Magazine, accessed 16 January 2019 at http://www-2.rotman.utoronto.ca/facbios/file/GuidetoNudging-RotmanMar2013.ashx.pdf. 15. In 2013, the British tax authority HMRC applied some of the insights presented in the paper of the Behavioural Insight Team (BIT 2012) to adopt a variation on the standard letter used to urge taxpayers to file their returns on time. 16. A similar experience emerges in Norway where the Stordalen Foundation launched GreeNudges. 17. The Unit still provides services to the UK government but also to private sector entities and foreign governments. http://www.behaviouralinsights.co.uk/. 18. At the time of writing the European Commission is engaged in a mapping exercise aimed at collecting practices of behavioural policy-making across its Member States. Its outcome is expected to be published in the form of a report by the end of 2015. 19. In Germany, the government announced it was hiring psychologists, behavioural economists as well as anthropologists to test new methods of ‘efficient government’, see Pickert and Beck (2014). 20. See: http://www.modernisation.gouv.fr/les-services-publics-se-simplifient-et-innovent/par-des-services-nu​ meriques-aux-usagers/le-nudge-au-service-de-laction-publique (last accessed 21 October 2014). 21. The report (in Dutch) is available on the WRR website: http://www.wrr.nl/publicaties/publicatie/article/ met-kennis-van-gedrag-beleid-maken/ (last accessed 18 March 2018). 22. Communication of the European Commission accompanying the Better Regulation Package, p. 3. 23. The Toolbox presents a comprehensive array of additional guidance to assist Commission officials in the application of Better Regulation. 24. For a similar perspective, see Alemanno and Spina (2014). 25. Judgment of 18 October 2011, Brüstle, C-34/10, EU:C:2011:669, paragraph 25. 26. Judgment of 11 July 2002, Carpenter, C-60/00, EU:C:2002:434.

REFERENCES Adler, M. (2008), ‘Bounded Rationality and Legal Scholarship’, in M .White (ed.), Theoretical Foundations of Law and Economics, Cambridge: Cambridge University Press, pp. 137–162. Alemanno, A. (2015), ‘How Much Better Is Better Regulation? Assessing the Impact of the Better Regulation Package on the European Union – A Research Agenda’, European Journal of Risk Regulation, 6 (3), 344‒356. Alemanno, A. and G. Garde (2013), ‘The Emergence of an EU Lifestyle Policy: The Case of Alcohol, Tobacco and Unhealthy Diets’, Common Market Law Review, 50 (6), 1745‒1786.

M4716-STRAßHEIM_9781785367847_t.indd 145

25/04/2019 13:32

146  Handbook of behavioural change and public policy Alemanno, A. and A. Garde (2014), Regulating Lifestyle Risks – The EU, Tobacco, Alcohol and Unhealthy Diets, Cambridge: Cambridge University Press. Alemanno, A. and A.L. Sibony (eds) (2015), Nudge and the Law  – A European Perspective, Oxford: Hart Publishing. Alemanno, A. and A. Spina (2014), ‘Nudging Legally: On the Checks and Balances of Behavioural Regulation’, International Journal of Constitutional Law, 12 (2), 429‒456. Ariely, D. (2008), Predictably Irrational: the Hidden Forces that Shape our Decisions, New York: Harper Collins. Baldwin, R. (2014), ‘From Regulation to Behaviour Change: Giving Nudge the Third Degree’, Modern Law Review, 77 (6), 831‒857. Baldwin, R., Cave, M. and M. Lodge (2011), Understanding Regulation. Theory, Strategy, Practice (2nd edn), Oxford: Oxford University Press. Bazerman, M.H. and A.E. Tenbrunsel (2011), Blind Spots: Why We Fail to Do What’s Right and What to Do about It, Princeton: Princeton University Press. Berns, W. (1963), ‘Law and Behavioral Science’, Law and Contemporary Problems, 28 (1), 185‒212. Behavioural Insight Team (BIT) (2012), ‘Applying Behavioural Insights to Reduce Fraud, Error and Debt’, accessed 16 January 2019 at https://www.gov.uk/government/uploads/system/uploads/attachment_data/fi​ le/60539/BIT_FraudErrorDebt_accessible.pdf. Borgesius, F. (2013), ‘Consent to Behavioural Targeting in European Law – What are the Policy Implications of Insights from Behavioural Economics?’, Conference paper for Privacy Law Scholars Conference (PLSC), 6‒7 June 2013, Berkeley, United States. Bovens, L. (2008), ‘The Ethics of Nudge’, in: T. Grüne-Yanoff and S.O. Hansson (eds), Preference Change: Approaches from Philosophy, Economics and Psychology, Dordrecht: Springer, pp. 207–220. Bovens, L. (2012), ‘Real nudge’, European Journal of Risk Regulation, 3 (1), 43‒46. Bubb, R. and R. Pildes (2014), ‘How Behavioral Economics Trims Its Sails and Why’, Harvard Law Review, 127 (6), 1593‒1678. Cabinet Office and Behavioural Insight Team (BIT) (2013), ‘Applying Behavioural Insights to Charitable Givings’, accessed 16 January 2019 at https://www.gov.uk/government/uploads/system/uploads/attachment_ data/file/203286/BIT_Charitable_Giving_Paper.pdf. Calo, R. (2014), ‘Digital Market Manipulation’, George Washington Law Review, 28 (4), 995‒1051. Conly, S. (2013), Against Autonomy – Justifying Coercive Paternalism, Cambridge: Cambridge University Press. Hansen, P.G. and A.M. Jespaersen (2013), ‘Nudge and the Manipulation of Choice: A Framework for the Responsible Use of the Nudge Approach to Behaviour Change in Public Policy’, European Journal of Risk Regulation, 4 (1), 3‒28. Hayden, G. and S. Ellis (2007), ‘Law and Economics After Behavioral Economics’, University of Kansas Law Review, 55 (3), 629‒676. Hood, C., Rothstein, H. and R. Baldwin (2001), The Government of Risk: Understanding Risk Regulation Regimes, Oxford: Oxford University Press. Jolls, C. (2010), ‘Governing America: The Emergence of Behavioural Law and Economics’, Max Weber Lecture Series, 2010/3. Jolls, C. and C.R. Sunstein (2006), ‘Debiasing Through Law’, Journal of Legal Studies, 35 (1), 199‒242. Jolls, C. et al. (1998), ‘A Behavioural Approach to Law and Economics’, Stanford Law Review, 50 (5), 1471‒550. Jones, R., Pykett, J. and M. Whitehead (2013), Changing Behaviours – On the Rise of the Psychological State, Cheltenham, UK and Northampton, MA: Edward Elgar Publishing. Kahneman, D. (2011), Thinking Fast and Slow, New York: Farrar, Straus and Giroux. Kochenov, D. (2013), ‘The Citizenship Paradigm’, University of Groningen Faculty of Law Research Paper No. 08/2013. Levitt, S. and S. Dubner (2005), Freakonomics: A Rogue Economist Explores the Hidden Side of Everything, New York: William Morrow. Lobel, O. and O. Amir (2009), ‘Stumble, Predict, Nudge: How Behavioral Economics Informs Law and Policy’, Columbia Law Review, 108 (8), 2098‒2137. Marteau, T. et al. (2011), ‘Judging Nudging: Can Nudging Improve Population Health?’, British Medical Journal 342: d228, cited in P. Rainford and J. Tinkler (2011), ‘Designing for Nudge Effects: How Behaviour Management Can Ease Public Sector Problems’, accessed 19 December 2018 at http://eprints. lse.ac.uk/37810/. McCrudden, C. (2015), ‘Nudging and Human Dignity’, VerfBlog, 2015/1/06, accessed 16 January 2019 at http:// www.verfassungsblog.de/nudging-human-dignity/. Miller, J. (2013), ‘City of London Calls Halt to Smartphone’s Tracking Bins’, in BBC News Monday 12 August 2013 accessed 16 January 2019 at http://www.bbc.co.uk/news/technology-23665490. Pickert, P. and H. Beck (2014), ‘Kanzlerin Angela Merkel sucht Verhaltensforscher’, FAZ, 26 August 2014, accessed 21 October 2014 at http://www.faz.net/aktuell/wirtschaft/wirtschaftspolitik/kanzlerin-angelamerkel-sucht-verhaltensforscher-13118345.html.

M4716-STRAßHEIM_9781785367847_t.indd 146

25/04/2019 13:32

Nudge and the European Union  147 Rebonato, R. (2012), Taking Liberties – A Critical Examination of Libertarian Paternalism, Basingstoke: Palgrave Macmillan. Selinger, E. and K. Whyte (2011), ‘Is There a Right Way to Nudge? The Practice and Ethics of Choice Architecture’, Sociology Compass, 5 (10), 923‒935. Sibony, A.L. and A. Alemanno (2015), ‘The Emergence of Law and Behavioural Science: A European Perspective’, in A. Alemanno and A.L. Sibony (eds), Nudge and the Law – A European Perspective, Oxford: Hart Publishing, Chapter 1. Sunstein, C.R. (2011), ‘Empirically Informed Regulation’, University of Chicago Law Review, 78 (4), 1349‒1430. Sunstein, C.R. (2013), ‘The Storrs Lectures: Behavioral Economics and Paternalism’, Yale Law Journal, 122 (7), 1826‒1899. Sunstein, C.R. (2014), ‘Nudging: A Very Short Guide’, Journal of Consumer Policy, 37 (4), 583‒588. Sunstein, C.R. (2015), ‘The Ethics of Nudge’, Yale Journal on Regulation, 32(2), 413‒450. Thaler, R.H. (2015), Misbehaving: How Economics Became Behavioral, New York/London: WW Norton. Thaler, R.H. and C.R. Sunstein (2008), Nudge: Improving Decisions about Health, Wealth, and Happiness, New Haven: Yale University Press. Waldron, J. (2014), ‘It’s All for Your Own Good’, New York Review of Books 61, 15, accessed 16 January 2019 at http://www.nybooks.com/articles/2014/10/09/cass-sunstein-its-all-your-own-good/. White, M.D. (2013), The Manipulation of Choice: Ethics and Libertarian Paternalism, Basingstoke: Palgrave Macmillan. Wright, J.D. and D.H. Ginsburg (2012), ‘Behavioral Law and Economics: Its Origins, Fatal Flaws, and Implications for Liberty’, Northwestern University Law Review, 106 (3), 1033‒1090. Yeung, K. (2011), ‘The Regulatory State’, in R. Baldwin, M. Cave and M. Lodge (eds), Oxford Handbook of Regulation, Oxford: Oxford University Press, pp. 80–81. Yeung, K. (2012), ‘Nudge as a Fudge’, Modern Law Review, 75 (1), 122‒148.

M4716-STRAßHEIM_9781785367847_t.indd 147

25/04/2019 13:32

11.  Nudging citizens’ knowledge in knowledge-based EU: the case of breast cancer screening programmes and participatory rights in choice architectures Luca Leone and Mariachiara Tallacchini

1. INTRODUCTION The idea of the law as a ‘process of social engineering’ dates back to the early twentieth century, when US legal scholar Roscoe Pound (1930) defined the law as an instrument for ordering social life while maximizing human interests. In the last decade, behavioural sciences have contributed to a renewed form of social engineering, known as ‘libertarian paternalism’ (Thaler and Sunstein 2003) and bound to promote legitimate goals and citizens’ welfare through specific choice architectures (Thaler et al. 2010). Under the theoretical umbrella of bounded rationality, namely the assumptions according to which individuals lack adequate rational skills in reasoning and performing (Simon 1947; Kahneman et al. 1982), the behaviourally-informed approach to regulation known as ‘nudge’ (Thaler and Sunstein 2008) has been deeply inspiring the normative strategies of several so-called knowledge-based economies, and definitely of the European Union (EU). Implementation of nudging instruments – the wide range of soft quasi-normative provisions encouraging specific individual and collective behaviour1 – are reframing several regulatory domains, despite the perplexities surrounding their overall legitimacy. This chapter deals with the meaning of European citizenship, how it has been politically and legally framed, and its relations to nudging. In the passage from the prevailing economic foundation of the European Communities to their (unfinished) transformation into a political entity, the EU started evoking European citizenship as a major element in legitimizing the related forms of techno-scientific and political innovation (Wynne et al. 2007). At least in principle, the EU vision of citizenship has been committed towards active citizen participation and engagement in the knowledge-based society. However, from the White Paper on European Governance (CEC 2001) to the Better Regulation Agenda (EC 2015), the need to revise and speed up legislation has altered this meaning of citizenship. Although the cognitive and normative challenges of governance call for more democratic citizen participation in the production, control and validation of scientific knowledge, the actual use of knowledge and power in European governance seems to proceed towards a different direction. As nudge was framed having in mind the notion of consumer (Abdukadirov 2016), it fails to fully meet the values connected to the idea of citizenship. Most nudging strategies imply soft, but still top down, paternalism, where pre-identified and often black-boxed values pretend to define individual and collective directions, with little attention paid to wrong assumptions, biases, potential mistakes, unforeseen developments in society, timely corrections. The two cases taken from the domain of health (breast cancer screening) and digital 148

M4716-STRAßHEIM_9781785367847_t.indd 148

25/04/2019 13:32

Nudging citizens’ knowledge in knowledge-based EU  149 privacy allow reflection on how nudging can lie, and also be hidden, at the core of a regulatory framework and not merely facilitate its implementation; how rights themselves can be commodified in the marketplace; how the idea of having rights remains ambiguous if rights protection is somehow imposed rather than offered to citizens. This is especially true when uncertain, controversial knowledge is at the basis of nudge policies, namely when the authoritative choice made by regulators depends on arbitrarily taking for granted the validity/reliability of specific scientific evidence. The approach proposed here consists in looking at choice architectures as a place where nudging can be properly explored and legitimized. To that end, the concepts of Participatory Design and Rights-in-Design are deemed relevant to open up and reframe architectures for public knowledge and normativity. Choice architectures are the place to look at for citizens’ moral and legal entitlements, a place where citizens are invested of an active role in using, experiencing, and controlling their options, powers, privacy, and data. This approach encourages different collaborative forms of knowledge production among a variety of expert and non-expert agents and a path towards renewed trust between citizens and institutions. The idea of ‘starting’ from citizens, who are the subjects and recipients of EU governance, entails important consequences. First, the reference to the term ‘citizen’ rather than ‘consumer’ frames the discourse in a broad political and social perspective, where learning how to collectively contribute to a better society is a primary concern. Second, the importance of involving citizens in both epistemic and normative choice architectures favours a more dynamic way of producing and assessing knowledge (and allowing flaws and mistakes to emerge) and supports a renewed vision of democratic rights. If regulation is increasingly seen as a ‘learning process’ (Von Schomberg and Davies 2010), citizens should participate in designing this process.

2.  LAW AND BEHAVIOURAL SCIENCES Since the early 1960s behavioural knowledge started gaining momentum and relevance in legal disciplines. In the words of Walter Berns, one of the first scholars questioning the role of behavioural science in the legal field, behavioural science may be applied to the law in a variety of contexts: expert testimony, court decisions, legal informatics, problems jurisprudence (Berns 1963, p. 186). With such a scenario, behavioural approaches to the law have become very appealing for several reasons. First, behavioural sciences seem to provide an evidence-based content to human behaviour for normative purposes. Second, the use of behavioural data to maximize economic rationality and efficiency is coherent with the dominant economic analysis of the law, both at the judiciary and legislative level. Moreover, from the perspective of bounded rationality and of the willpower and self-interest that people are supposed to have, behavioural studies run in connection with requests and needs arising from marketing strategies. Furthermore, the behavioural approach does not consist in simple deregulation, as it keeps under normative control substantial goals and ends, so as ‘to modify the implausible elements of rational choice theory and supplement the inadequate elements in order to create a tool with more predictive power in specific situations’ (Korobkin and Ulen 2000, pp. 1074‒1075). Behaviourally-informed regulation can help explaining both effects and contents of the

M4716-STRAßHEIM_9781785367847_t.indd 149

25/04/2019 13:32

150  Handbook of behavioural change and public policy law (positive function); comprehending how law might be used to achieve specific ends (prescriptive function); and assessing the goals and values of the legal system (normative function) (Jolls et al. 1998). Along these assumptions, the relationships between behavioural law and economics, and between behaviourally-informed interventions and society, have been increasingly advancing at the institutional level worldwide,2 impacting institutional structures and dynamics in several domains – employment, consumer protection, environment, taxation, health, neuroscience, and so on (Lourenço et al. 2016) – as well as groups representing categories of subjects and interests. However, if law and policy may legitimately consider experimental psychology and behavioural economics as evidence-based sources of knowledge, the overall issue of legitimate normativity is more complex than the rhetoric of better regulation suggests. Scholars have highlighted a number of difficulties related both to the epistemological soundness of behavioural knowledge and the conditions for its legitimate use for normative purposes. As to the former set of perplexities, the validity and reliability of behavioural knowledge, its variability through cultures and dependence on experimental settings have been questioned. As most behavioural data come from labs, where hypotheses are framed and tested, and are then translated to regulatory interventions (Schwartz 2015), the equivalence between decisions made by experimental subjects and real life agents cannot be taken for granted, and the actual reliability of these ‘translational policies’ (in analogy with the term ‘translational medicine’) should undergo better scrutiny. In addition to the discrepancies between the experimental and the societal validity of behavioural findings, the emotionally-sensitive biases related to the multiple different cultures among European countries are not adequately taken into account. Moreover, the path towards translating empirical data into actual policy interventions is not univocal and straightforward, and is in itself a separate topic of reflection (Feldman and Lobel 2015). As to the latter set of questions, the assumption that behavioural insights as an innovative basis to exercise public powers provide enough legitimacy, being grounded in the self-referential validity of science- and evidence-based knowledge, has been widely criticized: not only these practices fail to ensure a degree of democratic warrants comparable to legislative procedures, but they also leave more room for unchecked political agenda (Amir and Lobel 2012). Indeed, the narrow approach that deals with political issues just as scientific matters entrusted to experts implicitly recognizes science as the only authority in framing and legitimizing human choices. In the light of these difficulties arising in turning behavioural insights in the regulatory process, the concept of nudge has emerged as the mostly used tool in the framework of regulatory policies. In 2008 Thaler and Sunstein proposed and defined nudge as an alternative to traditional regulatory tools, characterized by being inexpensive and choice-preserving, and including three main behavioural tools: default rules, representing a category of tools that let people ‘opt in’ and ‘opt out’ of certain behaviours in many areas; smart disclosure, that is information having to be meaningful, useful and adequate in a given context; simplification measures, meant to provide clear messages to targeted groups about what they are expected to do. This definition has been, however, constantly refined and revised (Hansen 2016). Scholarly work spans from introducing an epistemic distinction between transparent and non-transparent nudges (Hansen and Jespersen 2013) to separating nudge from the dis-

M4716-STRAßHEIM_9781785367847_t.indd 150

25/04/2019 13:32

Nudging citizens’ knowledge in knowledge-based EU  151 tinct concepts of behavioural economics (Lunn 2014) and behavioural insights (Lourenço et al. 2016). Critics have argued about the concept of libertarian paternalism (Rebonato 2012; White 2013; Hurd 2016); the ethics of nudge (Bovens 2008) and its philosophical grounds (Conly 2012; Baldwin 2014). These criticisms, though, represent just a premise to the complex relationships between nudge and the law. On the one side, nudging not always meets the law (Sibony and Alemanno 2015; Alemanno and Spina 2014), but legal difficulties involved in nudging are related to constitutional questions of authority and accountability (McCrudden and King 2016); others raise the potential infringements of individual rights (van Aaken 2015); further concerns address a series of ‘gaps’ behavioural approach encounters between basic research findings and some of the empirical evidence needed for a fully effective legal analysis (Tor 2008). The legal use of behavioural evidence involves also difficulties concerning the ecological validity of experiments, the temporal dimension of influencing mechanisms, the heterogeneity problem (namely, the degree to which individuals display various biases), as well as the partial view of branches of behavioural research, and the granularity gap between the general character of legal rules and the context-dependence of behavioural knowledge (Alemanno and Sibony 2015). Although these perplexities are far from having been solved, since its start in 2014 the Juncker Commission has strongly supported behavioural strategies for policy as more targeted and efficient solutions at all stages of EU policy cycle, from designing to implementing EU regulations (Lourenço et al. 2016).

3.  REGULATING AND NUDGING IN THE EU 3.1  Innovative Knowledge, Normativity and Citizenship in the EU As shown by scholarly work (Straßheim et al. 2015) the rise of behavioural science in the framework of the EU policy agendas found its roots in two main factors, that is, the combination of the persuasiveness of experimental insights with the political practicality of heuristics and biases, and the re-definition of the role of citizens to improve their relation with public authorities. Both these factors have played a crucial role in pushing towards the process of rethinking of EU governance. The EU legal framework has increasingly looked at knowledge as a more objective and, therefore, shared ground to frame and inform legitimate regulatory processes. The unfolding uncertainties coming from complex, knowledge-based domains (such as health and the environment as well as emerging technologies) have required substantial adjustments in the way regulators make their choices in the face of ill-defined risks (van Asselt et al. 2014; Hilgartner et al. 2015). As a consequence, the law has been more often framed and presented as a ‘learning process’ that has to become smart(er) to cope with constant changes (Von Schomberg and Davies 2010; Rübig 2012). This growing trend in the vision of the law as essentially dependent on knowledge, and even as a learning process itself, correlates to the idea that contemporary democratic societies are and have to be knowledge-based. And although the role played by science and technology as drivers of social development has been crucial in this respect, this perspective is not new, but deeply rooted in modern legal theory as a rational foundation for normative power. Indeed, theoretical assumptions about the self-evident rationality of the law (and of

M4716-STRAßHEIM_9781785367847_t.indd 151

25/04/2019 13:32

152  Handbook of behavioural change and public policy the legislator: ‘la rationalité ou volonté du legislateur’) as well as the substantive and procedural knowledge requirements for the law to become binding (being approved, enacted, published, etc.) (Fuller 1964) highlight the cognitive and rational elements legitimizing norms. In the theorized modern State citizens trust legislators because of this shared rationality embedded in the social contract. The procedural requirements for legitimate legislations to be enacted are mostly stated at constitutional level as the highest foundational place for citizens’ shared values. Even more in democratic knowledge-based societies this implies that not only should citizens have access and contribute to create knowledge, but that they should also be entitled to legitimize new forms of normativity such as innovative forms of knowledge-based regulatory tools – at least through the legislative power. Indeed, since the early 1990s the notion of European citizen and citizenship has emerged and has been framed in close connection with issues of techno-scientific innovation and social and normative processes. The context has been that of the initial path toward European political integration and the Maastricht Treaty (1993) and the controversial deployment of biotechnology, with both processes requiring some forms of public legitimization. In 1991, well before the Charter of Fundamental Rights and the European Citizen Initiative were in place, while engaging in the construction of a successful market for biotechnology (CEC 1991), the Commission called for the need to take into account European citizens’ values in order to fully legitimize what was perceived as an unprecedented technological revolution. Even though ethics experts were eventually identified as those who could adequately represent/surrogate citizens in clarifying the values at stake (Tallacchini 2015), the link between citizenship and legitimation of innovative knowledge and normativity has been repeatedly confirmed. In 2002 when, in order to overcome the democratic deficit and to regenerate trust towards European institutions, the Commission launched the Science and Society Action Plan, the message was that: ‘In a knowledgebased society, democratic governance must ensure that citizens are able to make informed choice from the options made available to them by responsible scientific and technological progress’ (EC 2002, p. 3). This goal was deemed to be achieved through ‘the participation of citizens and civil society in the formulation and implementation of science policies in Europe, and the use of scientific knowledge complying with common ethical rules in the formulation of responsible policies’ (EC 2002, p. 3). And the very same message was conveyed through the White Paper on European Governance. However, despite having been framed around the issues of legitimizing innovative knowledge and normativity, the concept of European citizenship and its role has always been ambiguously implemented. The EU institutions constantly swing between top-down and bottom-up policies, paternalistic and empowering approaches, citizens as subjects of rights and objects of concerns, and the invariable tendency to flatten civic choices on consumer preferences and the needs of the market. The use of behavioural sciences to inform novel nudging regulatory tools represents a form of innovation on the side of both knowledge and normativity, and should find its legitimacy along these lines of respect and empowerment of citizens. 3.2  Nudging in Breast Cancer Screening and Digital Privacy: Two Cases The two topics explored below concern breast cancer screening programmes and privacy/ data protection. The reasons for looking at the sectors of health and private life are

M4716-STRAßHEIM_9781785367847_t.indd 152

25/04/2019 13:32

Nudging citizens’ knowledge in knowledge-based EU  153 manifold, including that both cases have been widely explored in nudge literature. The legal framing of several health and privacy issues has been increasingly informed by the idea that patients and users have to be seen as citizens, with an approach described as citizen–patient and citizen–user centred. Moreover, in both domains the individual right to obtain all relevant information and to provide free consent did represent a major step against paternalism and imbalance of knowledge and power between experts and lay citizens. And in both health and privacy, and now in their networked relations, it has become clear that informed consent is not always the optimal solution to effective citizens’ empowerment. Paternalism, if ever overcome, is re-entering the scene in the name of expert behavioural knowledge and pre-identified values and goals that are not given to citizens, but softly applied to them. The health domain increasingly appears as full of U-turns, where patients/citizens have been nudged towards certain knowledge and practices later proved wrong, from vitamins to antibiotics to hormone replacement therapy (Sukel 2016). The current trend toward nudging people to increase participation in clinical trials, for instance, is troublesome (van Epps et al. 2016), not only because it diminishes decades of work to make informed consent more effective, but also because the assessment of risks and benefits by ethics committees has often proved unreliable (Meyer 2013). As to privacy, the attempt to interpret and simplify all issues related to Information and Communications Technology (ICT) as privacy concerns not only prevented a more complex understanding of issues surrounding ICT (lack of agency, forms of unlearning), but also delayed an adequate reflection on them (agency, learning). In the cases presented here nudging lies at different levels. Breast cancer screening policies nudge the underlying knowledge, instead of recognizing the open controversies among researchers and involving citizens in a more direct dialogue. In digital privacy, the right to online private life is implemented in ambiguous paternalistic ways. Not only it is not designed as an active right/responsibility, but its implementation does not invest in citizens’ education and creates, instead, a market for privacy products. 3.2.1  Nudging in breast cancer screening Cancer screening programmes, and especially breast cancer population tests, represent an interesting example for nudging strategies which have been widely applied but have also raised serious criticisms. Despite having been enthusiastically adopted by health systems, from the United States to the European Union, breast cancer screening programmes keep raising debates about their actual impact on reducing mortality, risks of over-diagnosis and unnecessary or harmful treatments. Over-diagnosis occurs when mammography detects small tumours that may never affect patients’ health during their lifetime. In challenging the validity of screening programmes, these controversies also affect their efficacy. Nudging – primarily taking the form of automatic enrolment and pre-scheduled appointments – has therefore been seen as a potentially useful tool in increasing participation, even though the extent of its actual impact remains ambiguous and problematic. Breast screening programmes have been – and continue to be – the fulcrum of inexhaustible quérelles, spanning from their effectiveness on mortality reduction rate and on individual and collective prevention of diseases, to communication methods and practices, up to the benefits and risks of trials (Gøtzsche and Jørgensen 2013; Gøtzsche 2012; Welch 2012). In recent years, reviews of published work have been conducted and new

M4716-STRAßHEIM_9781785367847_t.indd 153

25/04/2019 13:32

154  Handbook of behavioural change and public policy guidelines proposed, showing divergent positions and results. In 2012, the analysis of the UK Panel on Breast Cancer Screening, led by Sir Michael Marmot, strongly stated that ‘breast screening programme extends life’ as it leads to a 20 per cent relative risk reduction of breast cancer mortality compared with no screening. The high rate of over-diagnosis still remained an open issue, with largely divergent interpretation of data (Marmot et al. 2013). But, in spite of the ongoing controversies among researchers, The Lancet suggested that the UK report would lead to ‘closing a chapter’ on the matter (The Lancet 2012). In the same year, the Nordic Cochrane Centre (Gøtzsche et al. 2012) proposed, instead, a different and very critical perspective on mammography, by ascribing mortality reduction to the availability of more effective therapies rather than for screening practices. In the face of this view, though, harms from over-diagnosis would remain extremely high and hardly reducible: ‘by avoiding to going to screening’, it was suggested, ‘a woman will lower her risk of getting a breast cancer diagnosis’ (Gøtzsche et al. 2012, p. 3). In 2015, the Guideline Update on screening mammography from the American Cancer Society (ACS) (Oeffinger et al. 2015; Myers et al. 2015) reaffirmed the evidence and the accuracy of the reduced mortality estimates, suggesting also moving up the screening entry age. The document has thus opened the debate between the ACS and the Nordic Cochrane Centre, which denied the scientific validity of the ACS position: ‘screening has not reduced total mortality, and therefore it is misleading to claim that “screening saves lives”’ (Jørgensen and Gøtzsche 2016). Meanwhile, in 2013 the Swiss Medical Board published a controversial report – defined as ‘unethical’ by several scientific organizations – advising the Swiss government not to introduce more screening programmes and to set a time limit on the existing ones. These recommendations were justified in the light of the erroneous results deriving from screening, unbalanced by a favourable cost-effectiveness ratio – though women generally overestimate the benefits of mammography (Biller-Andorno and Jüni 2014). In 2014, the analysis of the data produced in 25 years by the Canadian National Breast Screening Study did not detect a mortality reduction in women aged 40-59 undergoing annual mammography controls beyond that of physical examination, while showing a high percentage of cases of over-diagnosis (Miller et al. 2014). In 2017, a new study by Danish researchers concluded that breast cancer screening is not associated with a reduction in the incidence of advanced cancer and calculated that one in three breast tumours detected in women aged 50 to 69 was probably over-diagnosed (Jørgensen et al. 2017). The existing uncertainties and divisions within the scientific community are certainly reflected in the often high percentage of abandonment of screening from women, while the reduced rate of participation has negatively impacted, according to some scholars, the effectiveness and quality of the health system (Camilloni et al. 2013). In this scenario, nudging appeared as an important and strategic approach which improved the efficiency and effectiveness of screening programmes (Camilloni et al. 2013; Codagnone et al. 2014; Purnell et al. 2015). Nudge, however, has proved ambiguous – even though cultural differences and the need to accelerate specific educational processes in socially disadvantaged populations remains a topic of reflection (Purnell et al. 2015). According to Ploug et al. (2012), who have investigated the influence of nudging on the Danish mammography screening programme, libertarian paternalism not only did not help, but has also damaged the credibility of screening. While all choice architecture strategies – ‘opt-out’ rules, limited information on

M4716-STRAßHEIM_9781785367847_t.indd 154

25/04/2019 13:32

Nudging citizens’ knowledge in knowledge-based EU  155 benefits and risks, the Danish National Board of Health’s authoritative favour for screening (without providing compelling arguments), the absence of reference to critical positions – were expected to increase the efficiency of screening programmes, the results have been delusive and not higher than knowledge- and information-based campaigns (Camilloni et al. 2013). Indeed, the costs paid by women – in terms of psychological damages (anxiety, depression), information uncertainty, over-diagnosis and unnecessary surgical ­interventions – remain substantial and hardly justify the rationality of women’s participation within a welfare enhancing soft paternalistic framework. European institutions have shown their political commitment towards implementation of population-based screening programmes for cervical, breast, and colorectal cancers, by creating a comprehensive policy framework (EP 2006; CEC 2009; EC 2014) to support Member States in their efforts to tackle cancer. While acknowledging, already in 2003,3 that ‘screening can also have negative side effects for the screened population’ (Whereas 9), and that ‘it is an ethical, legal and social prerequisite that cancer screening should only be offered to fully informed people with no symptoms if the screening is proved to decrease disease-specific mortality, if the benefits and risks are well known, and if the cost-effectiveness of the screening is acceptable’ (Whereas 23), the EU institutions have adopted a precise epistemic position by stating that ‘evidence exists concerning the efficacy of screening for breast cancer … derived from randomised trials’ (Whereas 8). This un-nuanced epistemic stance per se introduces a form of nudged knowledge as no reasons are offered about which studies can be considered reliable and decisive, thus anticipating a general conclusion that is not merely technical but normative (Rizzo and Whitman 2009). The fourth edition of the EU guidelines does not focus so much on scientific aspects, but rather on the overall ‘quality’ of the process as a crucial factor for success (Lynge et al. 2012; Perry et al. 2013; Deandrea et al. 2015). The legitimacy of public decisions in a framework of structural uncertainty – common to many science-based policy areas – is increasingly built not as much on an alleged incontestability of the evidence, but rather on the reliability of the entire techno-scientific and decision-making process. However, this approach does not clarify if citizens are deemed to enter the process as subjects or as ‘nodes’ of behavioural data. 3.2.2  Nudging in EU digital privacy: from the right to the product Nudging lies at the core of Information and Communication Technology (ICT) as digital architectures embody rules and decisions in their own hard and soft structures as ‘factual normativity’ (Hildebrandt 2008). Already in 1980, in Do Artifacts have Politics? Langdon Winner highlighted that all machines, structures and technical systems should be assessed ‘for the ways in which they can embody specific forms of power and authority’ (Winner 1980, p. 121). These early observations (that have led to a number of developments in ICT, e.g. to make them more ‘human-centred’), have raised awareness about the choices implicitly embedded, packed, and black-boxed in programmes and devices, showing that ‘architecture matters’ (Kroes 2011), namely that ICT structures do not only have ethical and policy impacts, but have built-in values and choices that should be opened-up and unpacked. Nudging as a policy strategy, however, has formally entered the domain of ICT law through privacy and data protection. In the 1990s Ann Cavoukian, Information and Privacy Commissioner of Ontario (Canada), launched the idea of Privacy-by-Design

M4716-STRAßHEIM_9781785367847_t.indd 155

25/04/2019 13:32

156  Handbook of behavioural change and public policy (PbD) as an organization’s default mode of operation, a form of technological normativity for safe digital behaviour.4 In 2010, during the process leading to European Regulation 2016/679, the General Data Protection Regulation (GDPR),5 the European Data Protection Supervisor (EDPS) supported the idea that privacy should be engineered by-default and by-design, namely ‘embedded within the entire life cycle of the technology, from the very early design stage, right through to their ultimate deployment, use and ultimate disposal’ (EDPS 2010, p. 2). Moreover, PbD was proposed – which is Article 25 of the GDPR – as a general binding principle. Insights from behavioural economics literature casting light on individuals’ inconsistencies in keeping or sharing personal information (Acquisti et al. 2007) have contributed new instruments to nudging Internet users towards desired behaviours. Examples range from visual representations of the mobile app’s privacy rating – designed to nudge people away from privacy-invasive apps – to privacy settings impacting users’ ability to control how their information is shared; from software tools to prevent excessive information disclosure, to location sharing applications allowing users to control the conditions under which they make their location visible to others (Lai and Hui 2006; Choe et al. 2013). Data protection provisions have been widely influenced by behavioural sciences insights, also revealing the limits of informed consent (e.g. in the Internet of Things). Article 5 (3) of the so-called Cookies Directive6 introduced new methods of providing information and offering the right to refuse, in the view of the different objectives that information storage might involve, as well as ‘standard forms for specific methods to obtain verifiable consent’ (Article 85b); and GDPR Articles 35 and 36 highlighted, through the Data Protection Impact Assessment, the behavioural implications of processing personal data, such as the risks of profiling (Whereas 75 and Article 22) (Carolan and Spina 2015). Most nudging ‘by-design’ has the advantage to provide protection without requiring users to become more knowledgeable or responsible as to their personal data. However, a major way protection is delivered is through privacy products – so-called PET, Privacy Enhancing Technology. Through the by-design approach privacy has become a commodified good in the market: the ‘privilege of privacy’ for competitive advantage is sprouting up all over the digital economy, with products that are black-boxes. In this scenario where privacy is already feeding a profitable market, GDPR constitutes a unique instrument in legitimizing commodified privacy in the name of the ‘principle’ of PbD. Indeed, if privacy and data protection are listed as fundamental rights within the EU (Articles 7 and 8 of the Charter of Fundamental Rights of the European Union), the architecture of their digital protection is built more on the virtues of the market rather than on providing citizens with digital education and skills. The impression, again, is that in the EU social changes are addressed by creating new markets, while citizens are still primarily framed as consumers. 3.3  Nudging: from Improving to Co-designing Choice Architectures The relevance of open and transparent knowledge as a condition for legitimate nudging has been widely recognized. Nudge instruments are less reluctantly accepted even by critics when they are fully transparent and effective, if their rationale is not hidden, and if they do not limit freedom of choice (Hansen 2016). Some authors have gone farther, highlighting that full available knowledge and not just predefined, though transparent, information

M4716-STRAßHEIM_9781785367847_t.indd 156

25/04/2019 13:32

Nudging citizens’ knowledge in knowledge-based EU  157 is a pre-requisite for nudging. Floridi’s proposal of a ‘pro-ethical design’, a system capable of influencing choices, while remaining fully respectful of freedom, goes in this direction. Instead of shaping the actual options available to an agent and relying on human inertia and biases as the pragmatic motivation for action, information on the available options is provided with no suggestion, thus encouraging agents to ‘empower themselves’ through their informational choices. Floridi’s theory favours educating agents to make their own judgments and to assume responsibilities (Floridi 2015). The proposed system, however, does not engage with how desirable outcomes are defined, which knowledge has been deemed as relevant, and how the overall design of choice architecture works. A suggestion to improve current forms of nudging may come from the digital domain through the concepts of Rights-in-Design and Participatory Design, both centred on opening up digital architectures to make citizens/users more actively involved in producing and validating knowledge not only in relation to specific issues, but in the construction of the architecture itself. The idea of Privacy (as a right) in Design refers to ‘raising awareness about the processes through which values and norms become embedded in technological architectures. Privacy in design looks at the normativity of structural choices in an effort to promote transparency and protect rights and values of the citizens’ (EGE 2014, p. 32). Besides framing privacy more as a right rather than a protected value, the in-design approach can equally apply to other digital rights such as the (increasingly claimed) right of access to raw data (Lunshof et al. 2014). While still consistent with a market perspective, ICT products enhancing free choices may incorporate a human and civic dimension. As Kounelis et al. have pointed out, ‘(s)imply described, the development of software encompasses necessarily a front end, commonly described as a user interface and the (often black) “box” that it accesses.… Black-boxing needs to be opened up through structural spaces to allow individual personal decisions to be taken’ (Kounelis et al. 2014, p. 75). To this end a variety of normative and educational measures can be adopted, with engineers and information systems engineers working together with ethicists, lawyers and citizens to build collective transdisciplinary knowledge at the interface between technology and normativity, in a continuous process between agents and devices constantly redefined, renegotiated, renovated, and quantified when appropriate. As technologies are increasingly in charge of normative functions, through an ‘indesign’ approach digital architectures and their design become new spaces for rights to be actively used and choices to be made. This approach has the advantage of contributing to more trusted digital relationships, facilitating new learning experiences and skills. Access to technical and communication means to design, modify and create an artefact (object, system, application, etc.) allows for a greater variety of options and choices to be made regarding, for instance, personal health issues, environmental pollution or information about policy decisions. Even more, the concept of Participatory Design (PD) can prove crucial when institutions and citizens interact, from how institutional information is delivered to how laws are implemented. Theories of PD originated in the domain of Human-Computer Interactions (HCI), information systems and socio-informatics, especially in relation to working places (Kensing and Blomberg 1998; Simonsen and Robertson 2012); they reflect on building digital architectures through participatory procedures in order to control how choices and decisions are introduced and implemented in technological systems, with the aim of making knowledge and values embedded in technological systems more open and

M4716-STRAßHEIM_9781785367847_t.indd 157

25/04/2019 13:32

158  Handbook of behavioural change and public policy democratically shared. In PD users are conceived of ‘as co-designers during all stages of the design process … and have a say on how their future activities with the new technical solution will be’; in this context openness ‘means that decisions about possible design trajectories should … be open to the possibility of change and in ways that enable choices to be unmade or changed’ (Bratteteig and Wagner 2014, pp. 3, 6). As an interpretive concept Participatory Design suggests that, as agents of legitimization for innovative knowledge and normativity, citizens should be involved in structuring choice architectures. However, as law and policy are increasingly reframed and implemented through ICT and digitalization, scholars from the fields of informatics, policy and law claim that the public sphere and relations between institutions and citizens should be redesigned along these lines as a networked ‘common’ (Björgvinsson et al. 2010; Ostrom et al. 2012; Marttila et al. 2014). Participatory design projects already exist around the world as new ways to digitally implement specific policies, especially in relation to public services and in local administrations (Bratteteig and Wagner 2014; Teli et al. 2015). Reflecting on nudging in the light of Participatory Design means that all its practices should be thoroughly rethought and legitimized by citizens. Participatory Design blends the two instances both of integration between different forms of knowledge and redistribution of decisional power. Arguably, implementing and embedding these requirements into behaviourally-inspired initiatives will require a radical shift for most forms of nudging, in the view of a more democratically-shared vision focusing on promoting learning, human agency and participatory choices, while also offering the free choice of being nudged.

4.  SOME FINAL REFLECTIONS This chapter reflects on nudging and choice architecture in the light of how the concept of European citizenship has been designed. Since the beginning of the passage between the economic and the political European communities citizens have been seen as an essential component in legitimizing innovation both as knowledge and normativity. This role has been never fully given to European citizens and remained mostly rhetorical; however, the pressures to speed-up governance in Europe have led to new regulatory soft practices, such as nudge, that can deeply threaten the idea of citizenship. Two examples in the domain of health and digital privacy, where citizens are treated as ‘objects of concern and control’ rather than ‘subjects of rights’ reveal some of these ambiguities. The proposed approach in order to rethink nudging as a more legitimate form of normative innovation is inspired by the notions of Participatory Design and Rights-in-Design as ways to open up and reframe choice architectures. Behaviourally-based actions can deeply benefit from new normative practices of Participatory Design, thus gaining more robust epistemic and democratic credibility.

NOTES 1. According to Hansen, ‘A nudge is a function of (I) any attempt at influencing people’s judgment, choice or behaviour in a predictable way (1) that is made possible because of cognitive boundaries, biases, routines and habits in individual and social decision-making posing barriers for people to perform rationally in their

M4716-STRAßHEIM_9781785367847_t.indd 158

25/04/2019 13:32

Nudging citizens’ knowledge in knowledge-based EU  159

2. 3. 4. 5. 6.

own declared self-interests and which (2) works by making use of those boundaries, biases, routines, and habits as integral parts of such attempts’ (Hansen 2016, p. 4). In the US context, see, for example Sunstein (2010) and http://www.eenews.net/assets/2015/09/16/docu​ ment_cw_02.pdf (accessed 29 September 2016). For an overview of the activities carried out in Europe, see http://is.jrc.ec.europa.eu/pages/BE/BEindex.html (accessed 29 September 2016). See Council Recommendation 2003/878/EC of 2 December 2003 on Cancer Screening. https://www.iab.org/wp-content/IAB-uploads/.../fred_carter.pdf (accessed 30 September 2016). Regulation (EU) 2016/679 of the European Parliament and of the Council of 27 April 2016 on the protection of natural persons with regard to the processing of personal data and on the free movement of such data, and repealing Directive 95/46/EC (General Data Protection Regulation). Directive 2009/136/EC of the European Parliament and of the Council of 25 November 2009 amending Directive 2002/22/EC on universal service and users’ rights relating to electronic communications networks and services, Directive 2002/58/EC concerning the processing of personal data and the protection of privacy in the electronic communications sector and Regulation (EC) No 2006/2004 on cooperation between national authorities responsible for the enforcement of consumer protection laws.

REFERENCES Abdukadirov, S. (ed.) (2016), Nudge Theory in Action. Behavioral Design in Policy and Markets, London: Palgrave Macmillan. Acquisti, A., S. Gritzalis, C. Lambrinoudakis and S. De Capitani di Vimercati (eds) (2007), Digital Privacy: Theory, Technologies and Practices, London: Taylor and Francis Group. Alemanno, A. and A. Sibony (eds) (2015), Nudge and The Law. A European Perspective, Oxford and Portland, OR: Hart Publishing. Alemanno, A. and A. Spina (2014), ‘Nudging Legally. On the Checks and Balances of Behavioral Regulation’, International Journal of Constitutional Law, 12 (2), 429‒456. Amir, O. and O. Lobel (2012), ‘Liberalism and Lifestyle: Informing Regulatory Governance with Behavioral Research’, European Journal of Risk Regulation, 3(1), 17–25. Baldwin, R. (2014), ‘From Regulation to Behavior Change: Giving Nudge the Third Degree’, Modern Law Review, 77 (6), 831‒857. Berns, W. (1963), ‘Law and Behavioral Science’, Law and Contemporary Problems, 28, 185‒212. Biller-Andorno, N. and P. Jüni (2014), ‘Abolishing Mammography Screening Programs? A View from the Swiss Medical Board’, New England Journal of Medicine, 370, 1965‒1967. Björgvinsson, E., P. Ehn and P.A. Hillgren (2010), ‘Participatory Design and “Democratizing Innovation”’, Proc. PDC, 41‒50. Bovens, L. (2008), ‘The Ethics of Nudge’, in T. Grüne-Yanoff and S.O. Hansson (eds), Preference Change: Approaches from Philosophy, Economics and Psychology, Berlin and New York: Springer, pp. 207‒19. Bratteteig, T. and I. Wagner (2014), Disentangling Participation: Power and Decision-making in Participatory Design, Dordrecht and London: Springer. Camilloni, L., E. Ferroni, B.J. Cendales et al. (2013), ‘Methods to Increase Participation in Organized Screening Programs: A Systematic Review’, BMC Public Health, 13, 464‒479. Carolan, E. and A. Spina (2015), ‘Behavioural Sciences and EU Data Protection Law: Challenges and Opportunities’, in A. Alemanno and A. Sibony (eds), Nudge and The Law. A European Perspective, Oxford and Portland, OR: Hart Publishing, pp. 161‒178. Charter of Fundamental Rights of the European Union (2012), 2012/C 326/02, OJ C 326/391. Choe, E.K., J. Jung, B. Lee and K. Fisher (2013), ‘Nudging People Away From Privacy-Invasive Mobile Apps Through Visual Framing’, in P. Kotzé, G. Marsden, G. Lindgaard, J. Wesson and M. Wincklert (eds), Human– Computer Interaction – INTERACT 2013, Part III, LNCS 8119, Berlin and New York: Springer, pp. 74–91. Codagnone, C. et al. (2014), ‘The Challenges and Opportunities of Nudging’, Journal of Epidemiology and Community Health, 68 (10), 909–911. Commission of the European Communities (CEC) (1991), Commission Communication to Parliament and the Council. Promoting the Competitive Environment for the Industrial Activities Based on Biotechnology within the Community, Commission of the European Communities SEC (91) 629 final, Brussels. Commission of the European Communities CEC (2001), European Governance. A White Paper, COM(2001) 428 final, Brussels. Commission of the European Communities CEC (2009), Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee of the Regions on Action Against Cancer: European Partnership, COM(2009) 291/4, Brussels.

M4716-STRAßHEIM_9781785367847_t.indd 159

25/04/2019 13:32

160  Handbook of behavioural change and public policy Conly, S. (2012), Against Autonomy: Justifying Coercive Paternalism, New York: Cambridge University Press. Deandrea, S., C. Freeman, J. Lopez-Alcade et al. (2015), Review and Analysis of External Quality Assessment of Breast Cancer Services in Europe, JRC Science and Policy Report, Luxembourg: Publications Office of the European Union. European Commission (EC) (2002), Science and Society Action Plan, Luxembourg: Office for Official Publications of the European Communities. European Commission (EC) (2014), Report from the Commission to the European Parliament, the Council, The European Economic and Social Committee and the Committee of the Regions. Implementation of the Communication from the Commission, from 24 June 2009, on Action Against Cancer: European Partnership [COM (2009) 291 final] and Second Implementation Report on the Council Recommendation of 2 December 2003 on Cancer Screening (2003/878/EC), COM(2014) 584 final, Brussels. European Commission (EC) (2015), Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions. Better Regulation for Better Results-An EU agenda, COM(2015) 251 final, Strasbourg. European Data Protection Supervisor (EDPS) (2010), Opinion on Promoting Trust in the Information Society by Fostering Data Protection and Privacy (Opinion on Privacy By Design), OJ C 280, 16.10.2010, 1–15. European Group on Ethics in Science and New Technologies (EGE) (2014), Opinion 28 of the European Group on Ethics in Science and New Technologies, Ethics of Security and Surveillance Technologies, Brussels. European Parliament (EP) (2006), European Parliament Resolution on Breast Cancer in the Enlarged European Union, P6_TA(2006)0449. Feldman, Y. and O. Lobel (2015), ‘Behavioural Trade-Offs: Beyond the Land of Nudges Spans the World of Law and Psychology’, in A. Alemanno and A. Sibony (eds), Nudge and The Law, Oxford and Portland, OR: Hart Publishing, pp. 301‒324. Floridi, L. (2015), ‘Tolerant Paternalism: Pro-ethical Design as a Resolution of the Dilemma of Toleration’, Science and Engineering Ethics, 21, 1‒20. Fuller, L.L. (1964), The Morality of Law, New Haven CT: Yale University Press. Gøtzsche, P.C. (2012), Mammography Screening: Truth, Lies and Controversy, London: Radcliffe Publishing. Gøtzsche, P.C. and K.J. Jørgensen (2013), ‘Screening for Breast Cancer with Mammography’, Cochrane Database of Systematic Reviews, 6, CD001877. Gøtzsche, P.C., O.J. Hartling, M. Nielsen and J. Brodersen (2012), Screening for Breast Cancer with Mammography, The Nordic Cochrane Centre (2nd edition), accessed 8 January 2017 at http://nordic.cochrane.org/sites/nordic. cochrane.org/files/public/uploads/images/mammography/mammography-leaflet.pdf. Hansen, P.G. (2016), ‘The Definition of Nudge and Libertarian Paternalism: Does the Hand Fit the Glove?’, European Journal of Risk Regulation, 1, 1‒20. Hansen P.G. and A.M. Jespersen (2013), ‘Nudge and the Manipulation of Choice. A Framework for the Responsible Use of the Nudge Approach to Behaviour Change in Public Policy’, European Journal of Risk Regulation, 1, 3‒28. Hildebrandt, M. (2008), ‘Legal and Technological Normativity: More (and Less) than Twin Sisters’, TECHNE 12 (3), 169‒183. Hilgartner, S., C.A. Miller and R. Hagendijk (eds) (2015), Science and Democracy: Making Knowledge and Making Power in the Biosciences and Beyond, New York and London: Routledge. Hurd, H.M. (2016), ‘Fudging Nudging: Why ‘Libertarian Paternalism’ is the Contradiction It Claims It’s Not’, Georgetown Journal of Law and Public Policy, 14 (Special Issue), 703‒734. Independent UK Panel on Breast Cancer Screening (2012), ‘The Benefits and Harms of Breast Cancer Screening: An Independent Review’, The Lancet, 380 (9855), 1778–1786. Jolls, C., C.R. Sunstein and R.H. Thaler (1998), ‘A Behavioral Approach to Law and Economics’, Stanford Law Review, 50, 1471‒1550. Jørgensen, K.J. and P.C. Gøtzsche (2016), ‘Breast Cancer Screening: Benefit or Harm? Comment’, JAMA, 315 (13), 1402. Jørgensen, K., P. Gøtzsche, M. Kalager, et al. (2017), ‘Breast Cancer Screening in Denmark: A Cohort Study of Tumor Size and Overdiagnosis’, Ann Intern Med, accessed 9 October 2017 at http://www.med.uio.no/helsam/ forskning/grupper/klinisk-effektforskning/aktuelle-saker/2016/annals.pdf. Kahneman, D., P. Slovic and A. Tversky (eds) (1982), Judgment under Uncertainty: Heuristics and Biases, New York: Cambridge University Press. Kensing, F. and J. Blomberg (1998), ‘Participatory Design: Issues and Concerns’, Computer Supported Cooperative Work, 7, 167–185. Korobkin, R.B. and T.S. Ulen (2000), ‘Law and Behavioral Science: Removing the Rationality Assumption from Law and Economics’, California Law Review, 88 (4), 1074‒1075. Kounelis, I., G. Baldini, R. Neisse, G. Steri, M. Tallacchini and Â.G. Pereira (2014), ‘Building Trust in Human– Internet of Things Relationship’, IEEE Technology and Society Magazine, 73‒80. Kroes, N. (2011), Internet Essentials, OECD High Level Meeting on the Internet Economy, Paris.

M4716-STRAßHEIM_9781785367847_t.indd 160

25/04/2019 13:32

Nudging citizens’ knowledge in knowledge-based EU  161 Lai, Y.L. and K.L. Hui (2006), ‘Internet Opt-In and Opt-Out: Investigating the Roles of Frames, Defaults and Privacy Concerns’, in Proceedings of the 2006 ACM SIGMIS CPR Conference on Computer Personnel Research, pp. 253–263. Lourenço, J.S., E. Ciriolo, S.R. Almeida and X. Troussard (2016), Behavioral Insights Applied to Policy: European Report 2016, JRC Science Hub, accessed 29 September 2016 at http://publications.jrc.ec.europa.eu/ repository/bitstream/JRC100146/kjna27726enn_new.pdf. Lunn, P. (2014), Regulatory Policy and Behavioural Economics, OECD Publishing, accessed 13 December 2017 at http://www.oecd-ilibrary.org/governance/regulatory-policy-and-behavioural-economics_9789264207851-en. Lunshof, J.E., G.M. Church and B. Prainsack (2014), ‘Raw Personal Data: Providing Access’, Science, 343 (6169), 373‒374. Lynge, E., S. Törnberg, L. von Karsa, N. Segnan and J.J. van Delden (2012), ‘Determinants of Successful Implementation of Population-based Cancer Screening Programmes’, European Journal of Cancer, 48 (5), 743‒748. Marmot, M.G., D.G. Altman, D.A. Cameron, J.A. Dewar, S.G. Thompson and M. Wilcox (2013), ‘The Benefits and Harms of Breast Cancer Screening: An Independent Review’, British Journal of Cancer, 108 (11), 2205‒2240. Marttila, S., A. Botero and J. Saad-Sulonen (2014), ‘Towards Commons Design in Participatory Design’, in International Reports on Socio-Informatics (IRSI), Proceedings of the PDC 2014, Infrastructuring, Collaboration and Evolving Socio-Material Practices of Changing Our World, 11 (2), 39‒48. McCrudden, C. and J. King (2016), ‘The Dark Side of Nudging: The Ethics, Political Economy, and Law of Libertarian Paternalism’, in A. Kemmerer, C. Möllers, M. Steinbeis, G. Wagner (eds), Choice Architecture in Democracies: Exploring the Legitimacy of Nudging (Recht im Kontext, vol. 6), Baden: Nomos, pp. 75‒139. Meyer, M.N. (2013), ‘Regulating the Production of Knowledge: Research Risk-Benefit Analysis and the Heterogeneity Problem’, Administrative Law Review, 65 (2), 237‒299. Miller, A.B., C. Wall, C.J. Baines, P. Sun, T. To and S.A. Narod (2014), ‘Twenty Five Years Follow-up for Breast Cancer Incidence and Mortality of the Canadian National Breast Screening Study: Randomized Screening Trial’, BMJ, 348, accessed 13 December 2017 at http://www.bmj.com/content/bmj/348/bmj.g366.full.pdf. Myers, E.R., P. Moorman, J.M. Gierisch, et al. (2015), ‘Benefits and Harms of Breast Cancer Screening. A Systematic Review’, JAMA, 314(15), 1615‒1634. Oeffinger, K.C., E.T.H. Fontham, R., R. Etzioni et al. (2015), ‘Breast Cancer Screening for Women at Average Risk. Guideline Update from the American Cancer Society’, JAMA, 314 (15), 1599‒614. Ostrom, E., C. Chang, M. Pennington and V. Tarko (2012), The Future of the Commons: Beyond Market Failure and Government Regulation, Institute of Economic Affairs, London: Profile Books. Perry, N., M. Broeders, C. de Wolf, S. Törnberg, R. Holland and L. von Karsa (eds) (2013), European Guidelines for Quality Assurance in Breast Cancer Screening and Diagnosis (4th edn, supplements), Luxembourg: Office for Official Publications of the European Union. Ploug, T., S. Holm and J. Brodersen (2012), ‘To Nudge or not to Nudge: Cancer Screening Programmes and The Limits of Libertarian Paternalism’, Journal of Epidemiology and Community Health, 66, 1193–1196. Pound, R. (1930), An Introduction to the Philosophy of Law, New Haven: Yale University Press. Purnell, J.Q., T. Thompson, M.W. Kreuter and T.D. McBrided (2015), ‘Behavioral Economics: “Nudging” Underserved Populations to be Screened for Cancer’, Preventing Chronic Disease, 12 (1), 140346. Rebonato, R. (2012), Taking Liberties – A Critical Examination of Libertarian Paternalism, New York: Palgrave Macmillan. Rizzo M.J. and D.G. Whitman (2009), ‘The Knowledge Problem of New Paternalism’, BYU Law Review Collections, 4, 905‒968. Rübig, P. (2012), ‘The Changing face of Risk Governance: Moving from Precaution to Smarter Regulation’, European Journal of Risk Regulation, 2, 145‒146. Schwartz, D. (2015), ‘Regulating for Rationality’, Stanford Law Review, 67, 1373‒1410. Sibony, A. and A. Alemanno (2015), ‘The Emergence of Behavioral Policy-Making: A European Perspective’, in A. Alemanno and A. Sibony (eds), Nudge and The Law. A European Perspective, Oxford and Portland, OR: Hart Publishing, pp. 10‒11. Simon, H. (1947), Administrative Behavior, New York: MacMillan. Simonsen, J.R. and T. Robertson (eds) (2012), Routledge International Handbook of Participatory Design, London: Routledge. Straßheim, H., A. Jung and R. Korinek (2015), ‘Reframing Expertise: The Rise of Behavioral Insights and Interventions in Public Policy’, in A.B. Antal, M. Hutter and D. Stark (eds), Moments of Valuation. Exploring Sites of Dissonance, Oxford: Oxford University Press, pp. 249‒268. Sukel, K. (2016), ‘The Scary Reality of Medical U-turns, and How to Stop Them’, New Scientist, accessed 10 October 2017 at https://www.newscientist.com/article/mg23130880-300-the-scary-reality-of-med​ ical-uturns-and-how-to-stop-them. Sunstein, C.R. (2010), Memorandum for the Heads of Executive Departments and Agencies, Disclosure and

M4716-STRAßHEIM_9781785367847_t.indd 161

25/04/2019 13:32

162  Handbook of behavioural change and public policy Simplification as Regulatory Tools, OIRA, accessed 29 September 2016 at http://www.whitehouse.gov/sites/ default/files/omb/assets/inforeg/disclosure_principles.pdf. Swiss Medical Board (2013), Dépistage Systématique par Mammographie, accessed 8 January 2017 at http:// www.medical-board.ch/index.php?id=809&L=1. Tallacchini M. (2015), ‘To Bind or Not Bind? European Ethics as Soft Law’, in S. Hilgartner, C. Miller and R. Hagendijk (eds), Science and Democracy. Making Knowledge and Making Power in the Biosciences and Beyond, Routledge: London-New York, pp. 156‒175. Teli, M., S. Bordin, M.M. Blanco, G. Orabona and A. De Angeli (2015), ‘Public Design of Digital Commons in Urban Places: A Case Study’, International Journal of Human–Computer Studies, 81, 17‒30. Thaler, R.H. and C.R. Sunstein (2003), ‘Libertarian Paternalism’, The American Economic Review, 93 (2), Papers and Proceedings of the One Hundred Fifteenth Annual Meeting of the American Economic Association, Washington. Thaler R.H. and C.R. Sunstein (2008), Nudge. Improving Decisions about Health, Wealth and Happiness, London: Penguin Books. Thaler, R.H., C.R. Sunstein and J.P. Balz (2010), ‘Choice Architecture’, accessed 3 October 2016 at http://ssrn. com/abstract=1583509. The Lancet (2012), ‘The Breast Cancer Screening Debate: Closing a Chapter?’, The Lancet, 380. Tor, A. (2008), ‘The Methodology of the Behavioral Analysis of Law’, Haifa Law Review, 4, accessed 29 September 2016 at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1266169. van Aaken, A. (2015), ‘Judge the Nudge: In Search of the Legal Limits of Paternalistic Nudging in the EU’, in A. Alemanno and A. Sibony (eds), Nudge and The Law. A European Perspective, Oxford and Portland, OR: Hart Publishing, pp. 83‒112. van Asselt, M.B.A., M. Everson and E. Vos (eds) (2014), Trade, Health and the Environment. The European Union Put to The Test, London and New York: Routledge. van Epps, E.M., K.G. Volpp and S.D. Halpern (2016), ‘A nudge toward participation: improving clinical trial enrolment with behavioral economics’, Science Translational Medicine, 8 (348), accessed 13 December 2017 at http://stm.sciencemag.org/content/8/348/348fs13.full. Von Schomberg, R. and S.R. Davies (2010), Understanding Public Debate on Nanotechnologies. Options for Framing Public Policy, Luxembourg: Publication Office of the European Union. Welch, G.H. (2012), Over-Diagnosed: Making People Sick in the Pursuit of Health, Boston: Beacon Press. White, M.D. (2013), The Manipulation of Choice: Ethics and Libertarian Paternalism, New York: Palgrave MacMillan. Winner L. (1980), ‘Do Artifacts Have Politics?’, Daedalus, 109 (1), 121‒136. Wynne, B. et al. (2007), Taking European Knowledge Society Seriously, Report of the Expert Group on Science and Governance, Luxembourg: Publication Office of the European Union.

M4716-STRAßHEIM_9781785367847_t.indd 162

25/04/2019 13:32

12.  Behaviouralizing Europe: behavioural economics enters EU policy-making P.W. (Peter-Wim) Zuidhof

1. INTRODUCTION Many expected that the global financial and economic crisis would lead to a paradigm change within economics and economic policy-making. Some argued that the economic crisis announced the end of ‘orthodox’ or ‘neoclassical’ economics and forebode the demise of the neoliberal economic policy consensus (e.g. Stiglitz 2008). These hoped-for changes, however, appear not to have materialized as expected. Some change, however, came from an unexpected quarter. If there is one school of economic thought that emerged stronger after the crisis, it would be behavioural economics. With its focus on cognitive bias and the limited rationality of economic action, behavioural economics was a promising candidate for offering more adequate policy guidance after the crisis. Drawing on a long history of experimental studies, behavioural economists identified a wide array of cognitive biases that challenge the rationality assumptions of standard economic theory and policy-making. These range from nonstandard preferences such as the default or endowment effects, nonstandard beliefs due to overconfidence and priming, to nonstandard decision-making based on heuristics such as framing or menu effects (e.g. DellaVigna 2009). It is now common in behavioural economics to analyse decisionmaking as stemming from the interaction of two distinct systems, the one fast, automatic and uncontrolled, and the other slow, reflective and controlled (Kahneman 2011). Or using terminology introduced by Thaler and Sunstein (2009), when taking decision people often act as Humans rather than Econs. The main contribution of behavioural economics has been to provide overriding evidence that in real life, Humans systematically fail to act like Econs and do not conform to the rational model of homo economicus. While the insights from behavioural economics initially featured as part of an internal debate within economics, they have slowly been translated into policy (e.g. Shafir 2013). Key in popularizing behavioural economics for policy-making has been Robert Thaler and Cass Sunstein’s best-selling book, Nudge: Improving Decisions about Health, Wealth, and Happiness (2009). The book makes a convincing case for a behaviourally informed policy paradigm dubbed ‘libertarian paternalism’. It casts policy-makers in the role of ‘choice architects’ whose task is to design decision environments that ‘nudge’ human actors to overcome their biases and act more in accordance with their rational interests. The policy rationale of Nudge is to use insights from behavioural economics to design choice environments that nudge Humans to become more like Econs, enabling them to better serve their own welfare. The idea of nudge with its seemingly smart, simple, intuitive and non-invasive type of policy-making, immediately appealed to policy-makers. It led President Obama in 2009 to appoint Cass Sunstein (2013) as his “regulatory czar” and head of the Office of Information and Regulatory Affairs. And most famously, in 2010, 163

M4716-STRAßHEIM_9781785367847_t.indd 163

25/04/2019 13:32

164  Handbook of behavioural change and public policy Prime Minister Cameron of the UK created the Behavioural Insights Team – popularly known as ‘the Nudge Unit’ – directed by David Halpern (2015). The promise of behavioural economics and Nudge was not lost on EU policy-makers. Since 2008, one can witness the piecemeal entry of insights and techniques from behavioural economics into EU policy. These first emerged at the erstwhile Directorate-General for Health and Consumers (SANCO) – now DG Justice and Consumers – especially in the areas of consumer affairs and health policy. Behavioural science has informed legislation on pre-ticked boxes in e-commerce websites, the packaging of tobacco products, and the regulation of standard information documents for investment products. It influenced policy on online gambling, energy efficiency labelling, and information on bank fees. With the adoption of the Better Regulation Agenda in 2015, a number of these behavioural principles became part of the so-called Better Regulation Toolbox. The steady import of insights and techniques from behavioural economics has farreaching consequences for EU policy-making. This chapter is concerned with answering the question of how the adoption of behavioural insights affects the political orientation of the European Union: what are some of the political implications of adopting behavioural insights? More specifically, the chapter seeks to examine whether behavioural policies represent a retreat from the alleged dominance of neoliberal policies and a return towards more traditional liberal regulatory policies, or do they indicate a move in a different direction? Throughout this chapter, the term ‘liberal’ is used in a specific sense derived from Foucault (2007, 2008) to refer to conventional conceptions of economic policy where government achieves political goals by regulating or limiting markets. The term neoliberal by contrast, is used to refer to a new class of policies that apply market principles and construct markets to regulate other markets or even non-markets. A telling example of such neoliberal market making is the EU Emissions Trading Scheme (EU ETS), where the EU constructed an emissions market as a means to conduct its environmental policy. The ‘neo’ in neoliberalism therefore signifies an inversion of the liberal logic, in which markets are not merely objects of regulation, but are elevated into a norm for regulation. The political consequence of this inversion is that markets have been turned into a political norm that displaces former liberal principles such as equality or fairness. The chapter thus seeks to determine whether the behavioural policies as adopted by the EU are bolstering neoliberal market making, or whether they perhaps constitute a return to ‘liberal’ market regulation. Are nudges that seek to make Humans more like Econs new ways of limiting and regulating markets or are they imposing neoliberal market norms? Or could it be that behavioural policy is taking us beyond this traditional divide between ‘liberal’ and neoliberal policy-making, into uncharted territory? To answer these questions, the chapter first charts how insights from behavioural economics have slowly filtered into EU policy; in what policy areas did behavioural economics emerge, what insights and techniques were imported, and how can we explain that they emerged where they did? It secondly maps the actors and organizations that were instrumental in disseminating behavioural economics within European policy-making and how theoretical insights were translated into EU policy. The third part finally seeks to determine whether the behavioural turn presents a break with reigning neoliberal policy views within EU policy-making, whether it constitutes a return to liberal policy views, or if it could perhaps be construed as the advent of a post-neoliberal politics. The chapter concludes by arguing that it does neither. While behavioural approaches to EU policy remain modest to date,

M4716-STRAßHEIM_9781785367847_t.indd 164

25/04/2019 13:32

Behaviouralizing Europe  165 they present a significant shift in the political orientation of the EU that potentially has important political implications.

2. BEHAVIOURAL POLICY-MAKING IN THE EUROPEAN UNION Over the past decade there has been a growing interest in behavioural economics among policy-makers in the EU. Among its most visible achievements to date are four flagship cases where behavioural insights inspired EU legislation in the form of directives, official recommendations, and a regulation, with still a number of cases in the making. The first of these has been the inclusion in 2011 of a prohibition on pre-ticked boxes on e-commerce websites in the Directive on Consumer Rights (2011/83/EC, art. 22; see also Sibony and Alemanno 2015). The proposal for this directive prepared by the erstwhile DG Health and Consumers (now DG Justice and Consumers) dates back to 2008 and is advertised as one of the first visible uses of behavioural insights in EU policy-making.1 It reflects the behavioural insight that actors tend to suffer from default bias and the prohibition nudges consumers to remain closer to their interests. A second, oft-cited case is the 2009 competition decision of the European Commission banning Microsoft’s default option for Internet Explorer by requiring it to insert a ‘choice screen’ allowing consumers to decide which browser to install.2 Behavioural insights did not inform the Commission’s decision on the case, but were invoked when designing market tests to determine whether the technical solutions offered by Microsoft were cleared of default biases. A third flagship case is the new provisions on plain packaging and visual display bans in the Revised Tobacco Products Directive (2014/40/EU, art. 9). According to Alemanno (2012b, 2012a) the directive is indicative of a behavioural turn in tobacco regulation. Rather than regulating consumers through financial incentives like taxes or by providing health information, it modifies the decision environment and nudges consumers to take better care of their health interests. The new policies seek to address behavioural issues such as information overload characteristic of earlier policies. This is also the first case in which behavioural research directly guided the drafting of regulation. DG Health and Consumers commissioned Sambrook Research International, a market research consultancy, with reviewing the scientific knowledge on health and tobacco labelling.3 Many of the consultancy report’s recommendations on what type of warnings could effectively affect smoker’s behaviour, directly found their way into the new tobacco directive. The fourth flagship case involves the first study based on behavioural economics conducted by the Commission. In 2010, the DG Health and Consumers commissioned Decision Technology Ltd. to study the decision-making process of consumers in the market for retail investment services with a view of designing policy solutions that could assist consumers make ‘better decisions’. The resulting report Consumer Decision-Making in Retail Investment Services: A Behavioural Economics Perspective (2009) studied how consumer beliefs regarding investment decisions may suffer from various ‘behavioural biases’ such as framing effects, intertemporal biases, short-sightedness, the importance of heuristics, and risk or ambiguity aversion. A number of online and laboratory experiments were designed to study how subjects respond to various policy interventions directed at these biases. On the basis of these experiments, the study recommends simplifying and

M4716-STRAßHEIM_9781785367847_t.indd 165

25/04/2019 13:32

166  Handbook of behavioural change and public policy standardizing investment product information for instance by using a standardized ‘Key Information Document’ for investment products. The Commission presented the results of the latter study at its second Behavioural Economics Conference organized in November 2009. DG Internal Market subsequently started consultations on standards of protection in retail financial services.4 It led to a decision in 2014 to introduce standardized information documents with the Regulation on Key Information Documents for packaged retail and insurance-based investment products (EU No 1286/2014). A subsequent behavioural study was conducted by London Economics in 2015 to subject the proposed formats of the Key Information Document to consumer testing before the regulation would go into effect in 2017.5 It made the regulation on key information documents the first piece of EU legislation with an explicit base in behavioural economics. More behaviourally informed legislation is already in the making. Policy proposals and research studies are conducted in the areas of online gambling and energy efficiency labelling, for instance.6 In the slipstream of these legislative initiatives, there is a rapidly expanding set of behavioural policy initiatives in various stages of the policy process. Table 12.1 presents a non-exhaustive list of some of these policy initiatives. This overview makes clear that behavioural science has firmly established itself within a period of less than ten years as a legitimate and valuable new approach in EU policymaking and inspired new types of legislation. It also highlights some important patterns about the entry of behavioural insights into EU policy. First is to note in what policy areas behavioural insights make their first appearance. Behavioural policy-making has nearly exclusively been concentrated in the areas of consumer policy, environmental policy, health and food safety, and to a lesser extent in the areas of taxation or education, youth, culture and sport. Institutionally, it emerged first in health and consumer policy which at the time resided under a single Directorate-General, DG Health and Consumers (SANCO). After its first embrace of behavioural economics, DG SANCO started a cooperation with the Commissions’ Joint Research Centre (JRC). Strikingly, the first entry of behavioural economics is outside of policy areas that are typically considered ‘economic’. While behavioural economics has made substantial contributions to economic policy in macroeconomics, labour markets and especially finance (viz. Camerer et al. 2004, pp. 31‒36), its first EU applications are in ‘non-economic’ domains such as consumer, health and environmental policy. Although in line with the general popularization of behavioural economics, it remains striking that despite the appeal of behavioural studies to challenge the basic assumptions of conventional economic policy, it primarily enters EU policy to tackle policy issues that are peripheral to traditional economic policy. Even more important in the EU context, is the political context in which behavioural insights first emerge. These are areas where the EU has limited competence. Unlike the EU’s exclusive competence with regard to monetary and competition policy for instance, consumer protection, the environment, and public health are areas of shared competence between the EU and Member States and it has even less competence with regard to human health. The reason why behavioural policies first surface in areas where the EU has less competence, is that they are apparently much better suited to areas where the options for direct regulation are politically limited. While the EU cannot directly tax smokers, it can, with a reference to the internal market, issue directives on the labelling and packaging of

M4716-STRAßHEIM_9781785367847_t.indd 166

25/04/2019 13:32

167

M4716-STRAßHEIM_9781785367847_t.indd 167

25/04/2019 13:32

Digital Agenda Bank Fees Common Sales Law

Area

Taxation

Sustainability Information Food Information Eating Habits

Behavioural Economics and Taxation

Impact of information on patients’ choice within the context of the Directive 2011/24/EU of the European Parliament and of the Council on the application of patients’ rights in cross-border healthcare Study on the effects on consumer behaviour of online sustainability information displays Study on the Impact of Food Information on Consumers’ Decision Making Milan BExpo 2015: A behavioural study on food choices and eating habits

Testing CO2/Car labelling options and consumer information Study on the impact of the energy label – and potential changes to it – on consumer understanding and on purchase decisions Influences on consumer behaviour: Policy implications beyond nudging

Consumer behaviour in a digital environment Bank Fees Behaviour Study Testing of a Standardized Information Notice for Consumers on the Common European Sales Law

Study

Sources:  EC websites: e.g. website of the EC Joint Research Centre.7

Taxation: 2014

2015

2014

2014

Patient rights

Consumer Behaviour

2014

Health Policy: 2014

Energy Labelling

2013

Environmental Policy: 2013 CO2 Labelling

Consumer Policy: 2011 2012 2013

Year

Table 12.1  Behavioural policy initiatives in the European Union

EC Joint Research Council

Ecorys

TNS

Ecorys

London Economics

Ecological Institute

London Economics

LSE and Partners

London Economics TNS Gallup

Research Institution

DG Taxation and Customs Union

EAHC, DG Health and Consumers CHAFEA, DG Health and Consumers CHAFEA, DG Health and Consumers

CHAFEA, DG Health and Consumers

DG Environment

DG Energy

DG Climate Action

EP, DG Internal Affairs DG Health and Consumers DG Justice and Consumers

Directorate-General

168  Handbook of behavioural change and public policy cigarettes that nudge citizens to smoke less. Behavioural policies thus appear particularly serviceable in EU policy areas where more overt policy-making is politically contentious. Second, there is a striking pattern as to whom or what is the object of the EU’s behavioural policies. As it is currently taking shape, EU behavioural policy focuses exclusively on governing the behaviour of individuals and especially in their capacity as consumers, as for example in the areas of finance, the environment or health. With the adoption of behavioural policies, the focus of regulation shifts away from regulating markets and market failure, towards targeting the behaviour of individuals within those markets. Moreover, instead of addressing the behaviour of other actors such as business or financial institutions, its policies appear to single out the consumer as the primary object of behavioural change. Behavioural policy-making in the EU thus facilitates a novel focus in regulation towards governing the behaviour of individual actors rather than markets, and a targeted focus on consumers as opposed to other actors. A third pattern that emerges from the first encounters between behavioural economics and EU policy-making, is the method by which behavioural research is mobilized and applied in the policy process. All behavioural policies cited passed through similar policymaking trajectories, which in all cases involved the commissioning of behavioural research studies to private research consultancies. The studies typically consist of a literature review and the application of randomized controlled trials using laboratory, online or insitu experiments. Their aim is to generate firm evidence on the behavioural effectiveness of various policy options. The research studies reviewed, however, employ surprisingly little concepts and theories developed within behavioural economics and instead adopt a more general behavioural approach. Rather than studying the biases that prevent people from acting in their own interest, their purpose seems more mundane: to determine what works. Given that most studies are conducted by market research agencies, the main purpose instead seems to validate policies by making sure they are market tested. This evidently fits in a trend towards evidence-based policy-making, but has the effect of shielding policies from effective contestation in the political process. One can safely conclude that behavioural economics has made successful inroads into EU policy-making and established itself as a legitimate input for EU policy. It did so within a distinct political context. Behavioural theories and policies were initially only imported into the areas of consumer, health and environmental policy, leaving core economic policy-making unaffected. In these areas, where the EU’s political legitimacy was considerably restricted, they proved fertile tools for moving political agendas forward. The way behavioural studies are conducted turn them into policy solutions that are market tested, which makes them less politically contentious. These seemingly innocuous behavioural techniques however introduce new and pervasive approaches into politically contentious policy areas pushing these into new and unexpected directions.

3.  BEHAVIOURALIZING EU POLICY-MAKING To better understand how behavioural economics managed to swiftly capture the political imagination in the EU, it is worth examining in more detail how behavioural ideas travelled from the world of academia to that of EU policy. As always with the nexus between science and policy, the history of how behavioural ideas entered EU policy-making is

M4716-STRAßHEIM_9781785367847_t.indd 168

25/04/2019 13:32

Behaviouralizing Europe  169 multifaceted and in this case three distinct transmission mechanisms stand out that are of broader significance for understanding how behavioural insights were translated into actual policies. The first mechanism explaining the transmission of behavioural economics to EU policy is a set of push factors. Behavioural economics arose from a long internal debate within economics and between economists and psychologists that was mostly concerned with defining their own theoretical and empirical questions and with establishing behavioural economics as a respected field of inquiry (Heukelom 2014). Only in the 1990s as part of a new self-conscious phase in the history of behavioural economics, one begun popularizing behavioural economics as a novel policy paradigm. The first articulations of such a policy outlook appeared in the 2000s, like Thaler’s article with the legal scholar Sunstein on ‘Libertarian Paternalism’ published in the American Economic Review (Thaler and Sunstein 2003) and Camerer and others’ ‘Regulation for Conservatives: Behavioral Economics and the Case for “Asymmetric Paternalism”’ (Camerer et al. 2003). It took provocative metaphors and slogans to mobilize the first translation of behavioural economics into policy. The true popularization of the field, however, took off with the publication of a range of well-crafted and highly appealing books during the 2000s and around the global financial crisis, with Nudge being at its centre, that familiarized large audiences with the insights of behavioural economics. While gaining popular appeal, behavioural economics also started taking a hold in professional economic circles. Besides leading theorists grappling with the policy implications of behavioural economics (viz. Shafir 2013), a variety of national and international economic policy organizations begin exploring policy applications. The OECD for instance incorporated behavioural economics into its 2010 toolkit on consumer policy which is used as input by policy-makers (Lissowska 2011).8 It organized a conference in 2014 on Behavioural Insights and New Approaches to Policy Design bringing together theorists like Thaler and Sunstein, practitioners such as David Halpern of the BIT, international policy specialists, and EU policy-makers like Xavier Troussard of the EC Joint Research Centre (Lunn 2014).9 The World Bank with its World Development Report: Mind, Society and Behavior (World Bank 2015), the IMF (e.g. Blanchard 2011) and a host of national and international organizations, performed a similar function. EU documents often refers to these various push factors to make its case for why behavioural policy could be relevant for EU policy. A second transmission mechanism is found in a diverse set of intermediaries comprising policy entrepreneurs and policy networks. Especially since the publication of Nudge, there has been a growing number of academics, policy organizations and think tanks, research consultants, and other small-scale initiatives that devote themselves to translating behavioural economics into policy, some aiming explicitly at EU policy. The Behavioral Science & Policy Organization funded by a number of research universities and the Sloan and MacArthur foundations, for example aims ‘to foster dialog between social scientists, policymakers and other practitioners in order to promote the thoughtful application of rigorous,  empirical behavioural science in ways that serve the public ­interest’.10 To facilitate policy exchange, the Joint Research Centre of the EC has produced a register of the various state offices, research divisions of central banks and other official economic policy research organizations that seek to apply and promote behavioural policymaking.11 Entrepreneurial academics promote behavioural theory by publishing working

M4716-STRAßHEIM_9781785367847_t.indd 169

25/04/2019 13:32

170  Handbook of behavioural change and public policy papers with titles such as A Practitioner’s Guide To Nudging (Ly et al. 2013, see also Ly and Soman 2013). In the European context, the legal scholar Alberto Alemanno has been at the forefront of studying behavioural policy (viz. Sibony and Alemanno 2015) while also acting as policy entrepreneur. Alemanno, for instance, has been organizing a conference titled ‘Nudging in Europe’ that brought together important scholars and EU practitioners, has been involved in research projects, conferences and workshops on the topic, regularly blogs about nudging in the EU, and consults for the EU.12 Commercial policy entrepreneurs have also taken to advocating behavioural policy in the EU. iNudgeyou for example, is a platform run by a research consultancy in Denmark that advertises itself as ‘The Danish Nudge Unit’ and operates a blog (‘want to learn how to nudge?’) and organizes courses.13 In 2014, it was one of the initiators of The European Nudging Network (TEN), a network of behavioural policy practitioners in the European Union, a collaboration with Alemanno’s HEC Paris and the OECD.14 The most important policy transmission mechanism is probably the applied behavioural research conducted by entrepreneurial academics and their equivalents at research consultancies. Consultancy services have been at the forefront of translating behavioural economics into applied research that can be used in the policy process. All behavioural policy initiatives in the EU relied on commissioned research by policy research consultancies such as TNS, London Economics, Ipsos, and Ecorys, often in cooperation with academic institutes or individual academics. As often, a most expedient way for ideas to travel is when they are being paid for. The third and most important transmission mechanism for bringing behavioural economics to the EU are what could be dubbed policy labs. Policy labs are more or less institutionalized environments where one experiments with the first application of behavioural insights to policy. When these first experiments turn out successful, behavioural policies gain legitimacy and can be extended to other policy environments. The best example of such a policy lab is the Behavioural Insights Team (BIT) established in 2010 by David Cameron’s Cabinet Office. The so-called ‘Nudge Unit’ was headed by David Halpern (2015). It has since 2013 been semi-privatized and opened branches in Sydney and New York. BIT was the first to put behaviourally informed policies to practice, such as using behavioural nudges to improve tax compliance. BIT has been an important catalyst in the transmission of behavioural economics to policy and is often cited as a source of inspiration for any new behavioural policy initiative, including those within the EU. The successes of BIT inspired policy-makers of DG SANCO first in 2008 and later together with the JRC to form a behavioural policy lab. With the creation in 2015/16 by the JRC of the Foresight and Behavioural Insights Unit (FBIU), headed by Xavier Troussard, the EC further institutionalized its behavioural policy lab.15 It is now a central branch within a recent initiative aptly called ‘EU Policy Lab’ which is meant as an ‘experimental space for innovative policy-making’.16 Policy labs such as BIT or the EC’s FBIU, perform three related functions in the transmission of behavioural knowledge. The first is policy testing. Early policy initiatives serve the purpose of gaining experience with behavioural policy-making and initial successes motivate subsequent initiatives. After DG MARKT and SANCO commissioned a pilot study on retail investment services they shared its results at their second EU conference on Behavioural Economics in 2010. Advertising the conference, it stated its aims as:

M4716-STRAßHEIM_9781785367847_t.indd 170

25/04/2019 13:32

Behaviouralizing Europe  171 This conference will be an opportunity to present the results of the joint MARKT SANCO pilot behavioural study on retail investment services. The approach used in this first behavioural study could form the basis of a framework contract to be used by all Commission services to carry out behavioural experiments to road test regulatory or policy remedies. (DG SANCO, Conference Description)17

This example shows that successful pilot studies help legitimize the use of behavioural techniques and inspire future policy initiatives. Policy lab experiments translate behavioural science into policy and prove that it can be done. Second, policy labs promote policy mainstreaming. Successful policy experiments set the standard for future studies. The EU’s first behavioural study led to the establishment of the Framework Contract for the Provision of Behavioural Studies, which set the standard and enabled others to run similar studies. The involvement of the Joint Research Centre in 2012, was an important next step in mainstreaming behavioural policy-making that helped standardization and sharing of methods across divisions. In 2013, the JRC produced a highly accessible report Applying Behavioural Sciences to EU Policy-making (van Bavel et al. 2013) aimed at stimulating new policy initiatives. It explains the core ideas of behavioural science and how to apply these to policy-making, while devoting much attention to the feasibility of behavioural research and how to best commission a behavioural study. When the framework contract needed renewal in 2014, the JRC organized a workshop for EU professionals and research consultancies on Good Behavioural Research for EU Policy-Making (Rodríguez-Priego and van Bavel 2014) with a view of further practical guidelines for conducting and commissioning behavioural research for policy. The JRC’s role and the framework contract were important for setting research standards and for mainstreaming and disseminating behavioural policy in the EU. A third feature of a behavioural policy lab is policy promotion. After testing and mainstreaming, dissemination is only a natural next step. The Policy Lab is an example of such policy promotion and dissemination. With an attractive title and website, the EU Policy Lab advertises itself as ‘a collaborative and experimental space’ that employs advanced policy technologies such as ‘Foresight’ and ‘Design Thinking’, besides Behavioural Insight. Its mission is to ‘serve all Commission’s Directorates and focus on the creation of bridges and synergies’ through ‘Lab sessions’ and ‘Reaching out’. The EU Policy Lab managed to have behavioural insights included in the Better Regulation Agenda of 2015 and have them appear in various chapters of the Better Regulation Toolbox.18 Another project conducted as part of the EU Policy Lab has been to collect behavioural policy initiatives not just within the Commission but also among Member States and to build a network of behavioural policy practitioners for sharing and disseminating behavioural insights and techniques (Lourenço et al. 2016). It is an example that the EU’s behavioural policy promotion has not remained confined to EU policy but extends to embrace all levels of government in Europe. With these two initiatives, DG SANCO and later the JCR fulfilled the role of policy lab through which they succeeded in making behavioural economics a legitimate and accepted tool in EU regulation. The EU policy lab thus proved to be an effective instrument for behaviouralizing EU policy if not entire Europe. A striking feature about the transmission of behavioural economics to EU policy is that behavioural policies mostly travel through the replication of behavioural approaches and techniques by research institutions, consultancies and in policy labs. Especially policy labs prove fertile breeding grounds for the propagation of behavioural policy.

M4716-STRAßHEIM_9781785367847_t.indd 171

25/04/2019 13:32

172  Handbook of behavioural change and public policy Because behavioural techniques typically move through expert policy channels, political or ideological considerations hardly figure in the promotion of behavioural policies. One will look in vain for references to ‘libertarian paternalism’, for instance, or arguments about the bounded rationality of individuals. Instead, technocratic arguments showing that behavioural policies work better, and especially the quiet material multiplication of behavioural techniques through research institutions, consultancies and in policy labs seems to have prevailed.

4. THE POLITICS OF EU BEHAVIOURAL POLICY: BEYOND LIBERAL AND NEOLIBERAL? Having reviewed its gradual entry, the remaining question is whether and how behavioural policy instruments transform the political orientation of EU policy-making. How does the adoption of behavioural policies change the political rationale of government action? Reflecting on the political orientation of Nudge, Thaler and Sunstein claim that behavioural policy’s libertarian paternalism is ‘neither left nor right’ but offers a ‘new path’ and a ‘bipartisan’ foundation for ‘better government’ (Thaler and Sunstein 2009, p. 14). This section is rather asking whether the EU’s behavioural policies could be considered a return to liberal (re-)regulation of markets, or rather as a neoliberal attempt at constructing markets and imposing market norms, or even, as moving beyond these categories introducing a post-neoliberal type of politics. The distinction between liberalism and neoliberalism applied here, is inspired by Foucault (2007, 2008, see also Zuidhof 2014, forthcoming) and serves to highlight that various recent policy ideas and practices visibly depart from conventional liberal notions of market regulation. As illustrated by the example of emissions trading, what sets this new class of neo-liberal policies apart is that rather than regulating quasi-natural markets, they consist of applying market measures and principles. As opposed to limiting markets, neoliberal policies impose various kinds of market-like mechanisms and quasi-markets and consequently instil market norms and marketized political ideas (viz. Brown 2015). The EU’s behavioural policies clearly occupy an ambiguous position on this spectrum of liberal and neoliberal policies, sometimes advancing liberal attempts at market regulation in the areas of health, environment and consumer rights, while at the same time imposing market norms and principles by nudging consumers to abide by their economic interests. In a recent review, Will Leggett (2014) draws out the political paradoxes and contradictions of behavioural policy (see also Tyers, Chapter 23 in this volume). On the one hand Leggett cites those who criticize behavioural policy from a broadly ‘libertarian’ point of view. The latter charge behavioural policy-making with being a form of technocratic, research-based governance largely determined by experts. From such a libertarian point of view, behavioural policies challenge free market principles and constitute a repackaging of ‘big government’ which with a penchant for technocracy, even develops slight authoritarian features. In this view, behavioural policy is too ‘liberal’. A second set of critiques that Leggett calls ‘statist’, charge behavioural policies for being too neoliberal. In their view, behavioural policies tend to be overly pro-market and anti-state. By nudging economic subjects to act rationally, behavioural policies in fact smoothen the functioning of markets and advocate a simpler or leaner state that takes a

M4716-STRAßHEIM_9781785367847_t.indd 172

25/04/2019 13:32

Behaviouralizing Europe  173 back seat regarding economic affairs. Behavioural policies, in their view, pose no challenge to market principles, as the objective of behavioural policies is in fact to make humans more suited to the market. This way, behavioural policies are yet another attempt at neoliberal market constructivism. A third category criticizes behavioural policies from a ‘critical-democratic’ perspective. It argues that behavioural governance with its exclusive focuses on the individual while not addressing the functioning of markets, generalizes every (democratic) subject into a homo economicus. As such, behavioural policies are repackaging the state into an even more general neoliberal state, which maintains the centrality of market principles in government. So, following Leggett’s analysis, are the EU’s behavioural policies too liberal, too neoliberal, or even anti-democratically neoliberal? Adding to the latter perspective, McMahon (2015, p. 4) uses a Foucauldian governmentality approach to explain how behavioural economics contributes to an intensification of neoliberalism by introducing techniques that ‘deepen neoliberal practices and rationalities’. It does so in at least three ways. First, behavioural policy intensifies neoliberalism by shifting the object of regulation from the market to the individual. Where traditional liberal and neoliberal policies govern through the institution of the market, behavioural policy intervenes directly at the level of actors by addressing their interests, beliefs, and cognition. Whereas the EU ETS constructed a market to reduce pollution, behavioural policies use green labelling to nudge consumers to choose for environmental protection. Second and related, McMahon argues that behavioural policies more pervasively format economic subjects as homo economicus. Economic subjects in liberalism and neoliberalism were firstly ‘partners of exchange’. In addressing cognitive bias in general, behavioural policies universalize homo economicus to all forms of action. And rather than steering an innate homo economicus, behavioural policies actively produce Econs by tackling their cognitive biases. So McMahon concludes: ‘If the market cannot fix the individual, perhaps [behavioural] governmentality can fix individuals for the market’ (McMahon 2015, p. 13). Third, behavioural policies intensify neoliberalism by reinforcing the market as ‘depoliticized site of truth and veridiction’ (McMahon 2015, p. 3). The latter is a reference to Foucault (2008, p. 32) who argued that in neoliberal market-based policies, markets function as objective norm for neoliberal government. Behavioural policies similarly take rational market behaviour as their norm, when trying to nudge people away from cognitive bias to reach their inner economist. Using the market and unbiased rational behaviour as objective norm, rather than exterior political norms in effect depoliticizes policy. Policy-making is no longer a political activity but an empirical question. How does the EU’s application of behavioural policy fit in these considerations? As said, it does not fit easily in a simple spectrum of liberal and neoliberal policies. On the face of it, the behavioural interventions of the EU have a clear liberal streak, reminiscent of a state that is regulating markets in an attempt to do good. There are clear paternalistic overtones in aiming to reduce smoking, prevent gambling, or in caring for the environment. Yet to side with Leggett’s libertarians however, the EU interventions with their use of highly specialized behavioural studies take an increasingly technocratic form that quietly reshapes the EU into an intrusive, behavioural state. So, despite good old liberal intentions, the execution of behavioural policies conforms less to the image of a liberal state. On the other hand, behavioural policies also do not immediately conform to a

M4716-STRAßHEIM_9781785367847_t.indd 173

25/04/2019 13:32

174  Handbook of behavioural change and public policy standard neoliberal reading. They hardly show signs of a retreat of the state; at most the replacement of the regulatory state with a behavioural state. Behavioural policies may not lead to less state, but they do promise a leaner and simpler government, as exemplified by their inclusion in the Better Regulation Toolkit. More damning for a neoliberal reading, however, is that the EU’s behavioural policies do not appear to rely on market constructivist policies that govern by creating markets. None of the EU’s examples cite market principles, but do not actively challenge them either. There is a clear neoliberal sense however to how the legitimacy and effectivity of behavioural policies is determined through behavioural ‘market testing’, that elevates the market into a norm for policy. Behavioural policy measures thus appear neoliberal, but not in the way one was used to. Neither really liberal, nor typically neoliberal, the behavioural policies in the EU hence show most signs of a more advanced or intensified mode of neoliberal governance as outlined by Leggett and McMahon. For one, the behavioural policies of the EU shift the focus from regulating markets and regulating by means of markets to that of directly governing the interests, beliefs, and decisions of its citizens. Policies on energy labels, cigarette packaging, key information documents, and pre-ticked boxes no longer regulate markets, but only actors on these markets. Thus intervening deeper into the market tissue of society, it could be considered a first intensification of neoliberal governance. Second, these interventions moreover aim at making humans more like Econs or homo economicus and thus make subjects better suited for the market. Policies on bank fees, key information documents for investment products, energy labelling, patient choice in cross border health care, and a prohibition on pre-ticked boxes not just help citizens better care for their interests, but especially help them to become better market citizens. Intervening ever deeper into economic subjects and making them more suited to the market, behavioural policy is further entrenching the neoliberal market as a core policy principle. Another way in which behavioural policy leads to an intensification of neoliberalism is McMahon’s argument about how it depoliticizes policy-making and market regulation. Behavioural policies as they are adopted in the EU, progressively turn policy-making into a technical, empirical, and uncontested – that is, depoliticized – type of activity. The aim to help consumers overcome biases seems uncontroversial. Behavioural studies about how people respond to energy labels or investment documents, transforms policy-making into an empirical issue. With its good-hearted paternalism and behavioural biases essentially an empirical and technical matter, behavioural policies move beyond political contestation. The depoliticizing aspect of behavioural policy-making ties in with an important, specifically European dimension behind its rise in the EU. It is perhaps not surprising to see behavioural policy surface where it did. As an enhanced form of neoliberal and depoliticized policy-making, behavioural policy likely emerges in areas where the political competence of the EU is limited. When direct regulation is less of an option, the depoliticizing quality of behavioural policy may be a welcome way to circumvent political obstacles. Just as was the case of carbon reduction where direct regulation by means of taxes or quota were politically not feasible, one took the depoliticized neoliberal route of governing by means of establishing a market (Braun 2009). The same may apply to behavioural policy-making. When environmental or health policy cannot be conducted through direct political action, the advanced neoliberal means of depoliticized behavioural policy-making provide a convenient alternative.

M4716-STRAßHEIM_9781785367847_t.indd 174

25/04/2019 13:32

Behaviouralizing Europe  175

5. CONCLUSION: WHAT FUTURE FOR EU BEHAVIOURAL POLICY-MAKING? This chapter has shown that the EU’s behavioural policies occupy a contradictory position on the liberal–neoliberal continuum. On the one hand, the EU’s behavioural policies in the areas of health, consumer and the environmental affairs appear to be limiting markets and thus project a renewed form of liberal market regulation. At the same time, by nudging consumers and citizens to overcome cognitive bias and act more like Econs, the EU’s behavioural policies contribute to a further neoliberal extension of the reach of market norms while economizing new areas of policy-making. The way in which the EU’s behavioural policies do so, however. takes a form that departs from earlier forms of neoliberal market constructivism. Behavioural policies no longer rely on market constructivism, as for instance emissions markets did, to economize the behaviour of its subjects. Where ‘traditional’ neoliberal policies relied on markets and market-like mechanisms to economize behaviour and outcomes, behavioural policies dispense of markets altogether as economizing policy devices. The EU’s behavioural policies thus amount to a more advanced and pervasive form of neoliberal policies, that reaches straight for the economic subject. This way it could be considered the vanguard of a ‘neoliberalism after markets’ (Boeckler and Berndt 2012, p. 424). It was also observed that behavioural policy-making has depoliticizing effects that are particularly suitable to the European context. First of all, it provides a general lesson about how (advanced) neoliberal policies travel. Just like many neoliberal policies that preceded it, behavioural policy hardly travelled as a conscious political programme, but rather through the replication of behavioural techniques and practices in research labs for instance. This way of travelling is particularly effective in complex political environments such as the EU’s, where policy moves faster than politics because political agreements are harder to achieve. Second, the paternalism of behavioural policies, proves moreover especially suitable in those political contexts where the EU only has limited competence. The depoliticized methods of behavioural policy have allowed policy-makers to pursue political agendas in health care or environmental policy for instance where one would otherwise encounter considerable political obstacles. The smart and seemingly a-political behavioural policies promise a route of least political resistance. While there are no immediate reasons to be apprehensive about the ostensibly benign paternalism of the EU’s behavioural policies, its proliferation nonetheless requires caution. One concern is that because of the technocratic nature of behavioural policy-making, the advances of its paternalistic objectives and depoliticized neoliberalism remain largely hidden from public view. Moreover, by advancing market reasoning even deeper into the tissue of society, behavioural policies may be the next step in further de-legitimizing important liberal political values, such as justice, equality, and ultimately even freedom. If behavioural policies are indeed promoting liberal objectives but do so with advanced neoliberal means, what will guarantee that these objectives will always remain neatly liberal. It poses a risk that a depoliticized behavioural state with depoliticizing market policies, may in fact not strengthen but rather eclipse the liberal democratic state. The current fragile democratic state of the European Union may therefore not automatically stands to gain from a further depoliticization of its policy-making. Thaler and Sunstein’s (2009, p. 10) point that no context is nudge-free, may serve as

M4716-STRAßHEIM_9781785367847_t.indd 175

25/04/2019 13:32

176  Handbook of behavioural change and public policy a reminder that one better nudge for the right reasons. Leggett (2014) has proposed to democratize behavioural policy-making by making it serve liberal purposes again. Behavioural policy may only have a promising future if it manages to provide for a democratic, post-liberal and post-neoliberal future and nudges Humans not to become mere Econs, but flourishing democratic citizens. Current developments in behavioural policy-making in the EU unfortunately, do not yet show much signs of such promise.

NOTES  1. Proposal for a Directive of the European Parliament and of the Council on Consumer Rights, COM/2008/0196, art. 31.   2. Case COMP/C-3/39.530 – Microsoft (tying).   3. A Review of the Science Base to Support the Development of Health Warnings for Tobacco Packages (2009).   4. Working Document of the Commission Services (DG Internal Market), Consultation by Commission Services on legislative steps for the Packaged Retail Investment Products initiative (2010). See: http:// ec.europa.eu/internal_market/consultations/2010/prips_en.htm (last accessed: 6 April 2016).   5. London Economics and Ipsos, Consumer testing study of the possible new format and content for retail disclosures of packaged retail and insurance based investment products (MARKT/2014/060/G): http:// ec.europa.eu/finance/finservices-retail/docs/investment_products/2015-consumer-testing-study_en.pdf (last accessed: 6 April 2016).  6. European Commission, 2014, Commission Recommendation on online gambling (2014/478/EU); European Commission, 2015, Report from the Commission to the European Parliament and the Council, Review of Directive 2010/30/EU of the European Parliament and of the Council of 19 May 2010 on the indication of labelling and standard product information of the consumption of energy and other resources by energy-related products. COM(2015) 345.   7. A complete list of studies is maintained at the website of the Joint Research Centre of the European Commission: https://ec.europa.eu/jrc/en/research/crosscutting-activities/behavioural-insights/publica​tions (last accessed: 5 April 2016).   8. See also: OECD, Behavioural Economics, published at: http://www.oecd.org/gov/regulatory-policy/behav​ ioural-economics.htm (last accessed: 10 April 2016).   9. OECD website: www.oecd.org/naec/NAEC_Behavioural-Insights-Programme_23-Jan.pdf (last accessed: 10 April 2016). The conference resulted in a conference report see OECD (2015). 10. Website Behavioral Science and Policy Association. See: https://behavioralpolicy.org/ (last accessed: 10 April 2016). It operates from Durham NC. 11. EC Joint Research Centre, Behavioural Insights Applied to Policy. European Report 2016. See: http:// publications.jrc.ec.europa.eu/repository/bitstream/JRC100146/kjna27726enn_new.pdf (last accessed: 10 April 2016). 12. See for instance: Conference website Nudging Europe: What Can EU Law Learn from Behavioural Sciences. http://local.droit.ulg.ac.be/sa/nudging_europe/ (last accessed: 11 April 2016). Alberto Alemanno. Personal blog. http://www.albertoalemanno.eu/ (last accessed: 10 April 2016). Alberto Alemanno (2012). Nudging Europe: Why the European Commission should include behavioural insights in the design of regulatory proposals. http://www.politico.eu/article/nudging-europe/ (last accessed: 10 April 2016). Alberto Alemanno. The EU needs to embrace behavioural insights in the design of its policies 2012. http:// blogs.lse.ac.uk/europpblog/2012/06/26/eu-behavioural-insights-policies/ (last accessed: 10 April 2016). 13. iNudgeyou, the Applied Behavioural Science Group. See: http://inudgeyou.com/homepage/nudge-den​ mark (last accessed: 10 April 2016). 14. TEN – The European Nudging Network. http://tenudge.eu/ (last accessed: 21 November 2016). 15. EC Joint Research Council. Foresight and Behavioural Insights Unit. https://ec.europa.eu/jrc/en/research/ crosscutting-activities/behavioural-insights (last accessed: 11 April 2016). 16. EU Policy Lab. http://blogs.ec.europa.eu/eupolicylab/ (last accessed: 11 April 2016). 17. DG Health and Consumer Protection. Conference ‘Behavioural Economics, so What: Should PolicyMakers Care?’ 22 November 2010. http://ec.europa.eu/consumers/archive/conferences/behavioural_eco​ nomics2/index_en.htm (last accessed: 11 April 2016). 18. See tools numbered 11, 14, 15, 23, 28, 35, and 52 in: European Commission. Better Regulation Toolbox. http://ec.europa.eu/smart-regulation/guidelines/toc_tool_en.htm (last accessed: 11 April 2016).

M4716-STRAßHEIM_9781785367847_t.indd 176

25/04/2019 13:32

Behaviouralizing Europe  177

REFERENCES Alemanno, A. (2012a), ‘Nudging Smokers – The Behavioural Turn of Tobacco Risk Regulation’, European Journal of Risk Regulation, 3 (1), 32‒42. Alemanno, A. (2012b), ‘Out of Sight, Out of Mind – Towards a New European Tabacco Products Directive’, Columbia Journal of European Law, 18, 197‒242. Blanchard, O. (2011), ‘The Future of Macroeconomic Policy: Nine Tentative Conclusions’, IMF Blog: Insights and Analysis on Economics and Finance. https://blog-imfdirect.imf.org/tag/behavioral-economics/ (last accessed 10 April 2016). Boeckler, M. and Berndt, C. (2012), ‘Geographies of Circulation and Exchange III: The Great Crisis and Marketization “After Markets”’, Progress in Human Geography, 37, 424‒432. Braun, M. (2009), ‘The Evolution of Emissions Trading in the European Union: The Role of Policy Networks, Knowledge and Policy Entrepreneurs’, Accounting, Organizations and Society, 34 (3), 469‒487. Brown, W. (2015), Undoing the Demos: Neoliberalism’s Stealth Revolution. New York: Zone Books. Camerer, C., Loewenstein, G. and Rabin, M. (2004), Advances in Behavioral Economics, the Roundtable Series in Behavioral Economics. New York; Princeton, NJ: Russell Sage Foundation; Princeton University Press. Camerer, C., Issacharoff, S., Loewenstein, G., O’Donoghue, T. and Rabin, M. (2003), ‘Regulation for Conservatives: Behavioral Economics and the Case for “Asymmetric Paternalism”’, University of Pennsylvania Law Review, 151 (3), 1211‒1254. DellaVigna, S. (2009), ‘Psychology and Economics: Evidence from the Field’, Journal of Economic Literature, 47 (2), 315‒372. Foucault, M. (2007), Security, Territory, Population: Lectures at the Collège de France, 1977–78. Translated by: Burchell, G. New York: Palgrave Macmillan. Foucault, M. (2008), The Birth of Biopolitics: Lectures at the Collège de France, 1978–79. Translated by: Burchell, G. New York: Palgrave Macmillan. Halpern, D.S. (2015), Inside the Nudge Unit: How Small Changes Can Make a Big Difference. London: W.H. Allen. Heukelom, F. (2014), Behavioral Economics: A History. Cambridge: Cambridge University Press. Kahneman, D. (2011), Thinking, Fast and Slow, 1st edn. New York: Farrar, Straus and Giroux. Leggett, W. (2014), ‘The Politics of Behaviour Change: Nudge, Neoliberalism and the State’, Policy and Politics, 42 (1), 3‒19. Lissowska, M. (2011), ‘Overview of Behavioural Economics Elements in the OECD Consumer Policy Toolkit’, Journal of Consumer Policy, 34, 393‒398. Lourenço, J.S., Ciriolo, E., Rafael Almeida, S. and Troussard, X. (2016), Behavioural Insights Applied to Policy. European Report 2016, EC Joint Research Center. Lunn, P. (2014), Regulatory Policy and Behavioural Economics. Paris: OECD Publishing. Ly, K. and Soman, D. (2013), Nudging Around The World. Toronto: Rotman School of Management, University of Toronto. Ly, K., Zhao, M. and Soman, D. (2013), A Practitioner’s Guide to Nudging. Toronto: Rotman School of Management, University of Toronto. McMahon, J. (2015), ‘Behavioral Economics as Neoliberalism: Producing and Governing Homo Economicus’, Contemporary Political Theory, 14, 137‒158. OECD (2015), Behavioural Insights and New Approaches to Policy Design: The Views from the Field. Paris: OECD Publishing. Rodríguez-Priego, N. and van Bavel, R. (2014), Good Behavioural Research for EU Policy-Making: Workshop Report, Sevilla: EC Joint Research Council. Shafir, E. (ed.) (2013), The Behavioral Foundations of Public Policy. Princeton: Princeton University Press. Sibony, A.-L. and Alemanno, A. (2015), ‘The Emergence of Behavioural Policy-Making: A European Perspective’, in Sibony, A.-L. and Alemanno, A. (eds), Nudge and the Law: A European Perspective, Oxford: Hart Publishing, pp. 1‒25. Stiglitz, J.E. (2008) ‘The end of neo-liberalism? ‘, Project Syndicate (7 July 2008). http://www.project-syndicate. org/commentary/stiglitz101/English (last accessed 16 September 2011). Sunstein, C.R. (2013), Simpler: The Future of Government. First Simon & Schuster hardcover edition. New York: Simon & Schuster. Thaler, R.H. and Sunstein, C.R. (2003), ‘Libertarian Paternalism’, The American Economic Review, 93 (2), 175‒179. Thaler, R.H. and Sunstein, C.R. (2009), Nudge: Improving Decisions about Health, Wealth, and Happiness. Revised and expanded edition. New York: Penguin Books. van Bavel, R., Herrmann, B., Esposito, G. and Proestakis, A. (2013), Applying Behavioural Sciences to EU Policy-making, European Commission, Joint Research CenterEUR 26033 EN).

M4716-STRAßHEIM_9781785367847_t.indd 177

25/04/2019 13:32

178  Handbook of behavioural change and public policy World Bank (2015), World Development Report 2015: Mind, Society, and Behavior. Washington, DC: World Bank. Zuidhof, P.-W. (2014), ‘Thinking Like an Economist: The Neoliberal Politics of the Economics Textbook’, Review of Social Economy, 72 (2), 157‒185. Zuidhof, P.-W. (forthcoming), Imagining Markets: The Performative Politics of Neoliberalism. New York: Zone Books.

M4716-STRAßHEIM_9781785367847_t.indd 178

25/04/2019 13:32

PART III BEHAVIOURAL INSIGHTS AND INSTRUMENTS IN POLICY-MAKING

M4716-STRAßHEIM_9781785367847_t.indd 179

25/04/2019 13:32

13.  The enzymatic effect of behavioural sciences: what about policy-makers’ expectations? Kathrin Loer

1. NUDGING, BEHAVIOURAL CHANGE AND INSTRUMENT TYPOLOGIES ‘Nudging’ is often classified as a new policy instrument and there is a lively debate in science and in the media on it. This debate sometimes presents ‘Nudging’ as an instrument that holds promise with regard to specific, often complex political aims. Better compliance results are expected in fields like health care, consumer policy, environmental policies and in areas that depend on people responding to a policy instrument in a specific way in order to make the instrument effective. Two relevant aspects, however, are easily overlooked: Conceiving ‘nudging’ as a ‘new instrument’ is misleading as well as the assumption that policy-actors would prefer to replace previously known instruments by it. It is also misleading to narrow and misunderstand the view on ‘nudging’ to a form of ‘manipulation’, which quickly provokes defence reactions. To do the term ‘nudging’ justice, it is necessary to keep in mind that the term is merely a headline or an umbrella term and that its potential lies in unfolding and utilizing the variety of behavioural insights. Behavioural insights form the basis of different tools that in turn give existing policy instruments new spins, on the one hand, and, on the other, allow for combining and mixing different policy instruments. In order to better understand the mechanisms that characterize policy instruments in their traditional form, in this chapter I argue that we need to revise existing policy instruments and their typologies. This is the starting point for understanding the mechanisms that take effect when behavioural insights are incorporated into public policy-making. For policy-makers, incorporating behavioural insights into policy instruments is motivated by the perspective of being able to achieve better results (‘effective policy-making’1). But independent from empirical evidence that could answer the question of whether behavioural insights actually lead to effective policy-making, political scientists should offer a concept of how to deal with behavioural insights in connection to policy instruments. Here I put forward arguments that show that behavioural insights in public policy impact existing instruments in a way that changes our understanding of policy instruments and that have an effect on addressees. Despite the fact that including knowledge drawn from behavioural sciences does not lead to the introduction of additional instruments or to deriving specific principles thereof, they do have the potential to change the mechanisms at work in existing instruments or instrument mixes and in so doing they provoke a spin on the existing instruments. Due to insights won through behavioural sciences, their design changes and the spin resulting thereof can initiate remarkable changes. This process resembles enzyme reactions in living organisms, which we know from biology: enzymes speed up chemical reactions. An enzyme converts a molecule and provokes a metabolic 180

M4716-STRAßHEIM_9781785367847_t.indd 180

25/04/2019 13:32

The enzymatic effect of behavioural sciences  181 process in a cell, which has future implications. Hence, I call the spin that behavioural insights provoke in policy instruments the ‘enzymatic effect of behavioural sciences’. In developing this new typology and its ‘enzymatic effect’, I will integrate the mechanisms2 that behavioural insights trigger and will closely focus on the addressees, who so far have not been integrated into existing instrument typologies. This is not only theoretically true for instrument typologies in political science but seems to be the case also if we apply it to empirical cases. Although there are always underlying assumptions about the addressee, and policy-makers anticipate addressee will react to the chosen instruments or instrument mixes in a certain way, these aspects have not been included in the instrument debate. When looking at the use of ‘nudging’ and other tools inspired by behavioural sciences one might ask: Isn’t it boring just to talk about policy instruments? Isn’t it a highly technical topic with a large distance to aspects of daily life? No, it is not. A good example to demonstrate how relevant instrument use is in policy-making can be found in the vaccination topic. It is the state’s duty to protect its citizens and people wish to be protected from communicable diseases. Wouldn’t it be best to ‘nudge’ enough people to be vaccinated in order to ensure herd immunity? Progress in medical research has led to a variety of vaccines that prevent whole countries and regions from suffering from epidemic mushrooming. But this requires most people to be vaccinated. Figures on vaccination and herd immunity show that there are (sometimes larger) groups of people that do not want to be vaccinated. Effective public health policy is challenged by the dilemma between everybody’s right to insist on bodily integrity and not being vaccinated on the one hand and the state’s duty to protect its citizen on the other. In terms of herd immunity the state has an interest in reaching as comprehensive a protection against infections as possible (Thießen 2013). The state has an arsenal of instruments at its disposal in order to vaccinate enough people to reach herd immunity. These instruments can be separated into groups that either build on behavioural insights or that do not. To briefly sketch the idea of policy I would like to focus on the latter ones: First, the state could introduce vaccination as a duty and people would be forced to be vaccinated, which would also make an effective observation and sanctioning system necessary (authoritarian instrument: command and control). Second, the state could use incentives to move people to be vaccinated by, for example, paying people to be vaccinated or they could be attracted by other benefits and advantages (incentivizing instruments). Third, information campaigns could try to convince people by educating them on vaccinations, by establishing vaccinations as a routine of medical practitioners or other representatives of the health sector (informational instruments/capacity). And finally, the state could establish and foster specific organizations or promote cooperation, as an example one could think of a (state-led or regulated) infrastructure that would offer vaccinations, there could also be an arrangement between different groups or organizations that agrees on specific vaccination routines, for example, on the basis of voluntary agreements (organizational instruments). Despite some semantic differences, the typological terms used by scholars during the last thirty years are surprisingly similar3 and in the following I will continue to use the typology terms ‘authority’, ‘incentive’, ‘capacity’ and ‘organization’ (for a systematic overview see Ewert and Loer 2019). Everybody who reads the above list most probably will spontaneously either feel in favour of these instruments or not. Several objections are likely to come to mind when

M4716-STRAßHEIM_9781785367847_t.indd 181

25/04/2019 13:32

182  Handbook of behavioural change and public policy thinking about the political and social requirements that are linked to each instrument. What to many, however, is not clear is that our daily lives are more or less continuously affected by such policy instruments – might they be effective or not. The vaccination example shows that all these instruments have obvious advantages and disadvantages. To name just a few: First, a mandatory approach contradicts the principle of an individual’s bodily integrity. Second, if the state introduces incentives, it could be very costly. Third, information campaigns presume the addressee to be open minded, to be willing to learn, to have the ability to understand and to share the normative assumptions of the respective information. Fourth, specific organizational arrangements, cooperation and infrastructure could be costly but most notably depend on a variety of external factors that cannot be influenced easily. All of these instrument types presume that the addressee will (re)act in a certain way and they all have in common that the addressee, in one way or another, acts rationally. Researchers familiar with behavioural sciences will miss their findings when reading this listing of instruments as none of the mentioned instruments considers the behavioural dimension. Later in this chapter we will see how behavioural insights could play a role in policy-making. But, when it comes to policy instruments, it is not only political scientists that have been disregarding the behavioural dimension; political actors have, too. Although the concept of bounded rationality (Simon 1955) has generally found its way into political and social sciences decades ago, traditional instruments seemed to have been developed and used (if at all) mostly without systematically conceptualizing addressee’s behaviour for a long time (for fruitful historical examples of behaviourally inspired policies see Graf 2015). So far the addressee concept has not been introduced into the debate on instruments systematically in political science and little is known on how behavioural sciences can contribute to shaping policy instruments. In the following I will propose an idea on how to close this gap and connect the debate on policy instruments to the discussion on bringing behavioural insights into public policy-making. 1.1  This Chapter’s Focus As traditional policy instruments seem to be reaching their limits with regard to effectiveness (Hood and Margetts 2007, p. 4), it is no wonder that the question arises whether behaviourally-inspired strategies might be an elegant tool for public actors to make policies more effective. But how could these strategies be conceptualized and how can turning to instrument typology contribute to that goal? In this chapter I will put forward suggestions on how the existing instrument typologies can be advanced to better understand the mechanisms underlying the use of policy instruments in general and to integrate the mechanisms triggered by behavioural insights. I will argue that the available instrument typologies do not cover how behavioural insights impact traditional instruments, despite the fact that behavioural insights meanwhile actually play a role in policy-making. One reason why the debate on behavioural insights might not find its way to political science’s debate is that it does not meet the criteria of objectivity due the lack of precisely conceptualizing behavioural insights and their impact on policy-making. The same reason could apply to the media not taking up the concept as ‘nudging’ is understood as a stimulus word (e.g. newspaper reports on ‘nudging’) and behavioural approaches are presented as if they were one new instrument. An instrument type that manipulates

M4716-STRAßHEIM_9781785367847_t.indd 182

25/04/2019 13:32

The enzymatic effect of behavioural sciences  183 people and substitutes all well-known, tried and tested policy instruments in democratic societies. Since this interpretation does not do ‘nudging’ justice and the argumentation therefore is misleading, a structured approach and explanation on how to understand behavioural insights with regard to policy-making instruments is necessary. The instrument typologies as we know them typically leave the addressee out, they do not include any underlying assumptions about the addressees and the reasons why policy instruments fail to move addressees to behave according to the intended ends of the instrument in use are not searched for. The chapter is organized as follows: section 2 gives an overview of the debate on ‘nudging’ and behavioural insights and outlines how these concepts can be understood in the context of policy instruments. When choosing between instruments, it is essential to choose the addressee’s behaviour as a starting point, and section 3 introduces ways to conceptualize that starting point. This contribution closes not only with summarizing the findings but also with pointing to the requirements that have to be kept in mind if the enzymatic effect of behavioural sciences is to be triggered in public policy-making.

2.  ‘NUDGING’ IN POLITICS AND IN POLITICAL SCIENCES There are instances when public policy-making has an immediate effect on people’s daily lives: When mayor Bloomberg announced the ‘Soda Ban’ (Sugary Drinks Portion Cap Rule) for New York it meant that all people who were used to drinking large sodas had to change their habits. Although it was still possible to drink a large serving, the soda ban made it much more difficult. Some comments branded this ban as a sure sign of a ‘Nanny State’. Although the New York State Court of Appeals rejected the rule and the ‘Soda Ban’ no longer was effective, a debate over the consumption of sugary drinks and food had been set off. No matter whether the aim to restrict the consumption of sweet drinks and food was prudent, we could question the kind of instrument used here. Bloomberg did not want to prohibit drinking sodas but he introduced a quite strong regulation and intervened not only on the serving size, which is normally left to the vendor, but also in the alternatives a customer can choose between. In order to analyse what instrument was used in this case it is important to understand the mechanism underlying the instrument, the relationships between state and citizen as well as the relationships between the producers of goods and services and the instrument addressees. Understanding this part of the character of the instrument lays the ground for further debates about the goodness or fit of the instruments in use, as well as its implications on people’s behaviour and in so doing on the instrument’s effectiveness. On the following pages, I will analyse the character of policy instruments and their use. I will not discuss the political targets as such (e.g. energy use, fighting obesity) but rather focus on the rationales of policy-making. Policy-makers try to find solutions to (societal, economic, environmental …) problems and/or they (re-)act in a political context to demonstrate and stabilize power and strength. The question arises which types of instrument or instrument mix are available from which policy-makers expect to have an effect, be it an effect on problem solving or be it an effect in demonstrating power and strength. Howlett and Ramesh (Howlett and Ramesh 1993) refer to Doern and Wilson (Doern and Wilson 1974) who pointed out that governments would, for ideological reasons, ‘prefer to use the

M4716-STRAßHEIM_9781785367847_t.indd 183

25/04/2019 13:32

184  Handbook of behavioural change and public policy least coercive instruments available and would only “move up the scale” of coercion as far as necessary in order to overcome societal resistance to attaining their goal’ (Howlett and Ramesh 1993, p. 7). We would assume that in liberal democracies and liberal societies both citizens and politicians would prefer a least coercive way of being steered or ruled.4 The success of steering or ruling, the question of whether an instrument is effective or not directly depends on the addressee’s compliance (Weaver 2015, and Chapter 16 in this volume) and it is necessary to be effective in policy-making to secure political support and to be successful in (re-)elections. Therefore, the context between instrument use, its political acceptance and its effectiveness is highly relevant. But despite the fact of it being highly relevant, finding the right instrument type is not an easy task for political actors. Four major difficulties challenge the ability to find the right instrument. First, there is a variety of instruments and tools with different degrees of coerciveness. How coercive an instrument should or must be depends on the appropriateness in the particular context. Second, highly complex issues imply that sometimes there is a need for specific resources and expertise. High complexity also means that comprehensive solutions have to be found in order to make a policy effective. Third, the use of existing instruments can determine the future use of a specific policy as the institutional effect of instruments can provoke path dependency and narrow the view of policy-makers when choosing instruments. Last but not least, the ‘black-box’ addressee makes it difficult to find appropriate instruments or a mix of instruments that match the addressee’s behaviour. When it comes to policy instruments or instrument mixes, policymakers more often than not are familiar with the ineffectiveness of policy instruments or instrument mixes. In searching for new instruments or strategies to enhance the effectiveness of existing instruments, there are some leading political actors that believe they might have found a strategy to enhance the effectiveness of policy instruments in ‘nudging’ and drawing on behavioural insights. The discussion on ‘nudging’ was provoked by the seminal book of Thaler and Sunstein (Thaler and Sunstein 2009). I would argue that it is important to look at the boom the book Nudge set off more closely. Is it just rhetoric? Is something new presented? If we look at the expectations and promises that are connected to the public debate on nudging and behavioural insights, we will find that nudging focuses on enhancing the effectiveness of policy-making. Supra- and international organizations like the EU and the OECD have shifted their focus to include behavioural approaches into the design of policy instruments; the UK and the USA (ante Trump) are leading examples for institution-building aiming at a close nexus between behavioural expertise and governmental bodies (for an excellent overview of the dispersion of behavioural ideas see Straßheim 2017; Straßheim et al. 2015). The EU is interested in using a ‘broader repertoire of policy tools’ to deliver ‘more targeted and effective policy solutions’ (Lourenço et al. 2016, p. 6). The OECD also focuses on ‘making public policies work better’ (OECD 2017, p. 3). The British Behavioural Insight Team (BIT) goal in regards to behaviourally informed policy-making are spelled out in much more detail and they are determined to ‘help ensure that where possible we deliver policy aims by working with the way that people live their lives, rather than interposing – often to little effect – with the crude armoury of the legislating state; and where legislation is necessary, BIT can help ensure that it is designed correctly so that is has the greatest chance of achieving its desired ends’ (BIT 2016, p. 4). Under the Obama administration the ‘Social and Behavioral Sciences Team’ (SBST) aimed at ‘tools

M4716-STRAßHEIM_9781785367847_t.indd 184

25/04/2019 13:32

The enzymatic effect of behavioural sciences  185 for designing the kind of government Americans deserve’ (https://sbst.gov, accessed 15 November 2017) and ‘help Federal agencies advance their policy and program goals and better serve the Nation … to leverage social and behavioural science insights to advance the goals of their policies and programs, demonstrate the impact of these applications, and build capacity for applications of social and behavioural science across Federal agencies’ (SBST 2016, p. II). What these quotes have in common is the search for adequate and effective policies and strategies. One effect of these various searches is the mushrooming of ‘Nudge-Units’ or ‘Behavioural Insight Teams’ (Straßheim et al. 2015). Not only has the number of institutions (teams and units) grown, but also the term ‘nudge’ has grown in importance in scientific publications and the frequency with which it is used has nearly doubled over the past ten years (sourced on Web of Science and Google Scholar). Concurrently both precision and clarity of what the term exactly means have suffered. What needs to be undertaken now is to reach a clear understanding of what the terms ‘nudging’ and ‘behavioural insights’ stand for and through which indicators they can be mapped best. One source from which behavioural insights and ‘nudging’ are won is the knowledge scientists derive from behavioural economics. Narrowing the term to ‘behavioural economics’ means referring to only one specific scientific discipline (similar: Lourenço et al 2016, p. 10). Thaler points out: ‘I view behavioural economics to be economics that is based on realistic assumptions and descriptions of human behaviour. It is just economics with more explanatory power because the models are a better fit with the data’ (Thaler in Samson 2016, p. 23). There is definitely a difference between behavioural economics and ‘nudges’, which is also emphasized by Lunn in his OECD report from 2014: ‘It is important to recognise that behavioural economics and so-called “nudges” are distinct. The former is a scientific subdiscipline’ (Lunn 2014, p. 9). Another source of ‘behavioural insights’ and ‘nudging’ can be found in behavioural psychology, a discipline that can clearly be distinguished from behavioural economics. In order to be able to precisely define ‘nudging’ more deliberations are necessary and they need to take the specific character of the ‘nudging’ instrument, should it exist at all, into account. Thaler and Sunstein who brought the term into the world said ‘A nudge … is any aspect of the choice architecture that alters people’s behaviour in a predictable way without forbidding any options significantly changing their economic incentive. To count as a mere nudge, the intervention must be easy and cheap to avoid’ (Thaler and Sunstein 2009, p. 6). But these two sentences are not sufficient enough to define ‘nudging’ as being different from other instruments, for example, information (see next section). Lepenies and Małecka point out how difficult it is to evolve a distinct picture of what is meant by the term (Lepenies and Małecka 2016, and Chapter 24 in this volume). Their nudging definition sets off from the effects and origins of a nudge. A nudge can produce non-cognitive and non-normative effects, the origin of a nudge can be both cognitive and normative and also involve coercion. Similar to Thaler and Sunstein’s approach, this definition entails aspects that inform about the effect of nudges but the definition itself leaves some room for variation. Since nudges represent a further ‘ingredient’ to the instruments, Lepenies and Małecka’s definition suggests that ‘nudges’ modify instruments. It is therefore reasonable to deduct that the effect or even effectiveness of the instruments depends on whether nudges are included or not. Other authors offer broader ideas and definitions on nudges (e.g. Bornemann and

M4716-STRAßHEIM_9781785367847_t.indd 185

25/04/2019 13:32

186  Handbook of behavioural change and public policy Smeddinck 2016, p. 439) they often, however, do not draw distinctive lines to wellestablished instruments and do not address how the instruments’ mechanisms change. As we will see later, some nudge types cannot be distinguished as just one distinct nudge but share features with command and control, with incentivizing as well as with informational or organizational instruments. In 2015, interestingly enough Lepenies and Małecka suggested to ‘treat … as nudges these policy instruments that rely on the findings of behavioural science and that are intended to impact behaviour in a mode distinct from rational persuasion, command-and-control instruments, or (material) incentives’ (Lepenies and Małecka 2015, p. 428). As we will see later in this chapter the aspect that behavioural insights impact behaviour in a mode distinct from traditional instruments is appropriate but has to be understood as a variety of ‘nudging’ being a component of just these ‘classic’ instruments: Behavioural insights give a spin to these instruments and change the way they work (see section 3 of this chapter). This perspective resembles the EU’s preferred term of ‘behavioural insights’, which is understood as the ‘results of multidisciplinary research in different fields (economics, psychology, neurosciences) to better understand how people behave’ (Lourenço et al. 2016, p. 10). Coming back to Lunn who states that nudging ‘is a particular way to apply its findings to policy, which holds that policy makers should avoid regulations that limit choice (bans, caps, etc.) but can use behavioural science to direct people towards better choices’ (Lunn 2014, p. 9). Actually, as we will see later, they need to combine regulations with behavioural insights. To sum up we see that people’s choices and the way towards decision-making are at the centre of ‘nudging’ approaches and behavioural insights. After gaining a better understanding of decision-making procedures, a variety of tools can be applied to adequately stimulate the addressee. To understand why people behave how they behave and why they decide how they decide is puzzling not only to researchers in behavioural sciences also to but governments and international organizations. With the methods (!) and findings of behavioural sciences they have new strategies at hand. Although the public debate on ‘nudging’ typically refers to classical examples, for example, the image of a housefly in urinals, placing fruits in school cafeterias, organ donation, there is a variety of tools and policies that are assembled under the umbrella term ‘nudging’. Sunstein et al. emphasize that ‘some of these policies take the form of mandates, incentives, and bans, but a prominent set of behaviourally informed tools includes information, warnings, reminders, social norms, and default rules’ (Sunstein et al. 2017, p. 1). Behavioural insights provide the basis for developing adequate tools that ‘nudge’ people in the expected direction. The main focus is on ‘decision heuristics and biases’ on the part of the addressee and the ‘specific effect of the situation or decision context’ (Reisch and Zhao 2017, p. 191). By reviewing the literature on ‘nudging’ in politics and political science we learn that ‘nudging’ primarily serves as an umbrella term and that a number of strategies that deal with people’s behaviour can be grouped together. The term is best suited to mark the results of a process in policy-making that involves evaluating and analysing the circumstances of the addressee’s behaviour. Dependent on the effects that are expected when an instrument is used, the tool itself can differ. What counts most is that it has to impact behaviour and include factors that require behavioural science analytical methods. If the tool – as a result – then takes social or cognitive, normative or other facets into account, which depend on evidence behavioural sciences offer with regard to the respective case,

M4716-STRAßHEIM_9781785367847_t.indd 186

25/04/2019 13:32

The enzymatic effect of behavioural sciences  187 then it can be used as a complement to an existing instrument (or instrument mix) and trigger the ‘behavioural spin’.

3. PROPOSAL: FROM INSTRUMENT TYPOLOGY TO THE ‘BEHAVIOURAL SPIN’ IN PUBLIC POLICY Looking at the AICO-typology we see a typical mechanism that all single types of instruments have in common. The communality is that instruments function in a way that they expect one certain – in the broadest sense – rational response of the addressee that directly corresponds to the stimulus. The addressee is expected to react to the stimulus in a certain way. This stimulus can be an order, a ban, a financial offer, the fear of social exclusion, the wish to be socially accepted, a specific piece of information that leads to a certain behaviour or the offer to participate in a corporate action. Most important is the fact, that policy-makers who choose an instrument do either not take into account that different factors could disturb this assumed rational and coherent response to the stimulus and that they expect the addressee to react unambiguously to the stimulus. Or policy-makers completely ignore thinking about the addressee’s behaviour and the circumstances of his or her habits. This means that instrument users (= policy-makers) assume any form of rationality and do not consider other aspects that influence behaviour, which is something that behavioural scientists would have to be surprised at. Their research shows a variety of external influences that concretely disturb the addressee’s clear preferences or his or her rational considerations when he or she is acting. In the UK, for example, policy-makers introduced a subsidy for insulating lofts aiming at reducing energy consumption but had to recognize that this incentivizing instrument yielded very little effect. The BIT together with Richard Thaler tried to solve the problem of this instrument being ineffective. They found out that a lot of people had practical reasons to not applying for the subsidy: They would have to tidy up their loft first in order to be able to install the insulation tools. That was the reason why the initially designed instrument was not effective. After recognizing that, the subsidy was combined with a help for loft clearance and a commercial partner was integrated into the strategy (Cabinet Office 2011). This example is not particularly reliant on psychological insights but starts with thinking about the addressee and his or her reasons for not responding according to the instrument’s idea. It starts with figuring out the parameters of decision-making-procedures and activities. But not only a variety of external influences can disturb the addressee’s preferences and consideration. Research in behavioural sciences also shows how the human brain works. It underlines that not only does rationality comes into play when humans make decisions but that subconscious and non-rational factors do too. Beyond that: factors of heuristics and biases influence rationality or they are triggered in a way that is not intended. This corresponds with the findings above and explains why the well-known AICO-instruments often do not lead to the expected results. If we think about information, for example, it is extraordinarily important how information is presented. Studies can help to understand that the way in which people react if they are confronted with a medical diagnosis can differ depending on how the numbers are presented. Shafir and colleagues for instance show that ‘experienced physicians made markedly different choices between two

M4716-STRAßHEIM_9781785367847_t.indd 187

25/04/2019 13:32

188  Handbook of behavioural change and public policy a­ lternative treatments for lung cancer – surgery and radiation therapy – depending on whether the outcomes of these treatments were described in terms of mortality rates or in terms of survival rates’ (Shafir et al. 2002, p. 607). There are many other examples that show how the size of letters, colours, the language used and other aspects matter when information is transmitted to an audience. If policy-makers want to learn how to better design information, behavioural sciences come into play. The research conducted by behavioural scientist from different disciplines (economics, psychology) examines decision-making procedures and demonstrates how decisions are conducted. Every decision-making procedure is based on an architecture of choice be it with or without explicit knowledge of human behaviour. Based on experiments as well as mid- and long-time studies, behavioural scientists can show how this architecture of choice has to be built in order to increase the probability of a specific decision being made. Since the features of choice architecture differ and rely on the situation, we find a variety of tools that can be subsumed under the headline of ‘behavioural insights’ (inputs) or ‘nudging’ as the result of behavioural inputs. Reisch and Zhao give a comprehensive overview of these varieties and list default settings, the use of endowment effects, they consider mental accounting, sunk-cost effects, availability heuristics, salience heuristics, anchoring effects, present ways to enhance the power of simplicity or how to utilize the effect of contextual factors (Reisch and Zhao 2017).5 But what does this mean with regard to the existing instruments? Often in the debate on behavioural insights in public policy the claim is made that policy-makers would be holding a new, additional instrument in their hands. But as we see this is not the case since many empirical examples show that at least one of the AICOinstruments or often a combination of two or more AICO-instruments (instrument mix) are always involved; what is different is that by including behavioural science insights, the mechanisms at work change. Let me briefly show this by referencing the case of tobacco regulation. Many different strategies were and are followed to promote non-smoking or other forms of resisting tobacco use. However, policy-makers had the impression that these instruments in terms of taxes, information campaigns and organizational approaches and even strong regulation in the case of teenagers (prohibition) might not be (effective) enough. From previous educational campaigns and through studies on public health we know that effectiveness can be increased when information is framed differently (Roberto and Kawachi, 2015, present a broad spectrum of examples). A first step in this direction was the introduction of labelling, that is, presenting sentences resembling in their style obituaries (‘Smoking kills’, ‘Smoking seriously harms you and others around you’, ‘Tobacco causes cancer’). Here we see that some of the quotes fulfil the function of simple information: ‘Smoking kills’ or ‘Tobacco causes cancer’ – although one could doubt that people would have missed this information before. However, there is also one example of a sentence that shows not only the idea of conveying the information but also that it communicates on a moral level: ‘Smoking seriously harms you and others around you’. Both a societal aspect and a psychological dimension are addressed. Some might argue that this idea is merely a moral pointing finger, but it does take into account that psychological factors play a role. The fairly new and shocking tobacco pictures try to reach the addressee on an emotional level. Studies have shown that these pictures are effective especially in young people if they are still non-smokers and have yet to resist starting to smoke. Shocking pictures are also especially effective in young female smokers, who want to quit smoking

M4716-STRAßHEIM_9781785367847_t.indd 188

25/04/2019 13:32

The enzymatic effect of behavioural sciences  189 (for more information on both findings see Hammond 2011). These pictures try to affect the addressees’ emotion in a much stronger way than the moral sentences above. Simultaneously one could argue that the pictures contained information and that everybody who sees the picture(s) will have a better idea of what happens or could happen inside the body when people smoke cigarettes. Such and other kinds of labelling-schemes are often mentioned when ‘nudging’ is talked about. Deciding on which labelling level is most appropriate depends on what the addressee ‘needs’ to get the intended advice, and could be anything from using a colour-scheme or pictures, to playing with the size of letters, arranging texts and graphics or numbers in a specific way, for example. The crucial point is that experiments and tests (randomized control trials) have previously proven which kind of composition or design of choice architecture is able to powerfully spin the effect of information. Such experiments and tests help to predetermine the conditions of addressee’s behaviour as accurately as possible. Elaborating these conditions helps behavioural scientists to build an architecture of choice that will most probably produce the expected results. Policy-makers can learn how to build such architectures if they work with behavioural scientists. The charm – or to put it differently – the promise of behavioural science lies in the claim that people can still choose between two or more options. However, there is of course a bias in the architecture of choice – otherwise the use of behavioural science would not be attractive: If the instrument or instrument mix is re-designed on the basis of behavioural insights, the instrument or instrument mix gets a specific spin that makes it more probable to reach the desired aim (see Figure 13.1). “Enzymatic Effect” organ

tion opera

izatio

n/co

atio rganiz

o

behavioural insights

organization/cooperation

organization/cooperation

behavioural insights

behavioural insights

n/coo

perati

on

behavioural insights

Behavioural Spain: “Enzymatic mechanism” of behavioural insights activities

policy inputs deduced from a systematical analysis of human behaviour observation and experimentation: e.g. test, learn, adapt, share

technical limits

evidence from behavioural sciences

normative limits

deeply held values that conflict with the policy (Weaver 2015)

coin

(invisible) inputs to the policy process

Figure 13.1 Behavioural spin: ‘enzymatic mechanism’ of behavioural insights6

M4716-STRAßHEIM_9781785367847_t.indd 189

25/04/2019 13:32

190  Handbook of behavioural change and public policy This spin has to be integrated into the concept of instruments. Coming back to the AICO-typology we can now imagine policy inputs deduced from a systematic analysis of human behaviour observations and experiments, for example, test, learn, adapt, share. These inputs operate similar to an enzyme that works with a molecule (substrate) to enforce a reaction. In the biological procedure, the enzymatic mechanism accelerates the chemical reaction in a cell. Behavioural insights change the mechanism of an instrument or instrument mix. This change is possible due to the knowledge of expected human behaviour in a specific situation. Similar to all other single instruments there are also coins and limits to the behavioural spin: Obviously, the spin is only as good as the evidence from behavioural science. From a political point of view the normative limits carry more weight. And also deeply held values that conflict with the policy (Weaver 2014, p. 248) limit not only traditional instruments but also the chances of an effective behavioural spin since the addressee will not (re)act in the aspired manner.7 Weaver also mentioned ‘the heterogeneity of many target populations’ which ‘may make some segments of those populations unresponsive to nudges’ (Weaver 2015, p. 807). And Volkmann does even question ‘whether mere nudges are effective enough and really serve or at least further the purpose they have been introduced for’ (Volkmann 2016, p. 157). Interestingly enough, these critical views show that there is no automatic effect of nudges that manipulates everybody. The fear of an unjust intervention into the individual decision-making procedure could be seen in the light of these objections. Actually, it is most important to discuss under which democratic conditions policy-makers use behavioural inputs into policy-making (Lepenies and Małecka 2016). In terms of conceptual clarity, we could distinguish between ‘behavioural insights’, that is, a variety of different approaches resulting from research in behavioural sciences, as being the trigger of the behavioural spin (see Figure 13.1) and ‘nudges’ serving as the umbrella term for a variety of strategies that result from the enzymatic procedure. But it seems to be more important to again focus on the addressee, who has to be conceptualized as a multidimensional actor. Behavioural sciences consider that not only one specific behaviour impacts decision-making but that it is the interaction of different behavioural dimensions. Howlett suggests to conceptualize different kinds of ‘willingness’ as far as the traditional instruments are concerned (Howlett 2016, p. 17). This chapter proceeds to propose considering different varieties of behaviour that have to do with rational processes and conscious and wilful (Howlett 2016) decision-making processes and with subconscious spheres of human behaviour the way they are analysed by behavioural scientists. The list of categories proposed in Figure 13.2 is not a final list. The systematic analysis of human behaviour will probably add a multitude of such dimensions. However, policy-makers who apply behavioural insights in policy-making will trigger the behavioural spin. If effectiveness is really increased in the end is a question that empirical analysis will have to answer. As mentioned above the debate on nudging and behavioural insights often provokes refusals and emotional reactions. A reason for that could be that the variety of tools that can be used to spin policy instruments includes strategies that try to affect non-cognitive behaviour in a way that is invisible for the addressee. Such strategies strongly refer to the research of Kahneman and Tversky on how the human brain thinks (Kahneman and Tversky 1979, 1984) and who distinguish between system-1 thinking (fast, emotional, intuitive, subconscious) and system-2 thinking (slow, logical, calculating, conscious). In

M4716-STRAßHEIM_9781785367847_t.indd 190

25/04/2019 13:32

The enzymatic effect of behavioural sciences  191 Enzymatic mechanism of behavioural insights activities

policy inouts deduced from a systematical analysis of human behaviour observation and experimentation: e.g. test, learn, adapt, share

technical limits

evidence from behavioural sciences

normative limits

deeply held values that conflict with the policy (Weaver 2015)

coin

(invisible) inputs to the policy process

concept of addressee: the multidimensional actor

Examples

the learning actor

the complying actor the cooperative actor

the reasonable actor

the mis-calculating actor

the spontaneously reacting actor

the calculating actor the lethargic/ non-responsible actor

Examples

the socially embedded actor

the emotionally-led actor

Figure 13.2  Concept of the addressee: the multidimensional actor cases in which only system-1 thinking is addressed, the manipulation concern should be taken seriously. Advocates of ‘Nudging’ argue that ‘Nudges – the applications of libertarian paternalism – used as a policy tool should always be transparent and open for public discourse, and they have to be accepted and supported by the same democratic processes, public debate and critical scrutiny of their costs and benefits as are applied to other political instruments’ (Reisch and Zhao 2017 referring to Sunstein 2016, p. 201). In several contributions Sunstein and Reisch show that transparency and communication do not reduce the effect of ‘nudges’ and they advocate for such transparency. They also present their own worldwide studies on the approval of nudging as policy tools and state that most people welcome behavioural insights in policy-making provided that ‘a nudge has legitimate goals’ and that it ‘fits with the interest or values of most people’ (summarizing: Reisch and Zhao 2017, pp. 201ff).

4. RESULTS/CONCLUSION This chapter aimed at clarifying how the existing instrument typologies can be advanced to better-understand the mechanisms underlying the use of policy instruments. This also involves the question how behavioural insights affect existing instruments or instrument

M4716-STRAßHEIM_9781785367847_t.indd 191

25/04/2019 13:32

192  Handbook of behavioural change and public policy mixes. Most important is the finding, that there is not one (or more) new instrument(s) when we include behavioural insights into policy-making. Rather we have to understand that the mechanism of instrument or instrument-mixes change when addressees are focused on and are no longer merely conceptualized as rational agents. Applying the findings of behavioural research in policy-making requires shifting the focus on the addressee and the question of why they behave the way they do. Taking the AICO-typology of instruments as a starting point that distinguishes between four types it became clear that the addressee’s behaviour used to be expected as a direct and more or less rational reaction to the stimulus of the chosen instrument. Policy-makers and policy observers, however, often criticize these instruments for their poor effectiveness. This criticism, in turn, might have provoked the search for adequate and effective policies and strategies. In contrast to the so far dominating assumption of a rational addressee, behavioural sciences offer a more comprehensive and elaborate approach to human decision-making and have found its way into policy-making. But what happens to instruments from a political science point of view if behavioural insights are integrated into policy-making? Instruments and instrument mixes can get a behavioural spin if policy-makers use behavioural insights to develop policy instruments or mixes of instruments. From a conceptual point of view, this procedure could be labelled as the ‘enzymatic effect’ resulting from a behavioural spin that changes the mechanism of instruments. The debate on behavioural insights has had an impact on the question of how the addressee should be conceptualized. It is important to integrate the concept of the addressee into the existing typologies, which is independent from the use of behavioural insights. But, of course, the knowledge from behavioural sciences has also helped to conceptualize a multidimensional actor and policy-makers should be aware of this multidimensionality if they aim at developing adequate and effective instruments. Moreover, the debate on side effects of behavioural public policy-making is essential. The same is true for the requirements of a constitutional democracy: instruments always have to fulfil democratic requirements and this should also be the case if the instrument’s mechanism changes due to the behavioural spin. If policy-makers keep that in mind and behave responsibly when applying knowledge based on behavioural sciences, if there is clear evidence in a specific case and the addressees agree on the underlying aims, then there is a big chance in triggering the enzymatic effect of behavioural sciences (see van Aaken, Chapter 21 in this volume).

NOTES 1. In Germany, the idea of efficacy is reflected in the initiative ‘Wirksames Regieren’ (https://www.bundesr​ egierung.de/Content/DE/StatischeSeiten/Breg/wirksam-regieren/2017-06-21-mit-buergern-fuer-buerger-re​ gierungsstrategie-wirksam-regieren.html, accessed 1 December 2017). 2. Here, the term ‘mechanism’ is not understood as causal mechanism is used in political analysis (Falleti and Lynch 2009) but as describing a factual process. 3. The literature finds several terms to classify these four instrument types. Authoritative instruments (Schneider and Ingram 1990; Hood 1983) such as ‘command and control’ are known as ‘sticks’ (Vedung 2003), regulatory (Böcher 2012) or regulation (Howlett, 1991, p. 12 in adaptation of Doern and Phidd 1983). Incentivizing instruments are classified as treasure (Hood 1983), incentive (Schneider and Ingram 1990), carrots (Vedung 2003), economic instruments (Böcher 2012) or expenditure (Howlett 1991, p. 12 in adaptation of Doern

M4716-STRAßHEIM_9781785367847_t.indd 192

25/04/2019 13:32

The enzymatic effect of behavioural sciences  193

4. 5. 6. 7.

and Phidd 1988). Instruments that focus on capacity building are named as follows: nodality (Hood 1983), capacity (Schneider and Ingram 1990), sermons (Vedung 2003), persuasive instruments (Böcher 2012), exhortation (Howlett 1991, p. 12 in adaptation of Doern and Phidd 1988). Cooperative and organizational instruments fall into a broader category: organization is presented as a distinct category meaning bureaucratic administration or running public enterprises. Howlett follows a similar idea with ‘public ownership’ (Howlett 1991, p. 12 in adaptation of Doern and Phidd 1983). From a different perspective, another type of instrument is classified as ‘cooperative’ (Böcher 2012) with a view on voluntary agreements, round tables, mediation etc. The terms ‘nodality’ and ‘treasure’ were probably introduced by Hood (Hood 1983) to have an elegant acronym: ‘NATO’-typology. What to a degree is hidden behind this elegant acronym is that the terms ‘nodality’ and ‘treasure’ themselves for the most part lack a precise definition. Of course, one could think of contexts where addressees might wish for strict rules (coercive instruments) since it gives stability and enhances accountability (e.g. same environmental standards for companies instead of voluntary standards). Thaler et al. list incentives, the understanding of mappings, defaults, giving feedbacks, expecting errors, structure complex choices as being the six features and simultaneously the acronym of ‘nudges’ (Thaler et al. 2013). I thank Malte Jessen very much for designing the figures. A similar argument is made by Schubert with reference to different studies especially for ‘nudges’ that use the effect of social norms etc. Schubert argues that the impact of such inputs ‘depend on ideological or other predispositions of nudges (e.g., their ideological priors, their degree of empathy, or whether they hold individualistic or communitarian views)’ (Schubert 2016, p. 28).

REFERENCES BIT (2016): The behavioural insights team: update report 2015–16. Behavioural Insights Team, accessed 4 January 2018 at http://38r8om2xjhhl25mw24492dir-wpengine.netdna-ssl.com/wp-content/uploads/2016/09/ BIT_Update_Report_2015-16-.pdf. Böcher, Michael (2012), ‘A theoretical framework for explaining the choice of instruments in environmental policy’, Forest Policy and Economics, 16, 14‒22. Bornemann, Basil and Ulrich Smeddinck (2016), ‘Anstößiges Anstoßen? – Kritische Beobachtungen zur “Nudging-”Diskussion im deutschen Kontext’, Zeitschrift für Parlamentsfragen, (2), 437‒459. Cabinet Office (2011), ‘Behaviour Change and Energy Use’, London: Cabinet Office. Doern, G. Bruce and Richard W. Phidd (1988), Canadian Public Policy: Ideas, Structure, Process, Toronto: Nelson. Doern, G. Bruce and V. Seymour Wilson (1974), Issues in Canadian Public Policy, Toronto: Macmillan. Ewert, Benjamin and Kathrin Loer (2019), Understanding the Challenges of Policy-Making in Public Health. Theoretical and Political Implications of Behavioural Policies in the Field of Health Promotion and Disease Prevention, Cham: Palgrave Macmillan. Falleti, Tulia G. and Julia F. Lynch (2009), ‘Context and causal mechanisms in political analysis’, Comparative Political Studies, 42 (9), 1143‒1166. Graf, Rüdiger (2015), ‘“Heuristics and biases” als Quelle und Vorstellung’, Zeithistorische Forschungen/Studies in Contemporary History, 12 (3), 511‒519. Hammond, David (2011), ‘Health warning messages on tobacco products: a review’, Tobacco Control, 20 (5), 327‒337. Hood, Christopher (1983), The Tools of Government, London: Palgrave Macmillan. Hood, Christopher and Helen Z. Margetts (2007), The Tools of Government in the Digital Age, London: Palgrave Macmillan. Howlett, Michael (1991), ‘Policy instruments, policy styles, and policy implementation’, Policy Studies Journal, 19 (2), 1‒21. Howlett, Michael (2016), ‘Policy tools and their targets: beyond nudges and utility maximization in policy compliance’, IPSA 2016, 1‒30. Howlett Michael and M. Ramesh (1993), ‘Patterns of policy instrument choice policy styles, policy learning and the privatization experience’, Policy Studies Review, 12 (1),3‒24. Kahneman, Daniel and Amos Tversky (1979), ‘Prospect theory: an analysis of decision under risk’, Econometrica, 47 (2), 263–292. Kahneman, Daniel and Amos Tversky (1984), ‘Choices, values, and frames’, American Psychologist, 39 (4), 341. Lepenies, Robert and Magdalena Małecka (2016), ‘Nudges, Recht und Politik: Institutionelle Implikationen’, Zeitschrift für Praktische Philosophie, 3 (1), 487‒530.

M4716-STRAßHEIM_9781785367847_t.indd 193

25/04/2019 13:32

194  Handbook of behavioural change and public policy Lepenies, Robert and Magdalena Małecka (2015), ‘The institutional consequences of nudging: nudges, politics, and the law’, Review of Philosophy and Psychology, 6 (3), 427‒437. Lourenço, Joana Sousa, Emanuele Ciriolo, Emanuele, S.R. Almeida and X. Troussard (2016), Behavioural Insights Applied to Policy: Overview across 32 European countries, Brussels: European Union. Lunn, Pete (2014), Regulatory Policy and Behavioural Economics, Paris: OECD Publishing. OECD (2017), Behavioural Insights and Public Policy. Lessons from Around the World, Paris: OECD Publishing. Reisch, Lucia A and Min Zhao (2017). ‘Behavioural economics, consumer behaviour and consumer policy: state of the art’, Behavioural Public Policy, 1 (2), 190‒206. Roberto, Christina A. and Ichiro Kawachi (2015), Behavioral Economics and Public Health, Oxford: Oxford University Press. Samson, Alain (2016), The Behavioral Economics Guide 2016 (with an introduction by Gerd Gigerenzer), accessed at 4 January 2018 at http://www.behavioraleconomics.com. SBST (2016), Social and Behavioral Sciences Team 2016 Annual Report, Executive Office of the President National Science and Technology Council, accessed 4 January 2018 at https://sbst.gov/download/2016%20 SBST%20Annual%20Report.pdf. Schneider, Anne and Helen Ingram (1990), ‘Behavioral assumptions of policy tools’, The Journal of Politics, 52 (02), 510‒529. Schubert, Christian (2016), ‘Green nudges: do they work? Are they ethical?’, Ecological Economics, 132, 329‒342. Shafir, Eldar, Amos Tversky, E.E. Smith and E.E.D. Osherson (2002), ‘Decision making’, Foundations of Cognitive Psychology: Core Readings, 601‒620. Simon, Herbert (1955), ‘A behavioral model of rational choice’, The Quarterly Journal of Economics, 69, 99‒118. Straßheim, Holger (2017), ‘Die Globalisierung der Verhaltenspolitik’, in Katharina Hirschbrunn, Gisela Kubon-Gilke and R. Sturn (eds), Jahrbuch Normative und institutionelle Grundfragen der Ökonomik, Bd. 16, Kapitalismus, Globalisierung und Demokratie, Metropolis: Marburg, 211‒242. Straßheim, Holger, Arlena Jung and R.-L. Korinek (2015), ‘Reframing expertise: the rise of behavioural insights and interventions in public policy’, in Ariane B. Antal, Michael Hutter and D. Stark (eds), Moments of Valuation. Exploring Sites of Dissonance, Oxford: Oxford University Press, 249‒268. Sunstein, Cass R. (2016), The Ethics of Influence: Government in the Age of Behavioral Science, Cambridge: Cambridge University Press. Sunstein, Cass R., Lucia A, Reisch and J. Rauber (2017), ‘A world-wide consensus on nudging? Not quite, but almost’, Regulation & Governance, online first. Thaler, Richard H. and Sunstein, Cass R. (2009), Nudge. Improving Decisions about Health, Wealth, and Happiness, New York: Penguin. Thaler, Richard H, Cass R. Sunstein and J.P. Balz (2013), ‘Choice architecture’, in Eldar Shafir (ed.), The Bahavioural Fondations of Public Policy, Princeton, NJ: Princeton University Press, 428‒439. Thießen, Malte (2013), ‘Vorsorge als Ordnung des Sozialen: Impfen in der Bundesrepublik und der DDR’, Zeithistorische Forschungen / Studies in Contemporary History, 10, 409‒432. Vedung, Evert (2003), ‘Policy instruments: typologies and theories’, in Marie-Louise Bemelmans-Videc, Ray C. Rist and E. Vedung (eds), Carrots, Sticks and Sermons. Policy Instruments and Their Evaluation, New Brunswick, NJ: Transaction Publishers, 21‒58. Volkmann, Uwe (2016), ‘Nudging, education, paternalism: a philosophical perspective from the Old Europe’, in Alexandra Kemmerer, Christoph Möllers, M. Steinbeis and G. Wagner (eds), Choice Architecture in Democracies. Exploring the Legitimation of Nudging, Baden-Baden: Nomos, 141‒160. Weaver, R. Kent (2014), ‘Compliance regimes and barriers to behavioral change’, Governance, 27 (2), 243‒65. Weaver, R. Kent (2015), ‘Getting people to behave: research lessons for policy makers’, Public Administration Review, 75 (6), 806‒816.

M4716-STRAßHEIM_9781785367847_t.indd 194

25/04/2019 13:32

14.  The politics of nudge and framing behaviour change in health Muireann Quigley and Anne-Maree Farrell1

1. INTRODUCTION In recent years, behaviourally-inspired public policy has captured the political imagination. To a large extent, we can attribute this upsurge in political interest to Richard Thaler and Cass Sunstein’s work on nudges. In order to bring about desired behavioural change in the citizenry, they argue that public policy-makers need to focus on altering the contexts in which individuals make decisions, through trying to harness or eliminate their cognitive biases (Thaler and Sunstein 2009). Although nudges have been employed in a range of policy sectors, one area of recent focus has been on promoting individual behaviour change in relation to ‘lifestyle risks’, with significant focus on tobacco, alcohol, and dietary factors (Alemanno and Garde 2015). In this chapter we critically examine this focus, asking whether we should be embracing the use of behaviourally-inspired public policy initiatives as a way of encouraging and enabling individuals to lead healthier lives (see also Leone and Tallacchini, Chapter 11 in this volume). Or conversely, should we be wary of the potential promise they hold, given the political context which frames their use? In addressing these questions, we draw on the concept of framing in order to examine our key arguments. In doing so, we acknowledge that ‘frames’ and ‘framing’ analysis have been used in a diverse range of ways across a number of disciplinary literatures (see Goffman 1974; Tversky and Kahneman 1981; Snow et al. 1986; Schön and Rein 1994). Drawing inspiration from the approach taken to frame analysis within the policy studies literature, we are specifically interested in examining how and why nudges, and behavioural insights more generally, have been framed as an innovative way forward in health promotion. This requires that we critically examine what set of problems nudges are designed to address in what is a highly contested area of public policy, as well as why they are preferred by those in political leadership (see Bacchi 2009). Adopting such an analytical framework also allows us to explore any ‘mismatch’ between policy intent and outcomes and the reasons why this might be the case (van Hulst and Yanow 2016, p. 92). Specifically, we argue that the political context which frames the use of nudges in the field of health promotion has led to a ‘mismatch’ between policy intent and implementation. For present purposes, we define policy ‘intent’ as the desire to bring about behaviour change to improve individuals’ health; whereas policy ‘outcomes’ refers to the evidentiary basis which shows that such change has been achieved. In order to make our case, we first examine nudge as a political project, which includes briefly tracing the international spread of behavioural insights in public policy-making. Thereafter, we illustrate the mismatch problem with reference to two case studies: health promotion and organ donation. Regarding the first of these, we argue that behavioural health promotion is focusing too much on individual lifestyle preferences, with insufficient account being 195

M4716-STRAßHEIM_9781785367847_t.indd 195

25/04/2019 13:32

196  Handbook of behavioural change and public policy taken of the wider social determinants of health. Organ donation is then used to show how nudges may be an inadequate tool for addressing complex public policy problems in health. Following this we broaden our critique to briefly look at some current difficulties regarding the evidence base for behavioural public policy. At the end of the chapter we return to the political context in which the design and implementation of behavioural health-related policies has taken place, and consider how the framing of such strategies affects their political legitimacy.

2.  NUDGE AS A POLITICAL PROJECT The publication of Nudge provided the catalyst for renewed political interest in behavioural strategies (Thaler and Sunstein 2009). This has contributed to a recent upsurge in governmental or quasi-governmental units which seek to explicitly integrate findings from the behavioural sciences into public policy-making. Indeed, the use of nudges has been championed by those on both the conservative and progressive sides of politics (John et al. 2013). This has much to do with the way in which the use of the behavioural sciences generally, and nudges specifically, have been framed in political terms (Quigley 2013, 2018). First, the very use of the term ‘nudge’ has been promoted as fairly innocuous and something that we need not be overly concerned about, thus adding to its political allure (Quigley 2018). Second, they are often promoted as offering a bipartisan way forward out of political gridlock. Third, they seemingly offer a ‘low cost’ approach in terms of design and implementation, as well as achieving ‘more for less’ in delivering public services. Fourth, it has been suggested that they are ‘easy to avoid’ (thus purportedly countering any concerns about the potential of nudges to infringe upon individuals’ freedom of choice). Finally, they are promoted as a viable alternative to what is presented as cumbersome and inflexible regulation (see Thaler and Sunstein 2009; Dolan et al. 2010). Given the growing propensity for cross-border policy transfer (Dolowitz and Marsh 2000), the positive political framing of nudges has led to the international spread of behavioural policy units (Behavioural Insights Team 2016a). Such framing has made nudge strategies a malleable policy tool (see Rein and Schön 1996), although this very malleability has meant that they are susceptible to a range of applications, informed by diverse political preferences and ideologies. In the UK, for example, a focus on ideas generation, which has its roots in translational academic research, has become a feature of the marketing of political messages in recent decades. For Blair’s New Labour government in the 1990s, it was the Third Way agenda, inspired by the work of leading sociologist, Anthony Giddens (Driver and Martell 2000). For former UK Prime Minister David Cameron in the mid 2000s, it was about the promotion of nudges, underpinned by the work of Thaler and Sunstein. This also melded well with his government’s neoliberal Big Society agenda, which championed greater individual initiative and self-reliance at the expense of a more expansive welfare state (Burgess 2011; Quigley 2013, pp. 9‒10; Slocock 2015). During his time in government, Cameron supported the establishment of the Behavioural Insights Team (BIT) (‘Nudge Unit’). In 2013, this became co-owned by its employees, the UK government, and NESTA, the UK’s leading innovation charity (Plimmer 2014). The BIT describes itself as the ‘world’s first government institution dedicated to the

M4716-STRAßHEIM_9781785367847_t.indd 196

25/04/2019 13:32

The politics of nudge and framing behaviour change in health  197 application of behavioural sciences’ (BIT 2016b; see also Dolan et al. 2010). The use of behavioural insights in public policy-making was also taken up with enthusiasm by the then Obama administration in the US. Buoyed by the success of the BIT in the UK, the Social and Behavioural Sciences Team (SBST) was established with the aim of translating ‘findings and methods from the social and behavioral sciences into improvements in Federal policies and programs’ (United States Government 2016). Specifically, this included the need for US federal agencies to ‘identify policies, programs, and operations for which behavioral insights may yield substantial improvements in social welfare and program outcomes; develop strategies for implementing these insights and rigorously testing and evaluating their impacts; and to seek opportunities to strengthen agency relationships with the research community to better utilize findings from the social and behavioral sciences’ (White House 2015). At supranational level, the European Commission (Commission) also established the Framework Contract for the Provision of Behavioural Studies to EU policy-makers in order to ‘facilitate the running of behavioural studies in support of EU policy-making’ (van Bavel et al. 2013, p. 8). The commitment to this approach was also underlined by the EU Commissioner for Consumer Policy, who stated that ‘[b]ehavioural tools play a pivotal role in smarter regulation’ (Mimica 2013). In 2016, the Joint Research Centre – the Commission’s ‘in-house science and knowledge service’ – published an extensive report on the application of behavioural insights to policy in Europe (European Commission, Joint Research Centre 2016). This initiative is seen as aligning with the Commission’s ongoing commitment to the use of evidence-based policy, which informs its Better Regulation agenda (see also Alemanno, Chapter 10 in this volume). This has informed the most recent iteration of the Commission’s Better Regulation Guidelines and the accompanying Better Regulation ‘Toolbox’, which explicitly feature behavioural approaches (European Commission 2015, 2018). What the overview in this section highlights is both the alacrity and the enthusiasm with which insights from behavioural sciences research have been taken up in public policy-making at both national and supranational levels in recent years. It has crossed not only geographical, but also political, divides, albeit with differing emphases reflecting diverse political and bureaucratic cultures. Such developments also necessitate caution with respect to any claims about political neutrality with respect to the use of nudges. This is particularly so with respect to their application in politically sensitive sectors such as health promotion, which we consider in further detail next.

3. THE POLITICS OF BEHAVIOUR CHANGE IN HEALTH PROMOTION There is a long tradition within health policy of drawing on a range of strategies to try to change citizens’ health behaviours. A particular focus has been on changing individual behaviour vis-à-vis lifestyle risks: alcohol, smoking, and diet and exercise (Alemanno and Garde 2015). This broadly fits with the view that an individual’s health is a question of personal responsibility, subject to lifestyle choice. This is underpinned by a commitment to upholding individual freedom and choice, in circumstances where state intervention in the lives of its citizens should be kept to a minimum (Quigley 2013, pp. 591‒592). Such

M4716-STRAßHEIM_9781785367847_t.indd 197

25/04/2019 13:32

198  Handbook of behavioural change and public policy commitment is often to be found as a core political belief of a broader neoliberal worldview. In such circumstances, it may be resistant to change even in the face of evidence to the contrary (Sabatier 1998; Ayo 2012; Cahill et al. 2012). One of the benefits of such a worldview is that health interventions (be they campaigns or direct interventions) which focus on individual lifestyle risks, such as the need to reduce alcohol intake, may be quicker and easier to implement than the sustained effort and investment that would be required to alleviate poverty in vulnerable populations (Wikler 1987; Minkler 1999). It also means that managing any adverse health consequences also becomes a matter of individual responsibility, even in the face of evidence that they are likely to be borne predominantly by those who are worse off in terms of socioeconomic and educational status (McKinlay and Marceau 2000; Glass and McAtee 2006; Karpin and O’Connell 2017). This individualistic focus has continued with the new behavioural turn. For instance, in the UK, an early BIT report focused specifically on health. Whilst a range of interventions was discussed in the report, the focus on lifestyle risks comes through strongly. Indeed, half of the suggested interventions deal with smoking, alcohol, diet and weight, and physical activity (Behavioural Insights Team 2010). At EU level, smoking was a prominent health-related example in the JRC’s first report (van Bavel et al. 2013, pp. 4‒6). Meanwhile, the majority of examples provided in the health section of the most recent JRC report are again related to lifestyle risks: the salt content of food in Estonia, healthy eating in Iceland, childhood obesity in Croatia, and numerous initiatives in different countries relating to smoking (European Commission JRC Report 2016, pp. 23‒27). Running through these reports is the clear expectation that insights from the behavioural sciences can be marshalled in order to ‘nudge’ people towards making decisions (and taking actions) which are better for their health. The focus on risks associated with individual lifestyle factors is problematic for a number of reasons. First, evidence regarding the success of strategies which target these types of individual health behaviours can be described as mixed at best (Glass and McAtee 2006; Vallgårda 2012; Baum and Fisher 2014). For example, targeting individual behaviours to address obesity through weight loss interventions, have largely been proven to be ineffective (Bombak 2014). Moreover, a focus on interventions which target individuals have ignored the complexity of the problems they are aimed at solving, overlooking vulnerabilities experienced by target populations. These include the differing resources available when trying to make dietary changes, as well as the need to make trade-offs (economic and otherwise) in order to do so (Ferrer et al. 2014; Woolf 2011). Second, it has long been recognized that multiple social and other factors influence health outcomes (e.g. education, socioeconomic status, and employment) (Independent Inquiry into Inequalities in Health 1998; Marmot Review 2010; World Health Organization 2016). Changing behaviour to improve health, therefore, is a complex and multifaceted problem (Howlett and Mukherjee 2017; see also Weaver, Chapter 16 in this volume). It requires that policy-making for the purposes of health promotion needs to focus on prevention and protection, as well as structural health reform, to ensure access and equity for citizens. Adequately addressing these root causes of ill health requires political will and sustained financial and infrastructure investment (Commission on the Social Determinants of Health 2008; Marmot Review 2010). It also requires accepting a greater role and responsibility (for failure to act) on the part of those in political leadership. Yet

M4716-STRAßHEIM_9781785367847_t.indd 198

25/04/2019 13:32

The politics of nudge and framing behaviour change in health  199 the framing of behavioural public health policy problems and the proposed interventions take place in almost complete isolation from these broader health promotion debates and strategies. There is very little overlap between the literatures which deal with the use of behavioural sciences research in public health policy and those which focus on the social determinants of health (MacKay and Quigley 2018). The risk with this is that we will continue to reinvent the wheel of past policy failures and perpetuate existing mismatches between policy intent and implementation with respect to health promotion. For instance, in health promotion policy-making in general there has been a propensity towards ‘lifestyle drift’. This is the ‘tendency for policy initiatives on tackling health inequalities to start off with a broad recognition of the need to take action on the wider social determinants of health (upstream), but which, in the course of implementation, drift downstream to focus largely on individual lifestyle factors’ (Marmot Review 2010, p. 3). Behavioural public health policy in the main has no scope for ‘drift’ per se, because the framing of the problems and the proposed interventions have a built-in focus on individual lifestyle factors. The concern, therefore, is that this will continue to distract us from the fact that there are deeper socioeconomic and educational factors which substantively influence health throughout people’s lives. If this occurs, then significant improvements in health at a population level are likely to remain elusive. Third, for those who advocate for the use of nudges in health promotion, as well as behaviourally-informed public policy more generally, the tendency has been towards framing their use in opposition to the use of (what can only be presumed is ‘hard’) regulation to achieve desired individual behaviour change. In the UK, for example, this oppositional stance has more often than not been aligned with a neoliberal political agenda, expressed as a desire to cut ‘red tape’, avoid transactional costs, and promote flexibility associated with designing and implementing regulation (Behavioural Insights Team 2010; Quigley 2013). More recently, it has been used for political positioning purposes in response to the UK’s fraught relationship with the European Union (EU): ‘wherever possible, the Government will argue for alternatives to regulation at European level, drawing on behavioural science insights … [and] … wherever possible, [it will] seek to implement EU policy and legal obligations through the use of alternatives to regulation’ (HM Government 2013, General Principles, paras 2 and 5a). This oppositional stance is at odds with the more nuanced and complex view taken within the relevant regulatory studies literature. In this context, regulation is seen as being on a continuum – ranging from soft to hard measures – involving an increasingly diverse range of stakeholders and organized interests in regulatory governance, both within and beyond the state (LeviFaur 2013). In health promotion, regulation in its varying forms has long been part of the governance toolbox, designed to increase state capacity to act, to provide incentives, to prohibit conduct, and to create environments in which desired change can take place (Gostin 2000; MacKay 2011; Alemanno 2012). With this in mind, the oppositional framing of nudges and regulation in matters of health promotion should perhaps best be viewed as a heuristic device to serve a range of political ends, and which may coalesce at times with a neoliberal worldview depending on the health issue to be addressed. Current political enthusiasm for embracing behaviourally-inspired strategies, such as nudges, nevertheless warrants critical assessment in terms of their aims and objectives, as well as the evidence base for their effectiveness. It is these aspects which we consider in the next two sections.

M4716-STRAßHEIM_9781785367847_t.indd 199

25/04/2019 13:32

200  Handbook of behavioural change and public policy

4. ORGAN DONATION AND COMPLEX PUBLIC POLICY PROBLEMS An important, but underexplored question, is whether nudges and other behavioural techniques, which have been politically positioned as offering an innovative way forward in bringing about such change, are suitable for addressing complex public policy problems. We focus on the example of organ donation in order to explore this question. This is because, while organ donation is viewed by some as fitting for the application of behavioural insights, deeper analysis highlights the complexities involved in health-related policy problems. By way of background, it is important to note that the need to increase organ donation rates is recognized as a pressing problem in both public health and political terms. There is a growing gap between demand and supply for organs to address the increasing incidence of end stage organ failure. This is related to an ageing population, with a higher incidence of co-morbidities, such as diabetes and hypertension. It has occurred in lockstep with medico-scientific and technological advances that have substantially increased the potential for successful organ transplantation outcomes in recent years and further fuelled demand for organs (Murphy and Smith 2012). Two ways in which nudges could be harnessed to this end are opt-out organ donation and increasing organ donor register numbers. For both of these we note some difficulties. The majority of countries with developed organ donation and transplantation systems operate on an opt-in basis: that is, individuals (or their families after their death) must actively choose to become donors, usually by signing up to an organ donation register. A number of commentators, however, advocate for the use of opt-out systems of organ donation as the default position, as a way of increasing the number of organs available for transplantation (Saunders 2010; Rockloff and Hanley 2014). In such a system, individuals are treated as agreeing to become organ donors unless they have actively opted out during their lifetimes (Farrell and Quigley 2012). The behavioural rationale behind this approach is that people tend to display status quo bias. They display inertia and tend not to move away from the default option. While changing defaults can be powerful, it is not clear how effective it is in the organ donation context. There is some debate as to whether changing the legal default in this particular area causes the desired increase in the numbers of organs available. A number of reviews of existing studies suggest that it does (Rithalia et al. 2009; Palmer 2012). In contrast, other commentators have argued that the question of whether it is the change in the legislative default which makes the difference remains unanswered. This is because studies that have been conducted thus far have not, and indeed cannot, control for confounding factors. These include national cultural and social change; concurrent changes to national transplantation infrastructure and organization; and advertising campaigns about organ donation conducted at the same time as the change in the default (Willis and Quigley 2014). As such, there is a need for caution regarding any claims made that changing the default is causative. What then about opt-in systems of donation? One strategy is to try to increase the numbers of people who sign up to the national organ donor register, thus increasing the potential pool of organ donors. In this vein, the UK BIT conducted a Randomized Controlled Trial (RCT) to see if people could be nudged to do so. This involved comparing and contrasting various messages regarding

M4716-STRAßHEIM_9781785367847_t.indd 200

25/04/2019 13:32

The politics of nudge and framing behaviour change in health  201 organ donation that appeared on a UK government webpage, following completion of online renewal of their drivers’ registration. Eight different messages were randomly assigned over a five-week period, which appeared on the webpage as people came to the end of their online renewal process. During this period, over one million people viewed the messages. Some of the messages had images and the others had text, and all drew on insights from behavioural sciences research to see what worked best in terms of encouraging organ donor registrations. The findings from the study showed that the message which relied on reciprocity – ‘if you needed an organ transplant, would you have one? If so, please help others’ – proved to be the most popular. BIT suggested that the findings showed that if the reciprocity message was used over the course of a year, then it would be likely to generate a further 96 000 extra registrations for organ donation. Hailing the findings as ‘impressive’, BIT went on to state that the findings were already having an impact on the marketing messages used by the national transplant coordination agency (Behavioural Insights Team 2013). The BIT study is focused on maximizing organ donation registration rates. But how should we measure policy success here? The study was experimental and of limited duration, therefore extrapolation claims with regard to the potential increase in organ donor registration rates over the course of a year are speculative and need to be treated with caution. If the ultimate policy objective is successful organ transplant outcomes, then this requires that a complex interplay of ethical, institutional, clinical, and regulatory factors be taken into account in policy design and implementation (Quigley and Stokes 2015, pp. 64‒65). Indeed, deciding to register as an organ donor can be complicated and difficult. A ‘one click’ (or ‘one tick’) nudge cannot deal with the complexity of such decisions, as other studies have demonstrated. For instance, studies have been conducted in the UK to examine whether individuals could be ‘nudged’ towards registering as organ donors (John et al. 2013). One group of researchers found that using nudges in an information booklet, in the form of images and texts that promoted social norms supportive of organ donation, was effective in getting participants to register as organ donors. However, this proved not to be the case when the information booklet was combined with focus group discussion. This was because the discussion made participants more aware of the complex issues involved in deciding to become an organ donor and therefore made them uncertain about whether or not to register. The researchers concluded that there is a need for greater deliberation when complex issues such as freedom and informed consent are at stake and this was necessary if nudges and other behavioural techniques were to be considered as legitimate policy tools (John et al. 2013, pp. 101‒108). A further option to increase the rate of organ donation is a mandated choice model. Essentially, this is a required choice on the part of individuals as to whether to donate organs or not. This option is favoured by Thaler and Sunstein because individuals’ views regarding donation would be known, thus making it easier for families to come to terms with the donation of a family member’s organs after death. They also argue that it could be implemented at relatively low cost, with individuals required to tick a box at the time of renewal or registration of their drivers’ licences (Thaler and Sunstein 2009, p. 169). To a large extent, they reach this conclusion because mandated choice is likely to be more politically acceptable than an opt-out system. However, neither opt-out systems of donation nor the BIT RCT nor mandated choice

M4716-STRAßHEIM_9781785367847_t.indd 201

25/04/2019 13:32

202  Handbook of behavioural change and public policy models adequately engage with the complex and often fraught context in which family decision-making about organ donation takes place immediately following the loss of a loved one. Nor do they take account of the importance of professional duties and obligations impacting the doctor-patient relationship, as well as end of life care more generally in this context (Truog 2012). Studies such as the BIT one, which focus on effectiveness, also largely ignore what is a voluminous cross-disciplinary literature on the issue. These literatures explore the ethical, social and legal issues involved in the decision-making process to donate organs, as well as the significant support infrastructure required to facilitate successful transplant outcomes following organ donation (Farrell et al. 2011; Farrell 2015, p. 268). By focusing on effectiveness we only see part of the picture and risk missing the complexity of the problem at issue.

5.  THE EVIDENCE BASE As our discussion so far has illustrated, there are problems with the way that the behavioural sciences have been marshalled to health policy ends. One pressing issue, which has not yet been adequately addressed, is the evidence required to implement behavioural public policy. Some of the difficulties in this respect can be highlighted by considering the Commission’s tripartite taxonomy for classifying behavioural policy initiatives which is set out in their most recent report (European Commission JRC Report 2016, see also Ciriolo et al., Chapter 8 in this volume). The first category is behaviourally-tested initiatives. These are ones which are ‘explicitly tested, or scaled out after an initial ad-hoc experiment’ (p. 16). The second, behaviourally-informed initiatives, are ones ‘designed after an explicit review of previously existing behavioural evidence, although not benefiting from any specific prior ad-hoc experiment’ (p. 16). The third category consists of behaviourally-aligned initiatives. These draw on behavioural insights, although they ‘do not rely explicitly on any behavioural evidence, be it available literature or evidence coming from an ad-hoc test’ (p. 16). The difference between these categories is the strength of the evidence behind each of them. Yet they all arguably rely on policy-makers too readily buying into questionable data. For instance, even if real-world studies are done such as in category one, there may be problems of applicability and transferability between settings (e.g. between different countries or between the finance and health arenas, and so on). This is the case even if RCTs are used. These are a recent import from medical research into the policy arena and are widely viewed as representing an empirical gold standard (e.g. BIT 2012; Alemanno and Spina 2014, pp. 442‒443). Yet, all any particular RCT tells us is that an ‘intervention worked somewhere, at some time, under certain conditions’ (Quigley and Stokes 2015, pp. 73‒74; see generally Cartwright and Hardie 2012, pp. 122‒134). More is needed to get us from ‘this policy worked there’ to ‘our proposed policy will work here’ (Cartwright and Hardie 2012, p. 8). For instance, it cannot be taken as a given that a RCT conducted in one EU Member State will provide data that is applicable in another Member State. Heterogeneity exists along multiple dimensions within and across the EU: socially, economically, culturally, and so on. One or more such dimensions may impact on the transferability and applicability of behavioural policy studies as between EU Member States (Quigley and Stokes 2015). Moreover, as Deaton points out, RCTs are ‘unlikely to

M4716-STRAßHEIM_9781785367847_t.indd 202

25/04/2019 13:32

The politics of nudge and framing behaviour change in health  203 be helpful for policy unless they tell us something about why it works, something to which they are often neither targeted nor well-suited’ (Deaton 2009, p. 42; see also Deaton and Cartwright 2016). There are also problems with the weaker types of evidence represented by JRC categories two and three. Category two ones, in drawing on existing research but not being specifically tested in the particular policy context, may not capture the complexity of the day-to-day reality of human behaviour. This is because they may be relying on studies from a wide array of laboratory-type experiments, rather than those tested in real-world settings (Berg and Gigerenzer 2010; Quigley and Stokes 2015). Category three initiatives do not rely ‘explicitly on behavioural evidence’. Given this, they seemingly amount to nothing more than best guesses, rather than providing a strong evidence base for effective public policy. Leaving these aside, a further issue also compounds the empirical and methodological challenges regarding the use of the behavioural sciences in public policy-making. Behavioural psychology is undergoing what has been called a ‘replicability crisis’; that is, the replicability of some psychological research is poor, with suggestions that only up to 47 per cent of studies could be replicated (Open Science Collaboration 2015). Pressure to publish new empirical findings means that studies testing replicability are not given priority. The consequence is that the replications needed to produce robust results are missing. Despite this, particular research in the field has gone on to become ‘canon’ and is frequently cited, including beyond its original disciplinary borders, such as in legal and philosophical scholarship (Gilbert et al. 2016). This is problematic because policy-makers are attempting to apply the findings from a body of research which does not yet have a stable and critical core which can just be applied in practice (Priaulx and Weinel 2014, p. 372).

6. ANALYZING THE POLITICAL CONTEXT IN FRAMING BEHAVIOUR CHANGE IN HEALTH A critical analysis of how nudges, and behavioural strategies more generally, are framed in the political context is helpful in unpicking (some of) the reasons for the current political enthusiasm for their use. The surge in political interest in nudges occurred at a time of financial austerity brought about by the global financial crisis. Nudges were presented as low cost and easy to avoid, drawing on individual preferences and cognitive biases, whilst also preserving freedom of choice. It also coalesced with publics in many liberal democracies becoming increasingly disillusioned with traditional political parties, which precipitated the search for new ways of political thinking and action (Smith 2014; della Porta 2015). As mentioned earlier in this chapter, the use of nudges also focuses on individual behaviour change, rather than taking account of the social determinants of health. This chimes well with a neoliberal worldview which is concerned with market freedoms and individual responsibility for health and wellbeing. As a result, the recent behavioural turn in public policy on the part of national governments of different political persuasions (or ideologies) is not wholly unexpected. The appeal of nudges in health promotion is likely to continue, despite the methodological and evidential flaws in the approach taken to the use of behavioural insights in the

M4716-STRAßHEIM_9781785367847_t.indd 203

25/04/2019 13:32

204  Handbook of behavioural change and public policy field (see Galizzi and Wiesen 2017). Their continuing appeal derives from a combination of historical and biomedical factors favouring an individualized focus, and the fact that political leaders in many liberal democracies are wedded to a particular framing of policy problems which is rooted in a neoliberal worldview (see also Zuidhof, Chapter 12 and Tyers, Chapter 23 in this volume). This is likely to be resistant to change, notwithstanding evidence to the contrary. To do otherwise and to focus instead on the social determinants of health, for example, could be viewed as a direct challenge to existing structural and power asymmetries perpetuated by political, financial and corporate elites. This is something which the available evidence points to as having significantly contributed to growing health inequalities in liberal democracies (and beyond) (Commission on the Social Determinants of Health 2008; Hastings 2012; Baum and Fisher 2014). If we accept that the neoliberal worldview largely informs the framing and use of nudges in health promotion, what might this mean for the political legitimacy of such strategies in the context of democratic politics? It is suggested that questions of legitimacy can be assessed by reference to constitutional, functional, values-based and democratic criteria (Casey and Scott 2011). It may also include taking account of whether particular policy strategies have infringed rights or caused harm to citizens (see Yeung 2012). For present purposes, it is the latter three criteria which are particularly relevant. Functional legitimacy goes to questions of efficiency and effectiveness in using nudges in health promotion. As highlighted earlier in the chapter, the evidence base regarding whether they are effective in bringing about desired policy outcomes has not yet been established, with some success noted with respect to limited, small-scale (experimental) interventions. The values-based criterion engages with the underlying principles that should inform the use of nudges in health promotion (see generally Farrell 2015, pp. 277‒281). As things currently stand, the predominant principle is one based on individual choice and responsibility for one’s health and wellbeing, rather than a broader engagement with the social determinants of health at a population level. The democratic criterion goes to participation, transparency and accountability. Participation invites deliberation, which is highly relevant for addressing complex policy problems (Lafont 2015). As was highlighted in the organ donation example discussed earlier, the complex interplay of values, culture and clinical issues at stake requires active deliberation on the part of those deciding to register as organ donors, as well as their families in the event of their death. ‘One click’ nudges are simply unable to engage with such issues, making them unsuitable for decision-making in this context (Truog 2012; see Powys White et al. 2012). On the question of transparency involved in the use of nudges, Thaler and Sunstein (2009) have argued that we should rely on the Rawlsian publicity principle, which recognizes information disclosure as the primary method by which transparency is achieved (pp. 244‒245). It has been suggested that there may be policy areas where questions of transparency in the use of nudges do not loom large (Bovens 2008). However, this remains problematic in the context of democratic politics, given the reliance in designing nudges on unconscious cognitive biases in human decision-making. It also assumes that we can trust those policy-makers engaged in the design of nudges, and that they do not have their own political preferences or agendas in promoting the use of nudges and other behavioural techniques in health promotion (Farrell 2015, pp. 278‒280). While conceptualizing the relationship between transparency and accountability in

M4716-STRAßHEIM_9781785367847_t.indd 204

25/04/2019 13:32

The politics of nudge and framing behaviour change in health  205 governance processes is not straightforward (Hood 2010), there is a necessary connectivity between the two concepts in the context of policy design involving nudges. In these circumstances, the need for political accountability looms large for a number of reasons. First, public policy-makers engaged in health promotion will need to deal with the clout of financial and corporate elites, such as those involved in the tobacco and food industries. It is therefore important that policy-makers are accountable with regards to any potential or real conflicts of interest they may have in their dealings with such elites. Second, it is important that the ethical probity of public policy-makers is transparent. This is vital where public trust is integral to the continued success of public services, such as organ donation and transplantation programmes (Farrell 2015, p. 280). Finally, accountability on the part of policy-makers may be facilitated by the appropriate use of regulation, which was traditionally operated as a preferred technique of legitimation in the context of democratic politics (Brownsword 2008), redressing power asymmetries and socioeconomic inequalities that adversely impact upon health promotion activities.

7. CONCLUSION This chapter examined how the political context influences the framing of nudges, and other behavioural techniques, and their consequent impact on desired behaviour change in health promotion. Framing analysis calls attention to the dynamic process involved in constructing policy problems, which may result in a mismatch between policy intent and outcomes. We examined how the framing of nudge strategies as an innovative way forward in health promotion had resulted in insufficient account being taken of the interrelationship between individual lifestyle preferences and social determinants impacting health. This occurred even where there was little, if any, evidence that nudge strategies were effective in realizing desired behaviour change. What evidence was available pointed to limited success in relation to small scale interventions. Such strategies were also inadequate in addressing complex policy problems, particularly where there was a need to engage with the social, cultural, familial and clinical dynamics which impact the way in which health decision-making takes place. Nevertheless, current political enthusiasm for nudges, and other behavioural techniques, which focus on changing individual behaviour with regard to identified risky lifestyle factors remains high. This can be attributed in large part to the fact that it fits within an overarching neoliberal worldview that holds fast to the idea of a limited role for the state in the lives of its citizens, as well as support for individual choice and responsibility for one’s health and wellbeing. Such political beliefs inform the framing of nudge strategies, and are likely to be resistant to change, even in the face of evidence of limited effectiveness in terms of desired policy outcomes. It is therefore important that the politics of nudge are taken into account in critically assessing their usefulness in health promotion. This is in addition to considering their broader political legitimacy in the context of democratic politics.

M4716-STRAßHEIM_9781785367847_t.indd 205

25/04/2019 13:32

206  Handbook of behavioural change and public policy

NOTE 1. The first-named author would like to thank Kathryn MacKay for her research assistance; the second-named author’s contribution draws on research done while in receipt of an Australian Research Council (ARC) Future Fellowship (FT130101768). The support of the ARC is gratefully acknowledged.

REFERENCES Alemanno, A. (2012), ‘Nudging smokers: the behavioural turn of tobacco risk regulation’, European Journal of Risk Regulation 3(1), 32‒42. Alemanno, A. and A. Garde (2015), ‘Regulating lifestyles: Europe, tobacco, alcohol and unhealthy diets’ in A. Alemanno and A. Garde (eds), Regulating Lifestyle Risks: The EU, Alcohol, Tobacco and Unhealthy Diets, Cambridge: Cambridge University Press, 1‒20. Alemanno, A. and A. Spina (2014), ‘Nudging legally: on the checks and balances of behavioural regulation’, International Journal of Constitutional Law 12(2), 429‒456. Ayo, N. (2012), ‘Understanding health promotion in a neoliberal climate and the making of health conscious citizens’, Critical Public Health 22(1), 99‒105. Bacchi, C. (2009), Analysing Policy: What’s The Problem Represented To Be?, Frenchs Forest: Pearson Australia. Baum, F. and M. Fisher (2014), ‘Why behavioural health promotion endures despite its failure to reduce health inequities’, Sociology of Health and Illness 36(2), 213‒225. Behavioural Insights Team (2010), Applying Behavioural Insight to Health, London: Crown Copyright. Behavioural Insights Team (2012), Test, Learn, Adapt: Developing Public Policy with Randomised Controlled Trials, London: Crown Copyright. Behavioural Insights Team (2013), Applying Behavioural Insights to Organ Donation, London: Crown Copyright. Behavioural Insights Team (2016a), The Team and Locations, www.behaviouralinsights.co.uk/people, accessed 3 November 2016. Behavioural Insights Team (2016b), Who We Are, www.behaviouralinsights.co.uk/about-us, accessed 15 August 2015. Berg, N. and G. Gigerenzer (2010), ‘As-if behavioural economics: neoclassical economics in disguise?’, History of Economic Ideas xviii(1), 133‒165. Bombak, A. (2014), ‘Obesity, health at every size, and public health policy’, American Journal of Public Health 104(2), e60‒e67. Bovens, L. (2008), ‘The ethics of nudge’ in T. Grüne-Yanoff and S.O. Hansson (eds), Preference Change: Approaches from Philosophy, Economics and Psychology, Dordrecht: Springer, 207‒219. Brownsword, R. (2008), Rights, Regulation, and the Technological Revolution, Oxford: Oxford University Press. Burgess, A. (2011), ‘Nudging healthy lifestyles: the UK experiments with the behavioural alternative to regulation and the market’, European Journal of Risk Regulation 3(1), 3‒16. Cahill, D., L. Edwards and F. Stilwell (eds) (2012), Neoliberalism: Beyond the Free Market, Cheltenham, UK and Northampton, MA: Edward Elgar Publishing. Cartwright, N. and J. Hardie (2012), Evidence-Based Policy: A Practical Guide to Doing It Better, Oxford: Oxford University Press. Casey, D. and C. Scott (2011), ‘The crystallisation of regulatory norms’, Journal of Law and Society 38(1), 76‒95. Commission on the Social Determinants of Health (2008), Closing the Gap in a Generation: Health Equity through Action on the Social Determinants of Health, Geneva: World Health Organization. Deaton, A. (2009), ‘Instruments of development: randomization in the tropics, and the search for the elusive keys to economic development’, National Bureau of Economic Research Working Paper No. 14690, 1‒54. Deaton A. and N. Cartwright (2016), ‘Understanding and misunderstanding randomized controlled trials’, National Bureau of Economic Research Working Paper No. 22595, 1‒67. della Porta, D. (2015), Social Movements in Times of Austerity: Bringing Capitalism Back into Protest Politics, Malden, MA: Polity Press. Dolan, P. et al. (2010), MINDSPACE: Influencing Behaviour Through Public Policy, London: UK Cabinet Office and Institute for Government. Dolowitz, D.P. and D. Marsh (2000), ‘Learning from abroad: the role of policy transfer in contemporary policy-making’, Governance 13(1), 5‒23. Driver, S. and L. Martell (2000), ‘Left, right and the third way’, Policy and Politics 28(2), 147‒161.

M4716-STRAßHEIM_9781785367847_t.indd 206

25/04/2019 13:32

The politics of nudge and framing behaviour change in health  207 European Commission (2015), Strengthening Evidence Based Policy Making Through Scientific Advice: Reviewing existing practice and setting up a European Science Advice Mechanism, Brussels, European Commission. https:// ec.europa.eu/research/sam/pdf/strengthening_evidence_based_policy_making.pdf, accessed 8 February 2018. European Commission (2018), Better Regulation https://ec.europa.eu/commission/priorities/democraticchange/better-regulation_en, accessed 8 February 2018. European Commission, Joint Research Centre (2016), Behavioural Insights Applied to Policy: European Report 2016, Brussels: European Union, http://publications.jrc.ec.europa.eu/repository/bitstream/JRC100146/kjna​ 27726enn_new.pdf, accessed 3 October 2016. Farrell, A.M. (2015), ‘Addressing organ shortage: are nudges the way forward?’, Law Innovation and Technology 7(2), 253‒282. Farrell, A.M. and M. Quigley (2012), ‘Organ donation and transplantation’, in R. Chadwick (ed.), Encyclopedia of Applied Ethics Vol. 3, 2nd edn, San Diego: Academic Press, 288‒296. Farrell, A.M., M. Quigley, and D. Price (eds) (2011), Organ Shortage: Ethics Law and Pragmatism, Cambridge: Cambridge University Press. Ferrer, R. L., I. Cruz, S. Burge, B. Bayles and M.I. Castilla (2014), ‘Measuring capability for healthy diet and physical activity’, Annals of Family Medicine 12(1), 46–56. Galizzi, M. and D. Wiesen (2017), ‘Behavioural experiments in health: an introduction’, Health Economics, 26(S3), 3‒5. Gilbert, D.T. et al. (2016), ‘Comment on “estimating the reproducibility of psychological science”’, Science 351(6277), 1037a‒1038a. Glass, T.A. and M.J. McAtee (2006), ‘Behavioural science at the crossroads in public health: extending horizons, envisioning the future’, Social Science and Medicine 62(7), 1650‒1671. Goffman, E. (1974), Frame Analysis, New York: Harper and Row. Gostin, L.O. (2000), ‘Public health law in the new century: Part 1: law as a tool to advance the community’s health’, Journal of the American Medical Association 283(21), 2837‒2841. Hastings, G. (2012), ‘Why corporate power is a public health priority’, British Medical Journal 345, e5124. Hood, C. (2010), ‘Accountability and transparency: Siamese twins, matching parts, awkward couple?’, West European Politics 33(5), 989‒1009. Howlett, M. and I. Mukherjee (2017), ‘Policy formulation: where knowledge meets power in the policy process’ in M. Howlett and I. Mukherjee (eds), Handbook of Policy Formulation, Cheltenham, UK and Northampton, MA: Edward Elgar Publishing, 3‒22. Independent Inquiry into Inequalities in Health (1998), Report of the Independent Inquiry into Inequalities in Health, London: The Stationery Office. John, P et al. (2013), Nudge, Nudge, Think, Think: Experimenting With Ways to Change Civic Behaviour, London: Bloomsbury Academic. Karpin, I. and K. O’Connell (2017), ‘Social determinants of health and the role of law’ in A.M. Farrell et al. (eds), Health Law: Frameworks and Context, Cambridge: Cambridge University Press, 34‒47. Lafont, C. (2015), ‘Deliberation, participation and democratic legitimacy: should deliberative mini-publics shape public policy?’, The Journal of Political Philosophy 23(1), 40‒63. Levi-Faur, D. (2013), ‘Regulation and regulatory governance’, in D. Levi-Faur (ed.), Handbook on The Politics of Regulation, Cheltenham, UK and Northampton, MA: Edward Elgar Publishing, 3‒24. MacKay, K. and M. Quigley (2018), ‘Exacerbating inequalities? Health policy and the behavioural sciences’, Health Care Analysis 26(4), 380–97. MacKay, S. (2011), ‘Legislative solutions to unhealthy eating and obesity in Australia’, Public Health 125(12), 896‒204. Marmot Review (2010), Fair Society, Healthy Lives: Strategic Review of Health Inequalities in England post-2010, www.instituteofhealthequity.org/projects/fair-society-healthy-lives-the-marmot-review, accessed 3 October 2016. McKinlay, J.B. and L.D. Marceau (2000), ‘To boldly go…’, American Journal of Public Health 90(1), 25‒33. Mimica, N. (2013), ‘Applying behavioural insights to policy-making: results, promises and limitations’, Behavioural Economics Conference/Brussels, 30 September 2013, Speech/13/761. Minkler, M. (1999), ‘Personal responsibility for health? A review of the arguments and the evidence at century’s end’, Health Education Quarterly 26(1), 121‒41. Murphy, P. and M. Smith (2012), ‘Towards a framework for organ donation in the UK’, British Journal of Anaesthesia 108(S1), i48‒i55. Open Science Collaboration (2015), ‘Estimating the reproducibility of psychological science’, Science 349(6251), aac4716-1‒4716-8. Palmer, M. (2012), Opt-out systems of organ donation: International evidence review, Number 44/2012, Welsh Government Social Research. Plimmer, G. (2014), ‘UK Cabinet Office ‘nudge’ team to be spun off into private group’, Financial Times, 4 February 2014.

M4716-STRAßHEIM_9781785367847_t.indd 207

25/04/2019 13:32

208  Handbook of behavioural change and public policy Powys White, K. et al. (2012), ‘Nudge, nudge or shove, shove: the right way for nudges to increase the supply of donated cadaver organs’, American Journal of Bioethics 12(2), 32‒39. Priaulx, N. and M. Weinel (2014), ‘Behaviour on a beer mat: law, interdisciplinarity, and expertise’, Journal of Law Technology and Policy 2, 361‒391. Quigley, M. (2013), ‘Nudging for health: on public policy and designing choice architecture’, Medical Law Review 21(4), 588‒621. Quigley, M. (2018), ‘Libertarian paternalism, nudging, and public policy’, in K. Grill and J. Hanna (eds), The Handbook of the Philosophy of Paternalism, Abingdon: Routledge, 223‒235. Quigley, M. and E. Stokes (2015), ‘Nudging and evidence-based policy in Europe: problems of normative legitimacy and effectiveness’ in A. Alemanno and A-L. Sibony (eds), Nudging and the Law: A European Perspective, Oxford: Hart Publishing, 61‒82. Rein, M. and D.A. Schön (1996), ‘Frame-critical policy analysis and frame-reflective policy practice’, Knowledge and Policy 9(1), 85‒104. Rithalia, A., C. McDaid, S. Suekarran et al. (2009), ‘Impact of presumed consent for organ donation on donation rates: a systematic review’, British Medical Journal 338, a3162. Rockloff, M. and C. Hanley (2014), ‘The default option: why a system of presumed consent may be effective at increasing rates of organ donation’, Psychology, Health and Medicine 19(5), 580‒585. Sabatier, P.A. (1998), ‘The advocacy coalition framework: revisions and relevance for Europe’, Journal of European Public Policy 5(1), 130‒198. Saunders, B. (2010), ‘Normative consent and opt-out organ donation’, Journal of Medical Ethics 36, 84‒87. Schön, D.A. and M. Rein (1994), Frame Reflection, New York: Basic Books. Slocock, C. (2015), Whose Society? The Final Big Society Audit, London: Civil Exchange, www.civilexchange. org.uk, accessed 9 March 2015. Smith, M. (2014), ‘A crisis of political parties’, in D. Richards, M. Smith and C. Hay (eds), Institutional Crisis in 21st Century Britain, Basingstoke: Palgrave Macmillan, 101‒124. Snow, DA., E. Burke Rochford, S.K. Worden and R.D. Benford (1986), ‘Frame alignment processes, micromobilization, and movement participation’, American Sociological Review 51(4), 464‒381. Thaler, R. and C. Sunstein (2009), Nudge: Improving Decisions About Health, Wealth, and Happiness, London: Penguin Books. Truog, R. (2012), ‘When does a nudge become a shove in seeking consent for organ donation?’, American Journal of Bioethics 12(2), 42‒44. Tversky, A. and D. Kahneman (1981), ‘The framing of decisions and the psychology of choice’, Science, 211(4881), 453‒458. United States Government, National Science and Technology Council, Social and Behavioural Sciences Team (2016), Approach, https://sbst.gov/approach, accessed 3 October 2016. Vallgårda, S. (2012), ‘Nudge: a new and better way to improve health?’, Health Policy 104(2), 200‒203. van Bavel, R. et al. (2013), Applying behavioural sciences to EU policy-making, JRC83284 EUR 26033 EN, Brussels: European Commission. https://ec.europa.eu/jrc/en/publication/eur-scientific-and-technical-researchreports/applying-behavioural-science-eu-policy-making, accessed 8 February 2018. van Hulst, M.J. and D. Yanow (2016), ‘From policy “frames” to “framing”: theorizing a more dynamic, political approach’, American Review of Public Administration 46(1), 92–112. White House, President Barack Obama (2015), Executive Order – Using behavioral science insights to better serve the American people. https://obamawhitehouse.archives.gov/the-press-office/2015/09/15/executive-orderusing-behavioral-science-insights-better-serve-american, accessed 8 February 2018. Wikler, D. (1987), ‘Who should be blamed for being sick?’, Health Education Quarterly 14(1), 11‒25. Willis, B. and M. Quigley (2014), ‘Opt-out organ donation: on evidence and public policy’, Journal of the Royal Society of Medicine 107(2), 56‒60. Woolf, S.H., M.M. Dekker, F.R. Byrne and W.D. Miller (2011), ‘Citizen-centered health promotion building collaborations to facilitate healthy living’, American Journal of Preventative Medicine 40(1S1), S38–S47. World Health Organization (2016), Social determinants of health, www.who.int/social_determinants/en, accessed 14 October 2016. Yeung, K. (2012), ‘Nudge as fudge’, Modern Law Review 75(1), 122‒148.

M4716-STRAßHEIM_9781785367847_t.indd 208

25/04/2019 13:32

15.  Nudging to sustainability? Critical reflections on nudging from a theoretically informed sustainability perspective Basil Bornemann and Paul Burger

1. INTRODUCTION ‘Nudges’ and ‘nudging’ have become very popular terms recently. The terms were originally introduced by Thaler and Sunstein to denote a presumably new steering approach that makes use of knowledge from the behavioural sciences to improve people’s individual decisions about ‘health, wealth and happiness’ (Thaler and Sunstein 2008). More specifically, nudges are concerned with directing the behaviour of individuals in such a way that they adopt a specific (socially or individually desired) behaviour without forcing or incentivizing them to do so. Rather than using coercive power in the form of sanctions, economic incentives or the power of arguments, nudging pertains to soft, subtle and virtually unnoticed interventions in individuals’ choice architectures (Barton and Grüne-Yanoff 2015; Schubert 2017). While nudges are already widely present in the daily economic life, for example, in the product placement of supermarkets (Chriss 2015), governments and public authorities have only recently begun to deliberately integrate nudges into their repertoire of steering approaches (Pykett et al. 2014). Governments worldwide have started to build up expert groups and specialized units for designing and implementing nudges as means of political steering (Straßheim and Korinek 2015; Whitehead et al. 2018, also Chapter 7 in this volume). It is around these more recent efforts by public authorities that a controversial discussion about the prospects and limitations of nudging has emerged (Bornemann and Smeddinck 2016). Within this controversial discussion, an increasing number of (sometimes very harsh) critical voices emphasize the manipulative potential and fundamentally question the legitimacy of governmental nudging. Given its focus on shaping decision factors that are located, at least partly, in the non-conscious mind (see below), nudging is regarded to infringe with normative principles of the democratic public sphere, such as ‘autonomy’ and ‘individual freedom’ (Binder and Lades 2015; Bovens 2009; Fateh-Moghadam and Gutmann 2014; Goodwin 2012; Grüne-Yanoff 2012; Hausman and Welch 2010; John et al. 2009; Selinger and Whyte 2011). The proponents, in contrast, stress the prospects and opportunities of ‘governing with nudges’. Nudging would not only expand the governmental toolbox, but represents a particularly effective and efficient steering approach that is based on scientific evidence about the actual barriers and drivers of individuals’ behaviour (Mills 2013; Sunstein 2014). What is more, scholars and practitioners from various fields, stress the sustainability potential of nudging (Sunstein and Reisch 2014; Thorun et al. 2017). Governing with nudges is said to come with particularly promising potential for tackling sustainability issues. For example, nudging may help to change 209

M4716-STRAßHEIM_9781785367847_t.indd 209

25/04/2019 13:32

210  Handbook of behavioural change and public policy individual energy consumption patterns or food choices towards more sustainable ones (Kasperbauer 2017; Lehner et al. 2015). In this chapter we set out to further explore the alleged sustainability potential of nudging. More specifically, we aim to clarify whether and in what sense nudging represents a potentially promising approach to sustainability governance, that is, a way of organizing collective action such that societal transformations towards sustainability are unleashed and propelled (Lange et al. 2013). We do so not in terms of an empirical analysis, but by means of critical theoretical reasoning. Based on a systematic notion of sustainability, we ask: What conceptual tensions and critical issues come into view when we look at nudging from a sustainability perspective? By answering this question, we add to both the critical discussion about nudging in general and the more specific discussion regarding nudging as a sustainability-oriented steering approach. More specifically, we clarify the potential role of nudging for sustainability governance and reveal possible tensions between nudging and sustainability governance that should be addressed in further research and practice. Our argument proceeds as follows. We begin with some conceptual clarifications and delimitations regarding the meaning of nudging (2). We proceed with a brief review of the existing discussion that links nudging and sustainability, and an overview of the arguments in favour and against ‘nudging sustainability’ that have been raised so far (3). In a third step, we adopt a more sophisticated sustainability perspective to critically reflect on theoretical tensions between nudging and different sustainability conceptions (4). We conclude by pointing to a number of issues and perspectives for practice and further research on nudging to sustainability (5).

2. NUDGING AND LIBERTARIAN PATERNALISM: DEFINITIONS AND DELIMITATIONS Following Thaler and Sunstein, a nudge ‘is any aspect of the choice architecture that alters people’s behaviour in a predictable way without forbidding any options or significantly changing their economic incentives. To count as a mere nudge, the intervention must be easy and cheap to avoid’ (Thaler and Sunstein 2008, p. 6). Rather than incentivizing, forcing or convincing people to move into a desired direction by means of ‘carrots, sticks or sermons’ (Bemelmans-Videc et al. 2007), nudges are concerned with editing individuals’ choice architectures in such a subtle way that they almost automatically adopt a certain behaviour (Schubert 2017). Unlike economic incentives (‘carrots’), nudges do not (at least not significantly) affect a person’s economic incentive structure or even change the preference order of individuals. Unlike prohibitions or commands (‘sticks’), nudges operate without sanctions or penalties to ensure a certain behaviour. How nudges differ from informational interventions (‘sermons’), however, is not as clear, but rather controversial in the debate. Thaler and Sunstein themselves do not seem to make this distinction in an explicit way, and classify at least some types of information (e.g. warnings and labels) as nudges. Other authors call for a clear distinction between on the one hand ‘classical’ informational steering approaches addressing the conscious mind, and more subconscious nudges on the other (Barton and Grüne-Yanoff 2015; Ölander and Thøgersen 2014; Schnellenbach 2012; Schubert 2017; Selinger and Whyte 2011).1 In this chapter, we follow this latter position and treat nudges as a kind of intervention that differs from

M4716-STRAßHEIM_9781785367847_t.indd 210

25/04/2019 13:32

Nudging to sustainability?   211 factual information or even rational reasoning by addressing and using certain factors in the choice architecture of individuals that prevent people from acting rationally. These include cognitive boundaries, biases, routines or habits (Hansen 2016). This narrower understanding, which deviates from Thaler and Sunstein’s broader original concept, also seems to be consistent with the wider epistemic background of nudging. An important reference point in the discussion about nudging is scientific evidence on systematic rationality gaps in the behaviour of individuals, that is, systematic cognitive limitations of the human mind (Tversky and Kahneman 1974). According to these insights, people would behave far less rationally than suggested by the long-time dominant model of rational choice rendering the latter rather a normative approach on how humans ought to decide than a description of their actual behaviour. In fact, people hardly ever possess the kind of complete information, unlimited cognitive abilities, and perfect self-control to make best (or even second best) decisions as suggested by rational choice models (an idea that has already been articulated under the classical concept of ‘bounded rationality’). What is more, there is evidence that people normally behave within a decision environment that transgresses this rational mode of ‘system-2-thinking’ altogether. In situations, where fast and prompt decisions are required (which is regularly the case), people operate in a so-called ‘system-1-thinking’-mode (for a differentiated review of the literature see Evans 2003; Evans 2008; Kahneman 2013). This mode of thinking is dominated by factors of the decision environment that simplify complex decisions, including heuristics (e.g. rules of thumb), biases (e.g. status quo biases, following the herd) and emotions (e.g. relish, disgust). Drawing on these insights from behavioural science, nudging is frequently presented as a steering approach associated with system-1-thinking (Barton and Grüne-Yanoff 2015; Michalek et al. 2016). More specifically, nudges are interventions into those factors constituting the choice architecture of system-1-thinking, such as human biases, heuristics and emotions. They are concerned with taking account of and ‘working with’ these factors, by actively triggering, suppressing or overcoming them (Binder and Lades 2015; Hansen 2016). Referring to these various types of factors, different forms of nudges are being distinguished in the academic and political discussion. Among the most prominent are the following ones (Reisch and Sandrini 2015; Schubert 2017; Yoeli et al. 2017): 1. Default nudges make use of individuals’ reluctance to actively engage and decide by modifying the choice architecture in such a way that the ‘good option’ is the pre-defined standard choice. A prominent example of a default nudge towards more environmental benign behaviour is to set the default option of energy choice in energy supply contracts to ‘green energy’ (Chassot et al. 2013; Pichert and Katsikopoulos 2008; Sunstein and Reisch 2014). 2. Framing nudges refer to attempts to present and communicate choices so that people are cognitively or emotionally attracted to a particular option. For example, framing compensation for CO2-emissions from flying as ‘carbon offset’ attracts more people to take action than a ‘carbon tax’ frame (Hardisty et al. 2010). Also, specifying the total expected cost savings per year instead of the mere energy efficiency class increases the likelihood that people will opt for the efficient option (Camilleri and Larrick 2014). 3. Similarly, salience nudges serve to raise awareness of certain options over others.

M4716-STRAßHEIM_9781785367847_t.indd 211

25/04/2019 13:32

212  Handbook of behavioural change and public policy Ordering food choices in canteens such that the healthy option is displayed before the unhealthy one is the example par excellence. Using arrows or other means that lead people to staircases instead of elevators, or vegetables and salads instead of unhealthy food are other prominent examples (Payne et al. 2016; Thaler and Sunstein 2008). 4. Social norm nudges make use of people’s tendency to compare with or even to imitate the behaviour of their peer fellows (‘follow the herd’). The classic example of social norm nudges are energy reports that relate people’s own energy consumption to the (average) consumption behaviour of their neighbourhood (Allcott and Rogers 2014; Schultz et al. 2007) (although the authors do not refer to ‘nudges’). Apart from this or similar ways of distinguishing nudges by how they work, another typology relevant to our topic is based on the question of who the supposed beneficiary of nudging is (Barton and Grüne-Yanoff 2015; Hagman et al. 2015). Along this line, two types of nudges can be distinguished. On the one hand, there are ‘pro-self-nudges’ that are meant to steer people’s behaviour ‘in a private welfare-promoting direction’ (Barton and Grüne-Yanoff 2015, p. 344), that is, to increase individuals’ personal utility. Among this type are, for example, attempts to nudge people towards healthier eating patterns by modifying the order of food choices in a canteen or by making use of deterring pictures on cigarette boxes. On the other hand, there are nudges that are directed at enhancing societal welfare and realizing some form of public good. A classic example is an organ donor system in which people are nudged into by the default setting that everybody’s organs are automatically donated after death while, at the same time, allowing them to actively ‘opt out’ during their lifetime.2 From the outset of the discussion, ‘nudging’ has been tightly connected to the concept of ‘libertarian paternalism’. According to Thaler and Sunstein (2003), libertarian paternalism is concerned with supporting individuals to make ‘good’ decisions as judged by themselves, that is, decisions that are well-aligned with their own preferences (e.g. avoid drug consumption, living in a healthy way), and, thereby, also contribute to aggregated societal welfare. The paternalistic component is the belief that organizations are able to recognize people’s preferences and to design interventions in their choice architectures so that they can realize their preferences. The libertarian component lies in the presumption that people should not be forced into making certain decisions. This is mainly expressed in the call for ‘opt-out-clauses’, which allow the nudgee to easily withdraw from a decision environment that is being nudged (Hansen 2016). Both proponents and critics have viewed nudging and libertarian paternalism as intrinsically interlinked, and even used the two concepts interchangeably. However, we follow those who argue that they should be kept apart (Barton and Grüne-Yanoff 2015; Bornemann and Smeddinck 2016). Nudges can best be understood as a particular class of steering instruments, that is, as a certain kind of means for shaping the behaviour of individual actors through intervention in and modification of factors of their choice architectures. In contrast, libertarian paternalism represents a normative political philosophical rationale for justifying public or private organizations’ steering interventions in people’s behaviour and decisions. Whatever the rationale behind libertarian paternalism is, we follow the proposition to distinguish between nudging as an instrumental means to achieve a certain end (e.g. societal welfare), and libertarian paternalism as a specific

M4716-STRAßHEIM_9781785367847_t.indd 212

25/04/2019 13:32

Nudging to sustainability?   213 normative rationale to justify interventions into people’s lives for realizing that end. This distinction does not only allow for separating the discussion about nudging from the particular controversial normative implications of libertarian paternalism, but also opens the possibility to discuss nudging as an instrument with respect to other normative ideas such as sustainability.

3. NUDGING AND SUSTAINABILITY: EXPLORING THE DEBATE ‘Sustainability’ has been quite present in the nudging debate from the outset. In their seminal book, Thaler and Sunstein themselves advocate nudging as a highly promising tool for ‘saving the planet’ by ‘greening’ consumption and enabling more climate-friendly lifestyles (Thaler and Sunstein 2008, pp. 183–196). In addition, they introduce nudging as a tool to support people’s private savings behaviours with the aim to relieve the looming problems in the social security system (Thaler and Sunstein 2008, pp.  103–117). Since these early references, the link between sustainability and nudging has been further elaborated on, mainly by Cass Sunstein himself (Sunstein 2015) and in collaboration with other authors (Hedlin and Sunstein 2016; Sunstein and Reisch 2013, 2014; Thorun et al. 2017). These contributions have added examples and illustrations of ‘green nudges’ and clarified the conditions under which particular nudges, such as ‘green defaults’ or ‘smart metering’, might successfully shape individuals’ behaviour patterns towards more sustainable ones (Sunstein and Reisch 2014). Furthermore, there are reflections on how ‘choice architects’ can select among various design options of certain types of nudges, such as default settings (Sunstein and Reisch 2014, pp. 155–157). Apart from these references to sustainability by the inventors of the approach, nudging has been considered a promising approach of steering towards sustainability by a broader community of sustainability-oriented scholars from various disciplines (Heidbrink 2015; Lehner et al. 2015; Michalek et al. 2016; Ölander and Thøgersen 2014; Schubert 2017). The typical field of interest are patterns of (Western) consumption behaviour that are thought to be responsible for the unsustainable pathway of the global human society. Given the pervasiveness and persistency of these patterns, which have been fruitlessly addressed by other means of political steering, nudging has been welcomed as a promising new approach to sustainabilize individual consumption (Schubert 2017). Accordingly, many examples and related research (not always explicitly referring to ‘nudging’) can be found in areas such as household energy use (Allcott 2011; Kasperbauer 2017; Momsen and Stoerk 2014; Pichert and Katsikopoulos 2008; Yoeli et al. 2017), personal transport (Dogan et al. 2014), food consumption (Just and Wansink 2009; Mueller et al. 2017; Wansink and Hanks 2014; Wansink and Just 2015), shopping behaviour (Demarque et al. 2015), travel/tourism (Hall 2013), workplace behaviour (Michaels and Powell 2017), recycling behaviour or (household) waste production (Lehner et al. 2015; Milford et al. 2015; Rivers et al. 2017). Within these fields, empirical research has tested, mainly in small-scale lab and field experiments, how specific types of nudges, such as default settings, ordering patterns, or social norms work as well as how efficient and effective they are. Furthermore, there are some overall assessments of the comparative effectiveness of nudges across different fields. Based on their review of sustainability-oriented nudges

M4716-STRAßHEIM_9781785367847_t.indd 213

25/04/2019 13:32

214  Handbook of behavioural change and public policy in different fields, Lehner et al. (2015), for example, conclude that those types of nudges seem to be most effective, which are based on changes in the physical environment or default options. In comparison, the use of social norms and frames tends to be a less successful strategy. However, they also emphasize that context matters both on the side of the nudgers (in terms of available resources) and the nudgees (in terms of social and cultural environments), which implies a limited generalizability of the evidence regarding particular nudge designs (Loewenstein and Chater 2017). The debate regarding ‘nudging sustainability’ is not confined to academic circles. As more recent references to nudging and ‘choice editing’ in several policy documents show, the notion that nudging may contribute to sustainable transformations is not merely a theoretical construct, but has attracted the attention of policy practitioners in multiple contexts (European Commission 2016; German Environment Agency [GEA] 2017; Jones et al. 2014; Reisch and Sandrini 2015; for further references see also Schubert 2017 Fn. 8; Thorun et al. 2017; Whitehead et al. 2018; Yoeli et al. 2017). Given these developments, nudging will most certainly become more relevant in sustainability governance. Yet, what speaks for and what against nudging as a suitable steering approach to sustainability? In the following, we summarize the main arguments present in the existing debate that, directly or indirectly, support or reject nudging as steering approach to sustainability. 3.1  Arguments in Favour of ‘Nudging Sustainability’ A first main argument in favour of nudging as a steering approach towards sustainability refers to its supposed capacity to ‘crack’ routinized behaviour patterns and induce behavioural change in situations where people otherwise are not able to change their behaviour. More specifically, nudging is thought to support people in adopting the kind of sustainability-oriented behaviour they would like to adopt, but fail to realize (Heidbrink 2015). Hence, nudging addresses the famous attitude-behaviour gap that underlies many sustainability-related problems (Kollmuss and Agyeman 2002): According to repeated surveys on sustainable behaviour, many people dispose of sustainability-oriented knowledge, values and attitudes, but fail to act in accordance with these predispositions due to hindering factors, such as laziness or working routines. Nudging promises to allow ‘people [who] actually want to act sustainably, but have a limited ability of putting their willingness into practice’ (Heidbrink 2015, p. 173) to overcome their behavioural constraints. It enables people to act according to their own (sustainability-related) insights and attitudes, and thereby increases their own benefits as well as the aggregated societal welfare. Second, apart from furthering the goal attainment of those who actually subscribe to sustainability goals, nudging is also regarded as a promising tool to reach out to those people who don’t care for sustainability and who have turned out to be resistant to information campaigns and other ways of trying to raise their awareness for sustainability (Lehner et al. 2015). Precisely because it draws on factors such as heuristics or biases, nudging might be an effective approach to unleash broad societal transformations within the majority of disinterested. Accordingly, nudging comes with a new potential of scaling up and mainstreaming sustainability-oriented behaviour patterns (Heidbrink 2015; Rettie et al. 2014). That this potential should also be used is sometimes justified (albeit implicitly) by references to the alleged universality of nudging. According to this claim, basically every decision takes place in some choice architecture and is necessarily nudged in one

M4716-STRAßHEIM_9781785367847_t.indd 214

25/04/2019 13:32

Nudging to sustainability?   215 or the other way (Sugden 2009).3 If all choices are inevitably nudged however, there are supposedly good reasons to nudge these choices in the direction of commonly accepted ‘better’ ends, such as sustainability (Kasperbauer 2017; Momsen and Stoerk 2014). Third, one of the high hopes placed in nudging is that it comes with efficiency gains and increased decision capacities both on the side of the nudger and the nudgee (Thaler and Sunstein 2008). On the side of the nudger, nudging is a cheap way of intervention since it does neither involve the costs for economic incentives nor the costs of executing sanctions. On the nudgee’s side, nudging is regarded as way out of the complexity trap in face of information overload, and a necessary condition for maintaining agency (Sunstein 2014). It relieves individuals from reflecting on rather small and unimportant choices (such as the choice of healthy food in a canteen), and thereby frees their capacities for the kinds of decisions and actions that really matter, such as engaging in collective action oriented towards societal transformations. From this perspective, nudging is not a direct contribution to, but rather a necessary condition for sustainable transformations. 3.2  Arguments Against ‘Nudging Sustainability’ As compared to the somewhat rather explicit pro arguments, there are interestingly no direct objections against nudging as an approach to further sustainable development. As indicated above, the dominant nudging critique rather refers to values such as individual freedom or autonomy. However, there is some questioning as to whether nudging has the potential to effectively address the comprehensive, complex and long-term nature of grand societal challenges (Burgess 2012; Marteau et al. 2011; Selinger and Whyte 2012) strongly motivating the societal discourse on sustainable development. A first objection refers to the underlying framing of the problem in question in terms of consumerism. Even though there can be no doubts that modern consumerism is a major precondition of unsustainable development, some critics argue that nudging overemphasizes the role of consumption vis-à-vis production. Moreover, nudging as a class of steering instruments promotes ‘individualization’ of consumerism and its governance (Hall 2013; Schubert 2017). It represents a specific strategy of subjectivation of responsibility. Rather than problematizing and addressing structural conditions of unsustainable consumerism, such as infrastructures, collective norms and values, or power relations, nudging exclusively focuses on the responsibility of individuals. From a nudging perspective, it is individual consumers who are responsible for the (sustainability) problems through their behaviour. Consequently, it is basically only their behaviour that needs to be changed in order to solve these problems. According to the critics, such reasoning obscures that grand challenges and complex sustainability issues have also structural and not only individual causes (Brooks and Bryant 2014; Maniates 2014; Whitmarsh et al. 2013). A second related argument looks upon nudging as a means of reproducing and stabilizing existing consumption practices instead of changing them. Since consumption as such is not being challenged, but merely altered, nudging helps to maintain, rather than to challenge the established business models of the corporate world (Brooks and Bryant 2014). From this perspective, nudging is a strategy to sustain the unsustainable status quo, a manifestation of the dominant politics of unsustainability (Blühdorn 2007). Due to its focus on realizing given preferences, nudging implies that no changes of these preferences are needed to foster a sustainable transformation (Goodwin 2012). The conservative

M4716-STRAßHEIM_9781785367847_t.indd 215

25/04/2019 13:32

216  Handbook of behavioural change and public policy rather than transformative ambition is particularly expressed in nudging techniques that take advantage of human biases in favour of a given status quo (Sugden 2009). In a similar vein, a third strand of criticism points to the potential anti-transformative side effects of nudging. Critics argue that the extensive use of nudging would not only reproduce existing consumerism (see above), but paves the way towards a society that has entirely unlearned to actively transform itself. In particular, it is argued that nudging would impair on the learning capacity of modern societies (Selinger and Whyte 2011, 2012). Arguing that nudges systematically discourage agents from engaging in active choice and assuming that active choice is not only an instrumental means to come up with good results, but is embedded in broader individual and social processes of learning (from success and failure!) and identity creation (from interaction!), the use of nudges would in the long run, infringe on the fundamental self-esteem of modern societies (White 2013; Whitehead et al. 2018). A nudging state may thereby contribute to alienating individuals from their role as active participants in politics, that is, their role as citizens (see also Leone and Tallacchini, Chapter 11 in this volume). In that sense, nudges may have not only an alienating, but also a depoliticizing effect (Schubert 2017). A nudged society would not only lose its capacity of self-improvement through the making of mistakes; it would also lose its capacity to organize, engage in and carry out political deliberation.

4. NUDGING FROM A THEORETICALLY INFORMED SUSTAINABILITY PERSPECTIVE In the preceding section, we summarized arguments made in the present debate in favour and against ‘nudging sustainability’. While for some ‘sustainability’ serves as an important reference frame for arguing in favour of nudging, critical voices doubt that nudges are suitable means for realizing sustainability. Both positions, however, hardly build on the more sophisticated and elaborate understandings of sustainability that have emerged in the theoretical sustainability discourse over the last 30 years. Reflections on ‘green nudges’, for example, tend to focus on whether and how nudges effectively contribute to reducing the environmental impact of consumption behaviour, and sometimes also consider the legitimacy of nudging (Michalek et al. 2016; Schubert 2017). However, given this focus on (ecological) effectiveness and legitimacy, they tend to overlook the normative and evaluative implications of a much broader sustainability concept. We argue that these very implications need to be taken into account when advocating for or against nudging as a sustainability-fitting steering approach. This is why we now step back and proceed with a reflection on nudging from a qualified sustainability perspective. Any attempt to take a qualified position on sustainability faces the challenge that sustainability thinking has brought about an almost infinite number of conceptions. Joining other attempts for establishing a meta-order (Dobson 1996; Kates et al. 2005), we suggest distinguishing between three groups of conceptions: justice-based, resiliencebased and procedural conceptions of sustainable development, each of them coming with a specific understanding of ‘sustainable’ (Burger 2018). The first group encompasses conceptions that are explicitly based on ethical/moral investments, such as notions of wellbeing and intra- and intergenerational justice, respectively. Consequently, ‘sustainability’ designates a moral idea with intrinsic values focusing on achieving wellbeing for

M4716-STRAßHEIM_9781785367847_t.indd 216

25/04/2019 13:32

Nudging to sustainability?   217 all living and future generations (Burger and Christen 2011; Dobson 1998). The second group includes conceptions that define ‘sustainable’ based on the maintenance of life supporting system functions, such as global, regional or local ecosystems. Following this ecological perspective, sustainability is frequently identified with resilience and vulnerability (Berkes et al. 2006; Gallopín 2006; Robèrt 2000): Sustainable is whatever remains within the safe operating space of ecosystems (Rockström et al. 2009). Whereas justice-based and resilience-oriented conceptions of sustainability presume that sustainability can be defined in substantial terms, scholars in the third stream adopt a procedural understanding and conceive of ‘sustainability’ in terms of a moving target and essentially political concept, whose meaning is to be defined (and constantly re-defined) in societal discourse (Jacobs 1999; Norton 2005). In light of this diversity of conceptions, we suppose that there is not a straight-forward relationship between nudging and sustainability. Rather, nudging could be compatible with different sustainability conceptions in different ways, and, vice versa, different sustainability conceptions may point to different critical issues of nudging. In the following analysis, the link between sustainability conceptions and nudging shall be further established. As indicated in the introduction, we do not aim to assess whether nudging has the potential to achieve the intended goals – that would be an empirical endeavour well beyond the scope of this chapter. Rather, we take a conceptual critical point of view and focus on possible tensions between nudging as a class of specific steering instruments and the basic goals of the three types of sustainability conceptions. 4.1  Nudging and Justice-based Sustainability Looking at specific examples, such as default settings for a green energy option in energy contracts, we do not have fundamental doubts that nudging can be a useful approach to foster behaviour patterns that contribute to the realization of the kind of normative goals, such as inter- and intra-generational justice, advocated by proponents of a justice-based sustainability conception (Hoberg and Strunz 2018). However, there are at least three possible areas of tension. A first major concern from the perspective of justice-based sustainability is related to the consumption bias of nudging. The nudging approach presupposes a world of consumption and choice opportunities. However, access to this world is unevenly distributed since only the relatively privileged dispose of the resources to take part in consumption. Therefore, only the privileged might become ‘objects’ of nudging – and benefit from it. Nudging has no bearing at all when it comes to providing access to the world of consumption, for example, by redistributing resources from those who have choices to those who haven’t. It can accordingly even contribute to the reproduction of existing (unjust) resource distribution rather than overcoming it. Moreover, nudging comes with unequal redistributive effects for those who can take part in the world of consumption. For example, pro-social-nudges towards environmentally friendlier but costlier choices (such as the green energy default) may further disadvantage the already underprivileged people, given their limited resource basis for realizing these choices. Opt-out clauses may come with additional ‘hidden’ redistributive effects. Again, underprivileged people, who may not have the skills (information, time) to perceive and make use of opt-out clauses, may be disadvantaged by being nudged into costlier choices. In contrast, those who are better off anyway and have greater abilities and capacities

M4716-STRAßHEIM_9781785367847_t.indd 217

25/04/2019 13:32

218  Handbook of behavioural change and public policy to recognize and use opt-out clauses, could free-ride and further improve their relative position vis-à-vis the underprivileged (Goodwin 2012; Lehner et al. 2015; for empirical indications see Rivers et al. 2017; Schubert 2017). Second, those who subscribe to the capability approach (CA) as an appropriate metric of wellbeing and justice (Burger and Christen 2011; Nussbaum 2006; Sen 1980, 1999) will immediately recognize a tension between the CA and nudging. While the former is about ensuring and increasing individual freedoms, the latter cuts them down in certain respects. Even if one acknowledges opt-out clauses as freedom-ensuring mechanisms within nudges, the whole approach is about limiting choice options in the first place (Grüne-Yanoff 2012). At least, nudging does not contribute to enhance future capacities and valued choice options as implied by the CA. Furthermore, there are reasonable concerns that nudging infringes on established capabilities by, for example, taking away the chance of learning from mistakes and approaching some form of rational behaviour through constant improvement (Rizzo and Whitman 2009). Third, since nudging is about realizing pre-defined goals of (individual or social) wellbeing, the approach puts a blind eye on the justice-related consequences coming with different approaches to these goals. In the case of so-called pro-self-nudges, the goals of nudging are defined by (relatively) stable individual preferences. Since nudging takes these ‘given’ preferences for granted, it comes with a focus on effectiveness and efficiency, leaving considerations of sufficiency (which would imply a reconsideration of preferences) aside. More importantly, there are tensions between justice-oriented sustainability conceptions and pro-social-nudges, that is, those nudges that are directed to the societal goals expressed by ‘sustainable’. Given that in today’s mainstream policy-making wellbeing is still operationalized by traditional welfare-functions and expressed in GDP, pro-social-nudges risk to be directed towards goals standing in conflict with requirements stemming from intra- and intergenerational justice considerations (Stiglitz et al. 2010). 4.2  Nudging and Resilience-based Sustainability Looking at the many examples of nudges in sustainability-relevant fields, such as energy or food consumption, we do not have fundamental doubts that nudges principally can help influencing societal practices in such ways that their impacts remain within the carrying capacities of socio-ecological systems. However, we again see three potentially critical tensions. First, given its focus on individual (micro-)choices in daily consumption practices, it is unclear whether nudging may be sufficient to achieve the kind of largescale societal transformations that are needed to stay within certain socio-ecological boundaries. At least, there seem to be reasonable doubts as to whether nudging really is a promising strategy for dealing with ‘big impact issues’ such as mobility, land-use, biodiversity losses, or the impacts stemming from agricultural production. This concern is aggravated by opt-out clauses, which are constitutive for nudges (at least in the liberal paternalist sense). If many individuals opt out and choose not to be nudged by certain interventions, nudging turns out to be highly ineffective in maintaining defined societal or ecological boundaries. Second, in addition to this concern about the size of the lever (which might depend on contexts and problem areas), there is a more fundamental tension. It relates to the question of whether and how nudging can ensure that the aggregated impacts of the

M4716-STRAßHEIM_9781785367847_t.indd 218

25/04/2019 13:32

Nudging to sustainability?   219 (altered) behaviour of many people actually remains within certain socio-ecological boundaries (however defined and specified). At least there seems to be a contradiction between pro-self-nudges and resilience-oriented sustainability approaches. Although based on fundamental doubts about the empirical validity of rational choice assumptions, pro-self-nudges are specifically designed to help people overcome their rationality deficits and enable them to better realize their individual preferences.4 However, as is known from classics of resilience-oriented sustainability thinking, the collective effects of individually rational behaviour can turn out to be highly irrational (Dolšak and Ostrom 2003, Ostrom 1990). Therefore, pro-self-nudges bear the risk of increasing the rationality of individual decisions at the expense of unsustainable collective outcomes; attempting to nudge people into more rational patterns of behaviour could increase the tendencies to exploit and overuse public goods. Nudges can therefore aggravate the tragedy of the commons, which from a resilience-oriented perspective is at the centre of many sustainability problems. A third concern from a resilience point of view relates to the systemic side effects of nudges. Being designed as interventions into very specific behaviour patterns, such as food choices, nudges come with the risk of ‘problem-shifting’ that is, ameliorating one particular problem by amplifying another one. Such a pattern is frequently described as moral licensing and has been observed, for example, in a study on the interference between water and energy consumption (Tiefenbeck et al. 2013). Another example for systemic side effects is the well-known rebound effect that refers to increased consumption in a similar or another field as a result of efficiency gains (for example, lower CO2 emissions per travel distance, but larger distances travelled). It would be unfair to attribute moral licensing or the rebound effect to nudging alone, but it is fair to assume that the risks of such side effects by highly specific and isolated interventions of the nudging type need to be considered. 4.3  Nudging and Procedural Sustainability Again, we have no concerns about the potential of nudges to contribute to the achievement of sustainability goals resulting from the kind of social deliberation process promoted by advocates of a procedural sustainability conception. However, there are also tensions between nudging and such a conception. First, as already indicated above, nudging may come with conservative tendencies. Nudges are tailored towards realizing existing beliefs and values, or, more precisely, the kinds of individual or social welfare functions, policymakers deem representative for their steering objects, be it individuals or the society as a whole. Being oriented towards realizing some form of predefined individual preferences or societal goal, nudging seems to lack precisely the kind of transformative impetus that characterizes procedural sustainability conceptions. Instead of changing preferences and values, nudging could help to reproduce the very norms and values that have contributed to the emergence of sustainability problems. Second and corresponding to an already somehow established line of argumentation against nudging sustainability (see above), the extensive use of nudging risks the ability of a society to deliberate and learn. If individuals are constantly relieved of making meaningful decisions in their daily lives, they and ultimately society as a whole may lose the ability to make sound and socially robust judgments about complex societal issues. To hollow out these capacities, in turn, means to undermine precisely those social conditions

M4716-STRAßHEIM_9781785367847_t.indd 219

25/04/2019 13:32

220  Handbook of behavioural change and public policy that are considered prerequisites for procedural conceptions of sustainability. From the point of view of procedural sustainability, deliberation and social learning are in fact important in at least two respects. Because sustainability issues are characterized by great uncertainty and ambiguity (Voss et al. 2007), deliberation and learning are important on the one hand to accommodate different interests, values, and knowledge sets, and to find acceptable trade-offs between burdens and benefits. On the other hand, the ability to think and learn is a prerequisite for defining and constantly adapting socially robust and meaningful contextual understandings of sustainability, thus dealing with the concept’s inherent contestedness as such (Jacobs 1999). Obviously, these concerns about deliberation and learning touch upon the above-mentioned well-established criticisms of nudging regarding autonomy and individual freedom. As the proceduralists also relate the capacity to govern sustainable transformations to the realization of these very democratic principles, the asserted violation of these principles by nudging also undermines the capacity of sustainability governance (Dobson 2007). This leads to a third concern about nudging from a procedural sustainability perspective. Some advocates of nudging seem to suggest that it can overcome or, at least, bypass (conflicting) value judgments regarding appropriate ends of political steering by fully relying on scientific evidence not only about the explanation of behaviour (the choice architecture), but also regarding the ‘real’ preferences of individuals (for critical remarks, see Sugden 2009). This is, however, a rather problematic assumption especially when it comes to sustainability informed public choices (see for similar concerns Hoberg and Strunz 2018). On the one hand, and apart from reasonable doubts that experts can dispose of meaningful and comprehensive empirical knowledge regarding individual preferences (White 2017), there are good reasons that political steering of whatever kind inevitably involves normative issues. Questions regarding appropriate goals and means of steering cannot be answered on the basis of scientific evidence, but require ethical considerations and value judgments for which scientific experts cannot claim specific expertise and exclusive authority. Sustainability issues, in particular, are characterized by complex ethical questions, value conflicts, and trade-offs that cannot be simply translated, or even dissolved into, scientific evidence. On the other hand, the central role of scientific expertise for nudging points in the direction of technocratic elitism. It is the experts with the related scientific knowledge who decide where to go, and how. Aside from the argument just made that scientific experts have no exclusive expertise for value judgments, the emphasis on scientific experts contradicts an important normative element of procedural sustainability: the plea for participation in order to consider the values, interest and knowledge of different kinds of actors, such as stakeholders, citizens or local communities (Meadowcroft 2004).

5.  CONCLUSIONS AND OUTLOOK Starting from the observation that the nudging discussion refers to ‘sustainability’ as both an important concept of justification and relevant field of application, this chapter seeks to shed light on the conceptual relationship between nudging and sustainability, and thus attain a better understanding of the potential role of nudging in sustainability governance. After mapping existing arguments in favour and against ‘nudging sustainability’,

M4716-STRAßHEIM_9781785367847_t.indd 220

25/04/2019 13:32

Nudging to sustainability?   221 we reflected on nudging from a theoretically informed sustainability perspective to find out about tensions between nudging and sustainability conceptions. It turned out that the relationship between nudging and sustainability is neither straight-forward nor an easy one. Nudging cannot be expected to promote sustainability per se; rather, the relationship between nudging and sustainability is filled with tensions that vary between different sustainability conceptions. While justice-based conceptions of sustainability point to critical issues related to equity, resilience-based conceptions seem to be at odds with the potential leverage of nudging as well as with side effects; finally, procedural conceptions of sustainability lead to tensions related to nudging’s technocratic and conservationist tendencies. What are the consequences of these tensions both for the discussion on nudging and for the discussion on sustainability governance? First and foremost, we pointed to possible tensions. The presence of tensions does not mean that nudging is fundamentally incompatible with sustainability, nor is there a reason to exclude nudging from sustainability governance. Second, the analysis enriches the debate by going beyond an implicit equation between ‘sustainability’ and ‘greening’. By discussing nudging against the background of a theoretically sound understanding of sustainability, the analysis points to several potential negative side effects of nudging. Together with existing critiques focusing on values such as autonomy, freedom and democracy, our sustainability-oriented arguments underscore the proposition that nudging does not only come with benefits, but also societal risks. These risks need to be carefully taken into consideration and dealt with (either by deliberately accepting, reducing, compensating or actively working with them) when nudges are applied in practice. Third, the analysis certainly has made clear that nudging is not the only and all-curing remedy for sustainability problems as the current nudging hype sometimes seems to prompt. Rather, nudges might play their parts as instruments in more complex sustainability governance arrangements. This is in line with, on the one hand, existing views that regard nudges as complements to, rather than substitutes for, traditional incentive or rule-based measures (Lehner et al. 2015; Michalek et al. 2016; Schubert 2017), a view that is also shared by the main advocates of green nudging (Sunstein 2015). On the other hand, the view that nudges should be considered components of more complex governance settings is consistent with existing conceptualizations of sustainability governance that highlight the relevance of integrated policy mixes addressing different factors of societal transformations (Bornemann 2014; Voss et al. 2007). Sustainability-oriented policy mixes can benefit from nudges as long as there are other instruments that deal with and potentially outbalance the possible tensions between nudges and sustainability. Nudges can, for example, help to break up existing behaviour patterns while their tendency of hollowing out deliberation is countered by meaningful efforts to organize public deliberation and participation. In this vein, future design and practice of sustainability governance should take account of behavioural insights in governance design, rather than try to entirely design sustainability governance modes towards behavioural insights. Accordingly, to provide a basis for future governance design efforts, further theoretical and empirical research will have to scrutinize the potential and actual role of nudging in normatively ambitious and empirically complex governance settings. In addition to investing informed understandings of sustainability, such research needs to be based on informed understandings of ‘governance’ as well. Further research could focus on questions like: How can sustainability governance arrangements integrate nudges in

M4716-STRAßHEIM_9781785367847_t.indd 221

25/04/2019 13:32

222  Handbook of behavioural change and public policy such ways that their potential for promoting behaviour change is improved while their tensions and negative impacts are outbalanced? Addressing this issue will also need to further examine the institutional and power-related conditions and consequences of sustainability-oriented governance arrangements that integrate nudges.

ACKNOWLEDGEMENTS We are grateful for the comments and recommendations of the editors and two anonymous reviewers. They helped us a lot to improve the chapter. The research is part of the activities of SCCER CREST, which is financially supported by the Swiss Commission for Technology and Innovation (CTI).

NOTES 1. In fact, as Barton and Grüne-Yanoff (2015, p. 345) argue, the current controversial debate is not about information since informing is being generally considered as ethically unproblematic. The bulk of the literature has rather concentrated on ‘heuristics-triggering nudges’, which are considered to be ethically problematic. 2. Schubert distinguishes between individual-oriented paternalistic and society-oriented non-paternalistic nudges (Schubert 2017). 3. ‘In many cases, some kind of nudge is inevitable, and so it is pointless to ask government simply to stand aside. Choice architects, whether private or public, must do something’ (Thaler and Sunstein 2008, p. 237 emphasis in original). This argument, however, overlooks the conceptual difference between choice architecture and intervention into that architecture, that is, the nudge proper (see Barton and Grüne-Yanoff 2015; Schubert 2017, p. 336). 4. Accordingly, they are about improving the rationality of decisions by influencing non-rational decision factors.

REFERENCES Allcott, H. (2011), ‘Social Norms and Energy Conservation’, Journal of Public Economics, 95(9‒10), 1082–95. Allcott, H. and T. Rogers (2014), ‘The Short-Run and Long-Run Effects of Behavioral Interventions: Experimental Evidence from Energy Conservation’, The American Economic Review, 104(10), 3003–3037. Barton, A. and T. Grüne-Yanoff (2015), ‘From Libertarian Paternalism to Nudging—and Beyond’, Review of Philosophy and Psychology, 6(3), 341–359. Bemelmans-Videc, M.-L., R.C. Rist and E. Vedung (eds) (2007), Carrots, Sticks and Sermons: Policy Instruments and their Evaluation, New Brunswick, NJ: Transaction Publishers. Berkes, F., J. Colding and C. Folke (eds) (2006), Navigating Social-Ecological Systems. Building Resilience for Complexity and Change, Cambridge: Cambridge University Press. Binder, M. and L.K. Lades (2015), ‘Autonomy-Enhancing Paternalism’, Kyklos, 68(1), 3–27. Blühdorn, I. (2007), ‘Sustaining the Unsustainable: Symbolic Politics and the Politics of Simulation’, Environmental Politics, 16(2), 251–275. Bornemann, B. (2014), Policy-Integration und Nachhaltigkeit: Integrative Politik in der Nachhaltigkeitsstrategie der deutschen Bundesregierung, Wiesbaden: Springer VS. Bornemann, B. and U. Smeddinck (2016), ‘Anstößiges Anstoßen? – Kritische Beobachtungen zur “Nudging”Diskussion im deutschen Kontext’, ZParl Zeitschrift für Parlamentsfragen, 47(2), 437–459. Bovens, L. (2009), ‘The Ethics of Nudge’, in T. Grüne-Yanoff and S. Hansson (eds), Preference Change: Approaches from Philosophy, Economics and Psychology, Dordrecht: Springer, pp. 207–219. Brooks, A. and R. Bryant (2014), ‘Consumption’, in C. Death (ed.), Critical Environmental Politics, London, New York: Routledge, Taylor & Francis Group, pp. 72–82. Burger, P. (2018), ‘Sustainability, Sustainability Assessment, and the Place of Fiscal Sustainability’, in D. Malito,

M4716-STRAßHEIM_9781785367847_t.indd 222

25/04/2019 13:32

Nudging to sustainability?   223 G. Umbach and N. Bhuta (eds), The Palgrave Handbook of Indicators in Global Governance, London: Palgrave Macmillan, pp. 139‒159. Burger, P. and M. Christen (2011), ‘Towards a Capability Approach of Sustainability’, Journal of Cleaner Production, 19(8), 787–795. Burgess, A. (2012), ‘“Nudging” Healthy Lifestyles: The UK Experiments with the Behavioural Alternative to Regulation and the Market’, European Journal of Risk Regulation, 3(1), 3–16. Camilleri, A.R. and R.P. Larrick (2014), ‘Metric and Scale Design as Choice Architecture Tools’, Journal of Public Policy & Marketing, 33(1), 108–125. Chassot, S., R. Wüstenhagen, N. Fahr and P. Graf (2013), ‘Wenn das grüne Produkt zum Standard wird: Wie ein Energieversorger seinen Kunden die Verhaltensänderung einfach macht’, Organisationsentwicklung(3), 80–87. Chriss, J.J. (2015), ‘Nudging and Social Marketing’, Society, 52(1), 54–61. Demarque, C., L. Charalambides, D.J. Hilton and L. Waroquier (2015), ‘Nudging Sustainable Consumption: The Use of Descriptive Norms to Promote a Minority Behavior in a Realistic Online Shopping Environment’, Journal of Environmental Psychology, 43, 166–174. Dobson, A. (1996), ‘Environment Sustainabilities: An Analysis and a Typology’, Environmental Politics, 5(3), 401–428. Dobson, A. (1998), Justice and the Environment. Conceptions of Environmental Sustainability and Theories of Distributive Justice, Oxford: Oxford University Press. Dobson, A. (2007), ‘Environmental Citizenship: Towards Sustainable Development’, Sustainable Development, 15(5), 276–285. Dogan, E., J.W. Bolderdijk and L. Steg (2014), ‘Making Small Numbers Count: Environmental and Financial Feedback in Promoting Eco-driving Behaviours’, Journal of Consumer Policy, 37(3), 413–422. Dolšak, N. and E. Ostrom (eds) (2003), The Commons in the New Millennium.: Challenges and Adaptation, Cambridge, MA: MIT Press. European Commission (2016), ‘Behavioural Insights Applied to Policy: European Report 2016’, European Commission; Joint Research Centre. Evans, J.S. (2003), ‘In Two Minds: Dual-Process Accounts of Reasoning’, Trends in Cognitive Sciences, 7(10), 454–459. Evans, J.S.B.T. (2008), ‘Dual-Processing Accounts of Reasoning, Judgment, and Social Cognition’, Annual Review of Psychology, 59, 255–278. Fateh-Moghadam, B. and T. Gutmann (2014), ‘Governing [through] Autonomy. The Moral and Legal Limits of “Soft Paternalism”’, Ethical Theory and Moral Practice, 17(3), 383–397. Gallopín, G.C. (2006), ‘Linkages between Vulnerability, Resilience, and Adaptive Capacity’, Global Environmental Change, 16(3), 293–303. German Environment Agency (2017), ‘Shaping Environmental Policy in a Citizen-Friendly Manner: How Behavioural Science Findings Can Make Environmental-Policy Instruments More Effective’, accessed 10 December 2017 at www.umweltbundesamt.de/publikationen. Goodwin, T. (2012), ‘Why We Should Reject “Nudge”’, Politics, 32(2), 85–92. Grüne-Yanoff, T. (2012), ‘Old Wine in New Casks: Libertarian Paternalism Still Violates Liberal Principles’, Social Choice and Welfare, 38(4), 635–645. Hagman, W., D. Andersson, D. Västfjäll and G. Tinghög (2015), ‘Public Views on Policies Involving Nudges’, Review of Philosophy and Psychology, 6(3), 439–453. Hall, C.M. (2013), ‘Framing Behavioural Approaches to Understanding and Governing Sustainable Tourism Consumption: Beyond Neoliberalism, “Nudging” and “Green Growth”?’, Journal of Sustainable Tourism, 21(7), 1091–1109. Hansen, P.G. (2016), ‘The Definition of Nudge and Libertarian Paternalism: Does the Hand Fit the Glove?’, European Journal of Risk Regulation, 7(1), 1–20. Hardisty, D.J., E.J. Johnson and E.U. Weber (2010), ‘A Dirty Word or a Dirty World: Attribute Framing, Political Affiliation, and Query Theory’, Psychological Science, 21(1), 86–92. Hausman, D.M. and B. Welch (2010), ‘Debate: To Nudge or Not to Nudge’, Journal of Political Philosophy, 18(1), 123–136. Hedlin, S. and C.R. Sunstein (2016), ‘Does Active Choosing Promote Green Energy Use? Experimental Evidence’, Ecology Law Journal, 43(1), 107–142. Heidbrink, L. (2015), ‘Libertarian Paternalism, Sustainable Self-Binding and Bounded Freedom’, in D. Birnbacher and M. Thorseth (eds), The Politics of Sustainability: Philosophical Perspectives, London, New York: Routledge, pp. 173–194. Hoberg, N. and S. Strunz (2018), ‘When Individual Preferences Defy Sustainability — Can Merit Good Arguments Close the Gap?’, Ecological Economics, 143, 286–293. Jacobs, M. (1999), ‘Sustainable Development as a Contested Concept’, in A. Dobson (ed.), Fairness and Futurity. Essays on Environmental Sustainability and Social Justice, Oxford: Oxford University Press, pp. 21–45.

M4716-STRAßHEIM_9781785367847_t.indd 223

25/04/2019 13:32

224  Handbook of behavioural change and public policy John, P., G. Smith and G. Stoker (2009), ‘Nudge Nudge, Think Think: Two Strategies for Changing Civic Behaviour’, Political Quarterly, 80(3), 361–370. Jones, R., J. Pykett and M. Whitehead (2014), Changing Behaviours: On the Rise of the Psychological State, Cheltenham, UK and Northampton, MA: Edward Elgar Publishing. Just, D.R. and B. Wansink (2009), ‘Smarter Lunchrooms: Using Behavioral Economics to Improve Meal Selection’, Choices, 24(3), n.p. Kahneman, D. (2013), Thinking, Fast and Slow, London: Penguin. Kasperbauer, T.J. (2017), ‘The Permissibility of Nudging for Sustainable Energy Consumption’, Energy Policy, 111, 52–57. Kates, R.W., T.M. Parris and A.A. Leiserowitz (2005), ‘What is Sustainable Development? Goals, Indicators, Values, and Practice’, Environment: Science and Policy for Sustainable Development, 47(3), 8–21. Kollmuss, A. and J. Agyeman (2002), ‘Mind the Gap: Why do People Act Environmentally and What are the Barriers to Pro-environmental Behavior?’, Environmental Education Research, 8(3), 239–260. Lange, P., P.P.J. Driessen, A. Sauer, B. Bornemann and P. Burger (2013), ‘Governing Towards Sustainability— Conceptualizing Modes of Governance’, Journal of Environmental Policy & Planning, 15(3), 403–425. Lehner, M., O. Mont and E. Heiskanen (2015), ‘Nudging – A Promising Tool for Sustainable Consumption Behaviour?’, Journal of Cleaner Production, 134, 166‒177. Loewenstein, G. and N. Chater (2017), ‘Putting Nudges in Perspective’, Behavioural Public Policy, 1(1), 26–53. Maniates, M. (2014), ‘Sustainable Consumption – Three Paradoxes’, GAIA – Ecological Perspectives for Science and Society, 23(3), 201–208. Marteau, T.M., D. Ogilvie, M. Roland, M. Suhrcke and M.P. Kelly (2011), ‘Judging Nudging: Can Nudging Improve Population Health?’, British Medical Journal, 342, 263–265. Meadowcroft, J. (2004), ‘Participation and Sustainable Development: Modes of Citizen, Community and Organisational Involvement’, in W.M. Lafferty (ed.), Governance for Sustainable Development: The Challenge of Adapting Form to Function, Cheltenham, UK and Northampton, MA: Edward Elgar Publishing, pp. 162–190. Michaels, E. and M. Powell (2017), ‘Behavioural Economics: Using ‘Nudges’ for Promoting Pro-environmental Behaviours in the Workplace’, in P. Baranova, E. Conway, N. Lynch and F. Paterson (eds), The Low Carbon Economy: Understanding and Supporting a Sustainable Transition, Cham: Springer International Publishing, pp. 59–85. Michalek, G., G. Meran, R. Schwarze and Ö. Yildiz (2016), ‘Nudging as a New “Soft” Tool in Environmental Policy – An Analysis Based on Insights From Cognitive and Social Psychology’, Zeitschrift für Umweltpolitik und Umweltrecht(2‒3). Milford, A.B., A. Øvrum and H. Helgesen (2015), ‘Nudges to Increase Recycling and Reduce Waste’, Norwegian Agricultural Economics Research Institute, accessed 14 October 2016 at www.nilf.no/publikas​ joner/Discussion_Papers/2015/dp-2015-01.pdf. Mills, C. (2013), ‘Why Nudges Matter: A Reply to Goodwin’, Politics, 33(1), 28–36. Momsen, K. and T. Stoerk (2014), ‘From Intention to Action: Can Nudges Help Consumers to Choose Renewable Energy?’, Energy Policy, 74, 376–382. Mueller, M.P., S. Anzman-Frasca, C.E. Blakeley, S.C. Folta, P. Wilde and C.D. Economos (2017), ‘Ordering Patterns Following the Implementation of a Healthier Children’s Restaurant Menu: A Latent Class Analysis’, Obesity (Silver Spring, Md.), 25(1), 192–199. Norton, B.G. (2005), Sustainability: A Philosophy of Adaptive Ecosystem Management, Chicago: University of Chicago Press. Nussbaum, M.C. (2006), Frontiers of Justice: Disability, Nationality, Species Membership (The Tanner lectures on human values), Boston, MA: Harvard University Press. Ölander, F. and J. Thøgersen (2014), ‘Informing Versus Nudging in Environmental Policy’, Journal of Consumer Policy, 37(3), 341–356. Ostrom, E. (1990), Governing the Commons: The Evolution of Institutions for Collective Action, Cambridge: Cambridge University Press. Payne, C.R., M. Niculescu, D.R. Just and M.P. Kelly (2016), ‘This Way to Produce: Strategic Use of Arrows on Grocery Floors Facilitate Produce Spending Without Increasing Shopper Budgets’, Journal of Nutrition Education and Behavior, 48(7), 512‒513.e1. Pichert, D. and K.V. Katsikopoulos (2008), ‘Green Defaults: Information Presentation and Pro-Environmental Behaviour’, Journal of Environmental Psychology, 28(1), 63–73. Pykett, J., R. Jones, M. Welsh and M. Whitehead (2014), ‘The Art of Choosing and the Politics of Social Marketing’, Policy Studies, 35(2), 97–114. Reisch, L.A. and J. Sandrini (2015), Nudging in der Verbraucherpolitik: Ansätze verhaltensbasierter Regulierung, Baden-Baden: Nomos. Rettie, R., K. Burchell and C. Barnham (2014), ‘Social Normalisation: Using Marketing to Make Green Normal’, Journal of Consumer Behaviour, 13(1), 9–17.

M4716-STRAßHEIM_9781785367847_t.indd 224

25/04/2019 13:32

Nudging to sustainability?   225 Rivers, N., S. Shenstone-Harris and N. Young (2017), ‘Using Nudges to Reduce Waste? The Case of Toronto’s Plastic Bag Levy’, Journal of Environmental Management, 188, 153–162. Rizzo, M.J. and D.G. Whitman (2009), ‘Little Brother is Watching You: New Paternalism on the Slippery Slopes’, Arizona Law Review, 51, 685–739. Robèrt, K.-H. (2000), ‘Tools and Concepts for Sustainable Development, How Do They Relate to a General Framework for Sustainable Development, and to Each Other?’, Journal of Cleaner Production, 8(3), 243–254. Rockström, J., W. Steffen, K. Noone, A. Persson, F.S. Chapin, E.F. Lambin, T.M. Lenton, M. Scheffer, C. Folke, H.J. Schellnhuber, B. Nykvist, C.A. de Wit, T. Hughes, S. van der Leeuw, H. Rodhe, S. Sorlin, P.K. Snyder, R. Costanza, U. Svedin, M. Falkenmark, L. Karlberg, R.W. Corell, V.J. Fabry, J. Hansen, B. Walker, D. Liverman, K. Richardson, P. Crutzen and J.A. Foley (2009), ‘A Safe Operating Space for Humanity’, Nature, 461(7263), 472–475. Schnellenbach, J. (2012), ‘Nudges and Norms: On the Political Economy of Soft Paternalism’, European Journal of Political Economy, 28(2), 266–277. Schubert, C. (2017), ‘Green Nudges: Do They Work? Are They Ethical?’, Ecological Economics, 132, 329–342. Schultz, P.W., J.M. Nolan, R.B. Cialdini, N.J. Goldstein and V. Griskevicius (2007), ‘The Constructive, Destructive, and Reconstructive Power of Social Norms’, Psychological Science, 18(5), 429–434. Selinger, E. and K. Whyte (2011), ‘Is There a Right Way to Nudge? The Practice and Ethics of Choice Architecture’, Sociology Compass, 5(10), 923–935. Selinger, E. and K. Whyte (2012), ‘Nudging Cannot Solve Complex Policy Problems’, European Journal of Risk Regulation, 3(1), 26–31. Sen, A. (1980), ‘Equality of What’, in S. McMurrin (ed.), The Tanner Lectures on Human Values, Cambridge: Cambridge University Press, pp. 197–220. Sen, A. (1999), Development as Freedom, New York: Oxford University Press. Stiglitz, J.E., A. Sen and J.-P. Fitoussi (2010), Mismeasuring Our Lives: Why GDP Doesn’t Add Up, New York: New Press. Straßheim, H. and R.-L. Korinek (2015), ‘Behavioural Governance in Europe’, in J. Wilsdon and R. Doubleday (eds), Future Directions for Scientific Advice in Europe, pp. 155–162. Sugden, R. (2009), ‘On Nudging: A Review of Nudge: Improving Decisions About Health, Wealth and Happiness by Richard H. Thaler and Cass R. Sunstein’, International Journal of the Economics of Business, 16(3), 365–373. Sunstein, C.R. (2014), Why Nudge: The Politics of Libertarian Paternalism, New Haven: Yale University Press. Sunstein, C.R. (2015), ‘Behavioural Economics, Consumption, and Environmental Protection’, in L. Reisch and J. Thøgersen (eds), Handbook of Research on Sustainable Consumption, Cheltenham, UK and Northampton, MA: Edward Elgar Publishing, pp. 313–327. Sunstein, C.R. and L.A. Reisch (2013), ‘Green by Default’, Kyklos, 66(3), 398–402. Sunstein, C.R. and L.A. Reisch (2014), ‘Automatically Green: Behavioral Economics and Environmental Protection’, Harvard Environmental Law Review, 38(1), 127‒158. Thaler, R.H. and C.R. Sunstein (2003), ‘Libertarian Paternalism’, The American Economic Review, 93(2), 175–179. Thaler, R.H. and C.R. Sunstein (2008), Nudge: Improving Decisions about Health, Wealth and Happiness, New Haven: Yale University Press. Thorun, C., J. Diels, M. Vetter, L. Reisch, M. Bernauer, H.-W. Micklitz, K. Purnhagen, J. Rosenow and D. Forster (2017), ‘Nudge-Ansätze beim nachhaltigen Konsum: Ermittlung und Entwicklung von Maßnahmen zum “Anstoßen” nachhaltiger Konsummuster: Mit freundlicher Unterstützung durch Prof. Dr. Cass Sunstein’, accessed 10 December 2017 at www.umweltbundesamt.de/publikationen. Tiefenbeck, V., T. Staake, K. Roth and O. Sachs (2013), ‘For Better or for Worse? Empirical Evidence of Moral Licensing in a Behavioral Energy Conservation Campaign’, Energy Policy, 57, 160–171. Tversky, A. and D. Kahneman (1974), ‘Judgment under Uncertainty: Heuristics and Biases’, Science, 185(4157), 1124–1131. Voss, J.-P., J. Newig, B. Kastens, J. Monstadt and B. Nölting (2007), ‘Steering for Sustainable Development: A Typology of Problems and Strategies with Respect to Ambivalence, Uncertainty and Distributed Power’, Journal of Environmental Policy & Planning, 9(3), 193–212. Wansink, B. and A.S. Hanks (2014), ‘Calorie Reductions and Within-meal Calorie Compensation in Children’s Meal Combos’, Obesity (Silver Spring, Md.), 22(3), 630–632. Wansink, B. and D.R. Just (2015), ‘Trayless Cafeterias Lead Diners to Take Less Salad and Relatively More Dessert’, Public Health Nutrition, 18(9), 1535–1536. White, M.D. (2013), The Manipulation of Choice, New York: Palgrave Macmillan. White, M.D. (2017), ‘Nudging Merit Goods: Conceptual, Normative, and Practical Connections’, Forum for Social Economics, 25, 1–16. Whitehead, M., R. Jones, R. Lilley, J. Pykett and R. Howell (2018), Neuroliberalism: Behavioural Government in the Twenty-first Century, London, New York: Routledge, Taylor & Francis Group.

M4716-STRAßHEIM_9781785367847_t.indd 225

25/04/2019 13:32

226  Handbook of behavioural change and public policy Whitmarsh, L., S. O’Neill and I. Lorenzoni (2013), ‘Public Engagement with Climate Change: What do We Know and Where Do We Go from Here?’, International Journal of Media & Cultural Politics, 9(1), 7–25. Yoeli, E., D.V. Budescu, A.R. Carrico, M.A. Delmas, DeShazo J.R., P.J. Ferraro, H.A. Forster, H. Kunreuther, R.P. Larrick, M. Lubell, E.M. Markowitz, B. Tonn, M.P. Vandenbergh and E.U. Weber (2017), ‘Behavioral Science Tools to Strengthen Energy and Environmental Policy’, Behavioral Science and Policy, 3(1), 69–79.

M4716-STRAßHEIM_9781785367847_t.indd 226

25/04/2019 13:32

16.  Encouraging longer working lives: a behavioural perspective R. Kent Weaver

Populations in almost all advanced industrial countries – and in most developing countries as well – are living much longer than in the past. One of the results of this demographic trend is a dramatic increase in the percentage of the population that is aged 65 and above, the age that many countries long set as their standard (and in some cases mandatory) retirement age, though many countries have raised their retirement ages in recent years (OECD 2011, Chapter 1). Moreover, both past and already-legislated future increases in the minimum and standard state pension age have not kept up with increases in life expectancy in most countries that are part of the Organisation for Economic Co-operation and Development (OECD). Life expectancy for those reaching pensionable age averaged 18.5 years for men in 2010 and is expected to reach 20.3 years by 2050; the expected increase for women is slightly less dramatic because pensionable ages for women are rising faster (from a lower base): from 23.3 years in 2010 to 24.6 years by 2050. Thus despite important increases in retirement age policy in many countries, the burden of pension expenditures in most OECD countries is expected to increase unless benefits are reduced or people are required to work longer to receive pensions equivalent to those of today’s retirees. An aging population poses many challenges for governments in both industrial and developing countries in addition to higher costs in public pensions (see for example Barr 2006), notably increased health care and long-term-care costs, plus higher risk of many elderly persons outliving their assets, and lowered labour force participation by family caregivers caring for elderly parents or in-laws. Changing the behaviour of older workers by increasing their labour market participation is one way to address both affordability concerns of retirement income systems and the income needs of those seniors (Munnell and Sass 2006; for a behavioural perspective on young unemployed see Tosun and Hörisch, Chapter 18 in this volume). But if ­workers – or at least a substantial percentage of them – are not naturally inclined to labour, or have significant barriers to work, this is not a shift that is likely to be enthusiastically embraced by most older workers. Actual trends in labour force participation of older workers have not been linear, or uniform across countries. Both the average age at which workers exit the labour market and the level of labour force participation of older workers have changed substantially over time. Labour force participation rates of older (50‒64) aged men generally fell from around 1970 through around 2000 and then stabilized and (in some countries) increased slightly since then. Labour force participation rates for older women have steadily risen in most countries (Burtless 2013; OECD 2011, Chapter 2; Maestas and Zissimopolous 2010; on the United States, see Rix 2008). Both age of labour market exit and labour force participation rates of older workers vary substantially across OECD countries, however, with lower participation rates generally found in Southern Europe. Moreover, retirement 227

M4716-STRAßHEIM_9781785367847_t.indd 227

25/04/2019 13:32

228  Handbook of behavioural change and public policy is no longer an either/or proposition or a one-way street, with some workers now engaging in partial retirement, retiring and later returning to work on a full or part-time basis, or combining (usually reduced time) work and pensions (Maestas 2010; Fisher et al. 2016). Research clearly suggests that government policies on matters such as the ‘pensionable age’ for receiving state pensions without actuarial reduction for early retirement, as well as the availability of programmes that allow or encourage pension receipt prior to the standard retirement age, and whether or not mandatory retirement ages exist, exercise an important influence on labour market exit decisions, though they are by no means the only drivers of retirement behaviour. In the United States, for example, the Full Retirement Age (FRA) at which ‘standard’ Social Security retirement benefits are received was raised from 65 to 66 for cohorts born after 1938 in two-month increments every year for the next six years, with a consequently higher ‘discount’ on benefits received prior to that age (the earliest claiming age remained at 62). This change led to a modest increase in the age of claiming, with a spike at the time of the new Full (or ‘standard’) Retirement Age for each cohort (see Johnson et al. 2013; Song and Manchester 2007). A substantial and varied set of literatures now addresses the determinants of labour market exit by older workers from a variety of disciplinary perspectives (for reviews, see for example Vickerstaff 2010, Fisher et al. 2016; Wang and Shi 2014). The contribution of this chapter is to look at labour market exit from a multi-disciplinary behavioural perspective, that is to place multiple determinants of the behaviour of a very heterogeneous set of actors – older workers – at the centre of analysis. Government policies that seek to extend working lives are not ‘self-implementing’: they require changes in behaviour by workers, non-workers (who potentially can re-enter the labour force) and employers. This premise is obvious in itself, but it suggests that efforts to understand and reshape labour market decisions to extend working lives can be organized around four basic questions about behaviour change. First, what are the major barriers to delaying labour force exit across extremely heterogeneous populations of older workers in OECD countries? Second, what are the major options for addressing those barriers? Third, what are the major advantages and limitations of each option with respect to achieving the primary goal of extending working lives, as well as the additional goals of being fiscally sustainable for governments, minimizing unwanted disruptions to the lives of the elderly and near elderly, and creating a policy regime that has a reasonable probability of adoption and successful implementation and retention? Fourth, and more normatively, are there cross-national lessons that policy-makers can learn from, adapt and successfully implement and sustain from other wealthy countries? This chapter explores these questions, making three basic arguments. First, as noted above, labour force exit decisions are influenced by a variety of constraints/barriers and facilitating conditions. The focus here is on older workers who have not yet left the labour force. They may not currently be in employment due to voluntary or involuntary job loss, but are seeking work if they are not employed. Any individual actor is likely to be affected by multiple barriers to extending their working lives, and the immense heterogeneity of the target population of older workers means that patterns of ‘binding’ constraints – those that exert the greatest impact on labour market decisions – will vary across individuals and groups of individuals, notably those with different work and earnings histories, skill levels, and health conditions. Second, governmental efforts to extend working lives can use multiple instruments, with more or less intensive settings on each instrument (Schneider and

M4716-STRAßHEIM_9781785367847_t.indd 228

25/04/2019 13:32

Encouraging longer working lives  229 Ingram 1990; Weaver 2015). But simplicity and equity concerns mean that governments need to treat heterogeneous individuals according to a relatively simple set of rules that may be better-suited to changing the behaviour of some subsets of the target population than others. Third, policy responses are constrained by politics and public opinion: more intrusive policy instruments and settings on those instruments that impose greater costs on the target population face substantial opposition. The political power and positive ‘social construction’ of the elderly (that is, how they are viewed by the broad public; see Schneider and Ingram 1993) plus governments’ reluctance to impose visible losses on a politically popular constituency have generally led governments to avoid intrusive instruments that send the clearest signals or requirements to work longer. The structure of the chapter follows these questions. The first section briefly lays out the analytical framework. The second section analyses multiple barriers to extending working lives from the perspective of older workers. The third section briefly addresses strategic options for addressing those barriers, drawing on the experiences of policy choices made by governments in several OECD countries and identifying some promising practices to extend working lives. The conclusion summarizes the causal analysis and addresses potential extensions of the analysis.

1.  ANALYTICAL FRAMEWORK Factors such as political culture, societal wealth and attitudes towards the target population determine the formal policies and rules that may or may not attempt to change the behaviour of older workers. These formal policies and rules, which can be called the ‘compliance regime’, may vary in their intrusiveness and intensity, as will be discussed in the following two sections. Factors such as bureaucratic capacity influence how those policies are implemented (the ‘enforcement regime’; see Weaver 2014, for a fuller discussion of compliance and enforcement regimes). Attributes of individual workers – for example, the level of information and financial resources that they have, and peer effects among their social networks and family – are a third set of influences on individual labour market exit behaviour. Shortages of various types of resources as well as control exerted by other actors can interfere with the capacity of target populations to change their behaviour, even if they view it as in their own self-interest to do so. Labour market conditions act as a fourth constraint on individual labour market behaviour. Individual labour market exit behaviour in turn determines aggregate labour market exit rates. Feedback effects from aggregate labour market exit behaviour may in turn affect compliance regimes (e.g. by increasing the perceived severity of pension funding crises) and individual level factors that influence labour market exit. Government compliance and enforcement regimes can employ several different mechanisms that influence labour market exit behaviour in ways that either promote or discourage it. These include: ●●

Providing information on matters such as the prevalence of extended working lives, workers’ rights in resisting age discrimination, and so on. In addition to simply providing information, they may also provide mentoring over an extended period on how to put that information into practice.

M4716-STRAßHEIM_9781785367847_t.indd 229

25/04/2019 13:32

230  Handbook of behavioural change and public policy ●● ●● ●●

●● ●●

●●

Framing that casts working longer in a positive or negative light and admonitions to remain in or exit the labour market. Providing resources that help workers extend their working lives, such as job retraining and referral services. Manipulating choice architecture and defaults to encourage or discourage labour market exit without actually changing payoffs – for example, automatic payment of pensions at a particular age unless a worker opts not to receive them. Providing positive incentives for behaviour change, for example, increasing pension payoffs for later retirement. Imposing sanctions (negative incentives) to discourage certain behaviours – for example raising the age at which ‘full’ retirement benefits are received and increasing penalties for early retirement. Prohibiting or requiring specific behaviours, with penalties for non-compliant behaviour.

While some of these strategic options are natural ‘matches’ with particular barriers (providing information to address information barriers, for example), multiple instruments can be used to address most barriers. Moreover, governments can adjust the ‘settings’ on these instruments – for example, by providing stronger admonitions, more resources, or stronger incentives for behaviours that governments seek. And in some cases, governments may be able to identify sub-populations that are critical to achieving government’s overall behaviour change objectives and develop mechanisms that are especially tailored to achieve behaviour change in those groups, even if they apply to the entire population of older workers.

2.  BARRIERS TO EXTENDED WORKING LIVES Barriers to extended working lives can be seen in each of the categories outlined above. 2.1 Incentives Workers considering whether to leave the labour force or extend their labour market participation face a complex calculus. The most obvious incentive to extending working life for individual workers is the opportunity to increase current income (and perhaps retirement savings) and/or future pension entitlements. Clearly this will not be equally available to all older workers, as will be discussed below. Persons who have good physical health and cognitive capacity are more likely to have this option available to them than those who do not. More generally, the opportunity cost of foregoing consumption of pension and retirement savings benefits and leisure may work against extending working lives. This is likely to be particularly true for persons with low life expectancy. And the nature of the work/leisure trade-off also varies with the structure of benefit entitlement: systems that do not actuarially compensate for retirement beyond a standard retirement age and require labour force exit to receive benefits, for example, create an incentive to retire to avoid losing benefits. Other aspects of current social insurance arrangements may also affect the calcula-

M4716-STRAßHEIM_9781785367847_t.indd 230

25/04/2019 13:32

Encouraging longer working lives  231 tions by older workers. In particular, if older workers can qualify for disability insurance payments prior to the standard retirement age without lowering their ultimate retirement pension benefits, they have an increased incentive to seek those benefits and leave the labour force. Finally, in the United States, where workers and employers are responsible for providing health insurance prior to age 65 (the normal age of eligibility for Medicare), the high costs of paying for pre-retirement health insurance may act as a strong deterrent for labour exit prior to that age, especially for individuals who currently have employersubsidized health insurance. Other incentives may also increase the probability of early labour market exit: a European study, for example, found that becoming a grandparent increases the probability of retirement, though the effect is statistically significant only for women (van Bavel and De Winter 2013). The effect may operate both by increasing the rewards of family oriented leisure and by increasing requests for assistance in child care provision. 2.2  Monitoring and Enforcement Monitoring and enforcement, which are major issues in many areas of behaviour change, are probably a less critical issue in the case of extended working lives because, in most wealthy countries at least, payroll and income taxation mechanisms and payment procedures for public pensions are generally well-developed. The primary concern relates to workers who begin collecting pensions and work in the informal sector, avoiding payment of taxes that might accrue from their higher overall (earnings plus pensions) income. Little research appears to exist on how serious this problem is. 2.3  Informational and Cognitive Barriers Existing literature points to a number of information and cognitive barriers that may affect labour market exit decisions. The ‘rational actor’ perspective assumes that actors possess all the information they need to make decisions, that they respond rationally (i.e. in their perceived self-interest) to information and incentives, and that they have preferences that are consistent over time. But as a huge literature on bounded rationality has shown, individuals are likely to make sub-optimal decisions when they need to make decisions quickly and when information requirements and acquisition costs are high. Thus in weighing policy alternatives, people often ‘satisfice’ rather than continuing to search for optimal outcomes, act on impulse, discount risks (e.g. the risk of extreme longevity), procrastinate, and are fatalistic. Facing a high degree of information complexity and high uncertainty may lead workers to use simplifying decision-making heuristics. They may also be myopic: discounting future costs and benefits severely, or failing to draw connections between current actions and future consequences. Clearly many of these concepts apply to retirement decisions. Three factors in particular are especially relevant. The first is a high degree of information complexity and uncertainty. Many economic developments in recent years have increased the level of uncertainty in decisions regarding retirement and contributed to feelings of personal economic insecurity (see Feldman and Beehr 2011). Defined benefit pensions in the private sector have declined substantially over time in many countries; instead, workers are increasingly accumulating funds in defined contribution pension systems with an

M4716-STRAßHEIM_9781785367847_t.indd 231

25/04/2019 13:32

232  Handbook of behavioural change and public policy uncertain valuation over time, and even less certainty over what they will produce as a retirement income stream. Fluctuations in housing values have led to increased uncertainty over the value of what for many households is their largest asset and retirement ‘nest egg’. And cutbacks in public pensions in many countries and concerns about the solvency of others (notably the Social Security Program in the United States) have led many older workers to fear that public pension promises may not be kept. It is not clear, however, that these pressures will cause workers to extend their working lives; they may instead take reduced pensions as soon as they become eligible in hopes of ‘locking in’ those benefits. A second cognitive and informational barrier to extending working lives is the existence of ‘standard’ or ‘full’ retirement ages from which benefit reductions or increases for earlier or later benefit claiming are normally calculated. While these reference points are essentially arbitrary in systems where actuarial adjustments are made for earlier or later benefit claiming, the US data noted earlier on spikes in claiming of retirement benefits at the standard retirement age suggests that they do have an impact on behaviour. The importance of the ‘Full Retirement Age’ may occur through several cognitive mechanisms. ‘Standard’ retirement ages may simply serve as an ‘anchoring’ point that sets expectations on retirement. In addition, Behaghel and Blau (2012, p. 42) and others have noted that they may result from ‘an endorsement effect. Workers might take this “official” designation as implicit advice from the government about when to retire and claim benefits’, especially when gathering information independently about the merits of various claiming ages is difficult and costly. This ‘endorsement effect’ is likely to be especially strong among workers with limited information and cognitive skills. To avoid this problem, governments can emphasize ‘flexible’ rather than ‘standard’ retirement ages, as Sweden has done. But this may cause workers with little knowledge to assume that they can retire at the earliest eligibility age without any benefit penalty. Another cognitive bias created by the standard retirement age reference point, given the tendency towards loss aversion noted by prospect theory, is that it may both highlight for workers the trade-off between work and leisure, and the potential risk of not maximizing total benefit claiming if one dies relatively soon after retirement. If, on the other hand, a worker views Social Security benefits primarily as a form of longevity insurance for a very long life expectancy, then later retirement with higher monthly benefits is likely to be optimal (see Behaghel and Blau 2012; Alleva 2015). A recent British study found that extended working life ‘is framed as a loss, in terms of leisure time and active life years, which is not outweighed by financial gains from enhanced income’ (Weyman et al. 2012, p. 75). Of course, most workers do not make these calculations in a highly organized fashion: a 2008 study by the Employee Benefits Research Institute in the United States found that 22 per cent of retirees began serious retirement decision-making only in the last six months before retirement, and another 22 per cent between six months and a year; only 28 per cent did so for more than two years (EBRI 2008; see also Feldman and Beehr 2011). A third cognitive and informational barrier is that workers may underestimate their life expectancy and their income needs in retirement. The latter is especially important if they encounter expensive health shocks after retiring. One strategy to address these misperceptions is for governments to provide increased warnings of vulnerability to risk of inadequate retirement income, especially to those who are most vulnerable (those with low

M4716-STRAßHEIM_9781785367847_t.indd 232

25/04/2019 13:32

Encouraging longer working lives  233 state pensions and low retirement savings). However, such a strategy also risks provoking greater political demands for more generous state pensions. Moreover, it is not clear that increased awareness of vulnerability will actually lead to self-protective extended work behaviour if workers adopt the fatalistic attitude that there is little that they can do to change their retirement income prospects or they lack the resources (see the discussion below) to continue work or substantially impact their retirement income streams. 2.4  Attitudinal and Peer Effect Barriers In addition to information and incentives, individuals may also be influenced in their labour market exit decisions by broader sets of attitudes and beliefs and by what they observe about the behaviour of others (peer effects). Attitudes affecting the labour force participation of older workers include both expectations about the life-course and stereotypes held about older workers both by older workers themselves and by the general public (employer attitudes will be discussed under autonomy). Relevant attitudes include both preferences for retirement and internalized (and often stigmatizing) attitudes toward older workers. These attitudes may vary both cross-nationally and across gender lines. A recent survey of 23 European countries, for example, found substantial variation both on what the public saw as the ideal age of retirement, and when men and women were seen as too old to work twenty hours a week or more. In France, for example, the ideal age retirement age for men was 59.4, and the age at which they were perceived to be too old to work more than 20 hours was 60.8; the equivalents for women were 55.8 and 56.5. The figures for Denmark were 63.2 for ideal retirement age and 68.3 for more than part time work for men; the comparable figures for women were 61.7 and 66.3 (Sweet 2009). Older workers are frequently stigmatized as poor performing, less motivated, less willing to engage in training, more flexible and less healthy, among other attitudes (Ng and Feldman 2012). ‘[A]ge discriminatory stereotypes that portrayed older people as passive, acquiescent, family oriented and disinterested in social and political participation’ are also common (Walker and Maltby 2012, p. S118). Researchers have suggested that these stereotypes are particularly likely to affect behaviour if they are internalized by workers, leading both to early job exit and to psychological disengagement while still employed: ‘the decline of older workers’ interest in their job may be understood as coping by devaluing the domain where they are stigmatized’ (Desmette and Gaillard 2008, p. 171; see also Gaillard and Desmette 2010). Peer effects may act through several mechanisms. Individuals are likely to follow what they perceive to be the ‘descriptive norm’ (what people usually do) rather than the ‘injunctive norm’ (what is perceived to be the ‘right’ thing to do) when the two conflict. Thus when retirement at the earliest possible eligibility date for a public pension is the descriptive norm, many older workers are likely to follow this norm, even if government agencies advise to the contrary. Several other peer effect mechanisms may also be at work. As Brown and Lanschever (2012, p. 91) note: ‘First, retirement is a complicated financial decision and individuals may rely on others as a source of information or even simply mimic the behaviour of others. Second, individuals may also enjoy retirement more (or enjoy their time at work less) if their long-time colleagues are retired.’ Similarly, several studies show that workers who have a spouse who is already retired are

M4716-STRAßHEIM_9781785367847_t.indd 233

25/04/2019 13:32

234  Handbook of behavioural change and public policy more likely to retire themselves, while having a spouse still at work reduces labour market exit, controlling for other causal factors (Litwin and Tur-Sinai 2015). 2.5  Resource and Autonomy Barriers Individuals may also forego longer working lives because they lack the capacity to do so – through the closely-related factors of lack of resources and lack of autonomy. A number of resources are particularly related to remaining in the labour force, notably health, skills, and financial assets. Research suggests that both subjective perceptions of one’s own health (and particularly changes in those perceptions) as well as objective health conditions affect retirement expectations, and that health shocks either to the worker him/herself or to someone else who requires their care – are an important contributor to unplanned departures from the labour force (McGarry 2004; Roberts et al. 2010). Persons with chronic illnesses and workers in manual occupations are thus particularly likely to be prone to early labour market exit. Strong educational attainment, on the other hand, makes it more likely that individuals will stay in the workforce at higher ages. Gary Burtless found that in the United States both male and female workers ‘who are better educated tend to remain in the workforce longer than workers in the same birth cohorts who have less schooling’ (Burtless 2013, p. 22). However, the blue collar/white collar divide in employability at older ages should not be over-generalized; a recent study finds that for some white-collar occupations in which specific cognitive and cognitive abilities – for example, ‘fluid cognitive abilities, quick reaction times and fine motor skills’– are required, ‘retirement tends to occur relatively early’ (Belbase et al. 2016, p. 5). While good health and marketable skills make it easier to stay in the labour force, both of these factors tend to be correlated with a high level of financial assets and/or pension entitlements, which may promote early labour market exit by making continued work financially unnecessary; for many of these workers, continued work is likely to be perceived as involving significant opportunity costs. Autonomy problems involve situations in which targets may want to behave in a particular way, but lack control over their own decisions or behaviour due to either reduced capacity (e.g. addiction) or control exercised by others. In the case of older workers, several types of limited autonomy are particularly important: statutory or employerestablished mandatory retirement ages, employer pressure on older workers to retire (see for example Dorn and Sousa-Poza 2010) and employer resistance to retaining and hiring older workers. The literature cites a number of stereotypes of older workers (for overviews, see for example Posthuma and Campion 2009; Bal et al. 2011). However, the importance of these stereotypes should not be overstated; a study of US human resources managers found that older workers are ‘at least as attractive if not more attractive than younger workers on many attributes including their strong work ethic, reliability, and high level of skill’ (James et al. 2007, p. 1). And a study of Dutch managers found that ‘[o]nly a few managers felt that older employees were less productive than younger ones. Attitudes about the reliability of older staff were also positive. Older workers are, however, often associated with a lack of adaptability and a resistance to innovation and, in particular, technological innovations’ (Henkens 2005, p. 362).

M4716-STRAßHEIM_9781785367847_t.indd 234

25/04/2019 13:32

Encouraging longer working lives  235

3.  STRATEGIC RESPONSES Strategic interventions to try to lengthen working lives can, as noted in the first section, take many different forms, from relatively non-intrusive information provision to negative incentives and prohibitions and requirements. Governments may seek to match strategies to appropriate barriers, for example providing information where there are information gaps. They can also shift over time to employ more intrusive strategies and settings on those instruments if initial choices fail to achieve their objectives. They may also choose strategies that focus on particular sub-elements of the target population (e.g. older workers with high skills and good health). The following discussion outlines several types of strategic options for extending working lives in advanced industrial societies. 3.1  Information and Admonition Approaches Given the large numbers and positive social construction of older workers and retirees, it is not surprising that governments in most OECD countries have relied heavily on relatively ‘soft’ or non-intrusive mechanisms such as information and admonition to change their behaviour. One common strategy is to provide annual or at least regular statements of anticipated income from public pensions with retirement at different ages. In the United States, for example, the Social Security Administration sends statements to workers at ages 25, 30, 35, 40, 45, 50, 55 and 60 who are not receiving Social Security benefits and do not yet have a ‘MySocialSecurity’ online account with the agency. Beyond age 60, statements are sent annually until a worker retires. These statements show, on the second page, the worker’s likely Social Security retirement benefits if they retire at age 62, the Full Retirement Age (which varies by birth cohort), and age 70, amid a long list of benefits under different scenarios (disability, spousal benefit in case of death, etc.). In short, it creates a situation of overload akin to reading a complicated nutrition label on packaged food. The statement for workers near retirement also states on its fourth page: ‘Before you decide to retire … [c]arefully consider the advantages and disadvantages of early retirement. If you choose to receive benefits before you reach full retirement age, your monthly benefits will be reduced.’ Two additional pages of information on retirement options, including a bar chart showing benefits at different retirement ages and a link to an online benefit estimator. The bar chart estimates in the chart are not personalized, however, but are based on a hypothetical benefit level of 1 000 dollars at the FRA. Sweden’s annual pension statement to workers takes a very different approach that gives much greater prominence to the effects of early or later retirement. The front page of the statement has only four figures: the chart for a worker born in 1973, for example, gives projections for retirement at ages 61, 65, 68 years and 6 months and 70, noting on a later page that under Sweden’s Notional Defined Contribution pension system, which adjusts benefits automatically for changes in life expectancy, a worker ‘needs to work until you are 68 years and 6 months to receive a pension equal to what you would have had to retire at age 65 on average life if life expectancy had remained unchanged’ since the beginning of the system. Governments can try to lower the information burden in labour market exit decisions by providing simpler and more complete information about expected retirement incomes across the multiple retirement streams that many workers now have – for example, public

M4716-STRAßHEIM_9781785367847_t.indd 235

25/04/2019 13:32

236  Handbook of behavioural change and public policy pensions, as well as employer-provided pensions and tax-advantaged retirement savings. Sweden, for example, now tries through its integrated Minpension (‘My Pension’) portal to provide integrated information about expected income streams from the state pension, employer pensions and personal pension savings. But this strategy is likely to be much more difficult in countries like the United States where there are a large number of retirement income providers, where individuals often have multiple assets and retirement income streams, and there are many different mechanisms for converting accrued retirement entitlements into pension pay-outs. Some governments also provide online benefit calculators and other sources of policy advice, but these vary greatly in quality and depth, and governments have generally been very reluctant to go beyond pointing out trade-offs, which are almost always very complex and individualized, regarding earlier or later labour market exit. Government pension agencies could in theory provide more detailed information and counselling, combined with admonitions to continue work for those with expected inadequate retirement income streams. For many government pension agencies, however, this would involve a substantial change in their organizational mission and in the orientation of front-line workers. Nor is it clear that government agencies would be trusted as an impartial source of policy advice, especially if extended working lives were seen as part of a partisan political agenda (Weyman et al. 2012, p. 16). Given the complexity of many workers’ retirement decisions, making sound decisions is likely to require something more akin to prolonged mentoring than simple information transmission. As a result, it is likely to be more available to those whose employers provide it, those who can afford to pay for it, and those who are best connected to non-governmental organizations who provide it for free or at low cost. It is, in short, likely to be most available to those who need it least. Information and admonition is also likely to be more effective for some workers than others, moreover. As a recent report for the UK Department of Work and Pensions put it: [R]ates of behaviour change are likely to be highest from targeting individuals who are more intrinsically engaged with work; who work in sectors where flexible employment is available and whose preference is to continue working for their current employer. There are concentrations of these individuals, who we have characterised as ‘the receptive middle’ within certain occupations/ employment sectors, and there may be a case for apportioning intervention resources to take account of this. (Weyman et al. 2012, p. 5)

In particular, the report’s authors concluded that working with employers of these ­workers to change retirement norms may be particularly efficacious. The problem, of course, is that these are the same workers who are on average least likely to be at risk of experiencing inadequate retirement income in the absence of government interventions. Extending working lives for those who are in poor health, have limited opportunities for flexible work and less accommodating employers will be a much more difficult problem to address. 3.2 Changing Incentives and Norms: Increasing Minimum Pension Eligibility and ‘Standard’ Retirement Ages ‘Standard’ or ‘full’ retirement ages, as noted above, are the ages at which ‘full retirement benefits’ are awarded according to whatever formula the pension system uses. Standard retirement ages can affect retirement behaviour in two different ways. The first is by

M4716-STRAßHEIM_9781785367847_t.indd 236

25/04/2019 13:32

Encouraging longer working lives  237 affecting norms. As the OECD noted in a recent review (OECD 2011, p. 20) ‘the “retirement age” is the most visible parameter of the pension system … it sends a clear signal for people in choosing when to cease work’.1 The second is by affecting incentives: if the standard retirement age is raised by two years, and workers need to wait (and presumably work) two years longer to get the same level of benefits (as happened when the Social Security Full Retirement Age was raised from 65 to 67), it creates strong incentives to work longer. Standard and minimum ages for receiving a pension declined in a number of OECD countries in the latter half of the twentieth century. The more recent pattern has been the reverse. Several countries have recently enacted measures to increase pension eligibility ages, sometimes equalizing retirement ages that had previous been lower for women than for men. Some of these have already taken effect in whole or in part. As of 2011, moreover, 14 OECD countries had legislated retirement age increases that were scheduled to take effect in later years (OECD 2011, p. 222). Some of these increases were scheduled to increase retirement ages past the once seemingly sacrosanct age of 65; the UK, for example, is gradually raising the State Pension Age to 68 in 2046, and the government has floated the idea of increasing it further to 69. Overall, the pensionable age for men averaged across the OECD member countries rose from 62.4 in 1999 to 62.9 in 2010, and under current legislation will rise to 64.6 in 2050, though actual changes will presumably be higher. The equivalent figures for women are 61.1, 61.8 and 64.4 (OECD 2011, pp. 25‒26). Changes in both standard retirement ages and earliest eligibility ages usually occur with very long lead times, both to avoid disadvantaging workers who may find it difficult to adjust their retirement and savings behaviour late in their careers, and to limit political repercussions. An alternative to raising the standard retirement age is to try to do away with the concept entirely. The new Swedish Notional Defined Contribution pension system developed in the 1990s attempts to do this, setting a flexible retirement age between 61 and 67, while gradually eroding pension levels for those who do not work longer. A flexible retirement age can be a mechanism for lowering opposition to pension retrenchment if it is supplemented with incentive mechanisms that encourage workers to remain in the labour market longer. This approach also has drawbacks, however: a flexible retirement age is more politically palatable precisely because it does not send a clear signal about the need to work longer to maintain anticipated pension levels. The risk is that as a result of this ambiguous signal, many workers will not alter their behaviour, and will end up with inadequate pensions. Although Swedes now receive annual pension statements that clearly show the impact of delaying retirement on benefits, changes in retirement behaviour under the new pension system have been modest. The age of labour market exit has increased by 1.3 years since 1994, but labour force participation rates for workers aged 65‒74 measured in full time equivalents have been stagnant since 2008, and increased life expectancy has meant that the gap between the age of labour market exit and average life expectancy has actually increased slightly since the late 1990s (SOU 2013, pp. 124‒126; see also Pensionsmyndigheten 2014a). The dominant trend has been a dispersion of retirement ages under the new, more flexible retirement age system rather than an increase (Pensionsmyndigheten 2014b). In addition to political feasibility of increases in the ‘standard’ pension age, those increases also raise important social concerns that need to be addressed by ­policy-makers

M4716-STRAßHEIM_9781785367847_t.indd 237

25/04/2019 13:32

238  Handbook of behavioural change and public policy who are considering such steps. A first concern is its impact on populations who are disadvantaged. Evidence from the United States, for example, suggests that minority workers who experienced difficulties in entering the labour market are also especially likely to experience job dislocations and involuntary labour market exit as they approach retirement age (Flippen and Tienda 2000). This effect is likely to be especially pronounced during economic downturns (Coile and Levine 2007). In addition, studies of West European countries have found differences at age 65 between high-education and low-education groups (a proxy for socio-economic status) in both life expectancy and disability-free life expectancy (DFLE) that are strong and consistent across countries (Majer et al. 2011). Moreover, those disparities appear to be increasing over time in many countries (Singh and Siahbush 2006; Kibele et al. 2013), as better off groups benefit disproportionately from increases in life expectancy. Thus if increased inequality of outcomes is to be avoided, policies that require longer working lives must be balanced by policy initiatives that provide adequate disability coverage to those who cannot work longer, especially many people who spent their working lives in manual occupations. 3.3  Prohibitions and Requirements Policies that prohibit labour market exit prior to a specified age or require continued work beyond a specified age are neither practical nor politically feasible given the potential for disability. However, policies setting state pension eligibility requirements and benefits and policies on withdrawal of tax-advantaged retirement savings have such strong incentive effects that their practical effect can be close to a requirement for some older workers. Policies that bar mandatory retirements can also have a major influence on extending working lives for those who already have a preference for continued work. Preventing employers from discriminating against older workers and allowing those who want to work longer to do so without fear of being forced into retirement by their employers (especially before they reach pensionable age) are especially important, as is monitoring and enforcement of those policies.

4.  CONCLUSION AND LESSONS This discussion provides several lessons for policy designers and programme implementers. First, it suggests that for extending working lives, as with most policies that have diverse target populations, it is critically important to look at multiple barriers to behaviour change – notably complex incentives and a high degree of uncertainty that challenge the cognitive capacity of most workers, plus substantial peer effects – and to develop policy interventions that address those barriers. Second, it is equally important to recognize that the population of older workers is very heterogeneous in the barriers that they confront and conditions facilitating extension of their working lives. They will not all respond to information, incentive structures or provision of resources in the same way. Rather than a single strategy to extend working lives, therefore, it is important to identify sub-populations that confront particular constellations of barriers and focus on specific strategies to address those barriers – although

M4716-STRAßHEIM_9781785367847_t.indd 238

25/04/2019 13:32

Encouraging longer working lives  239 most policies must apply to the entire population of older workers, some will have more impact on the behaviour of some sub-populations than others (Weyman et al. 2012). Incentive-based strategies, for example, may be most likely to change the behaviour of workers of middle-income and reasonably good health. Providing new resources and opportunities – notably opportunities for flexible work opportunities – may be more likely to change the behaviour of those with health difficulties and those (disproportionately women) with substantial caring responsibilities. Third, very substantial changes in individual and aggregate levels of behaviour are likely to require intrusive measures and ‘intense’ settings – but that sparks lots of opposition. Thus policy-makers also need to analyse how much compliance with the objective of extending working lives – and by whom – is ‘good enough’ to achieve their objectives. Clearly it will not be practical to get all older workers to work at higher ages, given the resource (notably health) and autonomy limitations that many of them face. Deciding which workers can work at higher ages and which cannot because of disability, will impose a higher burden on the administrative capacity of the state, lower budgetary savings from extended working lives and – if disability programmes are perceived by the public to be riddled with fraud – undermine trust in and support both for government and the welfare state. Fourth, policy-makers need to think strategically for the long-term in trying to extend working lives. This will almost certainly involve multiple phases, as in the fifty-year long campaign against smoking. Incremental changes and less-intrusive policy instruments for changing the behaviour of individual workers will be the primary tools in the short-term. A final note of caution is in order: most of the research that is cited in this chapter both on the barriers to extending working lives and to the effectiveness of policies to delay labour market exit relates to efforts to prevent early exit – that is, exit before the ‘standard’ retirement age at around age 65. How generalizable these results are to efforts to extend working life beyond that age, and in particular to the upper 60s and beyond, remains to be seen.

NOTE 1. Retirement age is in itself a fuzzy concept in many countries, because in some countries pension eligibility is linked to years of work as well as reaching a minimum age – those who have worked a specified number of years may be eligible for a pension before reaching that age. Special retirement ages may also be set for particular occupations, or on other grounds.

REFERENCES Alleva, B. (2015), ‘Minimizing the Risk of Opportunity Loss in the Social Security Claiming Decision’, Journal of Retirement, 3 (1), 61‒87. Bal, A., A Reiss, C. Rudolph and B. Baltes (2011), ‘Examining Positive and Negative Perceptions of Older Workers: A Meta-Analysis’, The Journals of Gerontology B: Psychological Sciences and Social Sciences, 66B (6), 687‒698. Barr, N. (2006), ‘Pensions: Overview of the Issues’, Oxford Review of Economic Policy, 22 (1), 1‒14. Behaghel, L. and D. Blau (2012), ‘Framing Social Security Reform: Behavioral Responses to Changes in the Full Retirement Age’, American Economic Journal: Economic Policy, 4 (4), 41‒67.

M4716-STRAßHEIM_9781785367847_t.indd 239

25/04/2019 13:32

240  Handbook of behavioural change and public policy Belbase, A., G. Sanzenbacher and C. Gillis (2016), ‘How Do Job Skills That Decline With Age Affect WhiteCollar Workers?’, Center for Retirement Research at Boston College Issue in Brief 16-6. Brown, K. and R. Laschever (2012), ‘When They’re 64: Peer Effects and the Timing of Retirement’, American Economic Journal: Applied Economics, 4 (3), 90‒115. Burtless, Gary (2013), ‘Who Is Delaying Retirement: Analyzing the Increase in Employment Among Older Workers’, in G. Burtless and H. Aaron (eds), Closing the Deficit: How Much Can Later Retirement Help? Washington, DC: The Brookings Institution, pp. 11‒35. Coile, C. and P. Levine (2007), ‘Labor Market Shocks and Retirement: Do Government Programs Matter?’, Journal of Public Economics, 91 (10), 1902‒1919. Desmette, D. and M. Gaillard (2008), ‘When a “worker” becomes an “older worker”’, Career Development International, 13 (2), 168‒185. Dorn, D. and A. Sousa-Poza (2010), ‘“Voluntary” and “Involuntary” Early Retirement: An International Analysis’, Applied Economics, 42(4), 427‒438. Employee Benefits Research Institute (EBRI). 2008. EBRI 2008 Recent Retirees Survey: Report of Findings, Issue Brief No. 319 (July). Feldman, D. and T. Beehr (2011), ‘A Three-Phase Model of Retirement Decision Making’, American Psychologist, 66 (3), 193‒203. Fisher, G., D. Chaffee and A. Sonnega (2016), ‘Retirement Timing: A Review and Recommendations for Future Research’, Work, Aging and Retirement, 2 (1), 230‒261. Flippen, C. and M. Tienda (2000), ‘Pathways to Retirement: Patterns of Labor Force Participation and Labor Market Exit Among the Pre-Retirement Population by Race, Hispanic Origin, and Sex’, Journal of Gerontology: Social Sciences, 55 (1), S14‒S27. Gaillard, M. and D. Desmette (2010), ‘(In)validating Stereotypes About Older Workers Influences Their Intentions to Retire Early and to Learn and Develop’, Basic and Applied Social Psychology, 32 (1), 86‒98. Henkens, K. (2005), ‘Stereotyping Older Workers and Retirement: The Managers’ Point of View’, Canadian Journal on Aging, 24 (4), 353‒366. James, Jacquelyn B., Jennifer E. Swanberg and Sharon P. McKechnie (2007), Responsive Workplaces for Older Workers: Job Quality, Flexibility and Employee Engagement, Center on Aging and Work at Boston College, Issue Brief 11. October. Johnson, Richard W., Karen E. Smith and Owen Haaga (2013), How Did the Great Recession Affect Social Security Claiming?, Washington, DC: Urban Institute Program on Retirement Policy, Brief # 7. Kibele, E., D. Jasiolonis and V. Shkolnikov (2013), ‘Widening Socioeconomic Differences in Mortality Among Men Aged 65 Years and Older in Germany’, Journal of Epidemiology and Community Health, 67 (5), 453‒457. Litwin, H. and A. Tur-Sinai (2015), ‘The Role of the Social Network in Early Retirement Among Older Europeans’, Work, Aging and Retirement, 1 (4), 340‒349. Maestas, N. (2010), ‘Back to Work: Expectations and Realizations of Work after Retirement’, Journal of Human Resources, 45, 718‒748. Maestas, N. and J. Zissimopoulos (2010), ‘How Longer Work Lives Ease the Crunch of Population Aging’, The Journal of Economic Perspectives, 24 (1), 139‒160. Majer, I., W. Nusselder, J. Mackenbach and A. Kunst (2011), ‘Socioeconomic Inequalities in Life and Health Expectancies Around Official Retirement Age in 10 Western-European Countries’, Journal of Epidemiology and Community Health, 65, 972‒979. McGarry, K. (2004), ‘Health and Retirement: Do Changes in Health Affect Retirement Expectations?’, Journal of Human Resources, 39 (3), 624‒648. Munnell, A.H. and S.A. Sass. (2006), Working Longer: The Solution to the Retirement Income Challenge, Washington, DC: The Brookings Institution. Ng, T. and D. Feldman (2012), ‘Evaluating Six Common Stereotypes About Older Workers with MetaAnalytical Data’, Personnel Psychology, 65 (4), 821‒858. OECD (2011), Pensions at a Glance 2011: Retirement-Income Systems in OECD and G20 Countries, Paris: OECD Publishing. Pensionsmyndigheten (2014a), Medelpensioneringsålder och utträdesålder 2013 [Average Retirement Age and Labour Market Exit Age 2013], Stockholm: Pensionsmyndigheten. Pensionsmyndigheten (2014b), Orange Rapport: Pensionssystemets årsredvisning 2013 [Orange Report: Annual Report of the Pension System 2013], Stockholm: Pensionsmyndigheten. Posthuma, R. and M. Campion (2009), ‘Age Stereotypes in the Workplace: Common Stereotypes, Moderators, and Future Research Directions’, Journal of Management, 35 (1), 158‒188. Rix, S. (2008), ‘Age and Work in the United States of America’, in Philip Taylor (ed.), Ageing Labour Forces:  Promises and Prospects, Cheltenham, UK and Northampton, MA: Edward Elgar Publishing, pp. 111‒140. Roberts, J., N. Rice and A. Jones (2010), ‘Early Retirement Among Men in Britain and Germany: How Important is Health?’, The Geneva Papers on Risk and Insurance - Issues and Practice, 35 (4), 644‒667.

M4716-STRAßHEIM_9781785367847_t.indd 240

25/04/2019 13:32

Encouraging longer working lives  241 Schneider, A. and H. Ingram (1990), ‘Behavioral Assumptions of Policy Tools’, The Journal of Politics, 52 (2), 510–529. Schneider, A. and H. Ingram (1993), ‘Social Construction of Target Populations: Implications for Politics and Policy’, The American Political Science Review, 87 (2), 334–347. Singh, G. and M. Siahpush (2006), ‘Widening Socioeconomic Inequalities in US Life Expectancy, 1980–2000’, International Journal of Epidemiology, 35 (4), 969‒979. Song, J. and J. Manchester (2007), ‘Have People Delayed Claiming Retirement Benefits? Responses to Changes in Social Security Rules’, Social Security Bulletin, 67 (2), 1‒23. SOU (2013), Åtgärder för ett längre arbetsliv [Measures for a Longer Working Life], Statens offentliga utredningar, 2013:25. Sweet, S. (2009), When Is a Person Too Young or Too Old to Work?: Cultural Variations in Europe. Sloan Center on Aging and Work at Boston College. Global Issue Brief No. 2. van Bavel, Jan and Tom De Winter (2013), ‘Becoming a Grandparent and Early Retirement in Europe’, European Sociological Review, 29 (6), 1295‒1308. Vickerstaff, S. (2010), ‘Older Workers: The “Unavoidable Obligation” of Extending Our Working Lives?’, Sociology Compass, 4, 869–879. Walker, A. and T. Maltby (2012), ‘Active Ageing: A Strategic Policy Solution to Demographic Ageing in the European Union’, International Journal of Social Welfare, 21 (s1), 117–130. Wang, M. and J. Shi (2014), ‘Psychological Research on Retirement’, Annual Review of Psychology, 65, 209‒233. Weaver, R.K. (2014), ‘Compliance Regimes and Barriers to Behavioral Change’, Governance, 27 (2), 243‒65. Weaver, R.K. (2015), ‘Getting People to Behave: Research Lessons for Policymakers’, Public Administration Review, 75 (6), 806–816. Weyman, A., D. Wainwright, R. O’Hara, P. Jones and A. Buckingham (2012), Extending Working Life: Behaviour Change Interventions, Department for Work and Pensions, Research Report No 809.

M4716-STRAßHEIM_9781785367847_t.indd 241

25/04/2019 13:32

17.  Behavioural economics and development policy Christian Berndt

1. INTRODUCTION There is growing recognition that Behavioural Insights (BIs) – by focusing on how people actually make choices – contribute to delivering more targeted and effective policy solutions. The understanding of human behaviour is already informing policy-making and contributing to the design of new forms of intervention, as well as complementing traditional approaches (i.e. regulations, incentives, and information requirements). The last few years have seen major developments in the application of BIs to different policy areas. (Lourenço et al. 2016, p. 6)

After the journey of behavioural economics from the margins to the heart of mainstream economic thinking had found a preliminary endpoint with the award of the Nobel Prize in Economics to cognitive psychologist Daniel Kahneman in 2002, it quickly caught the eye of policy-makers and consultants. Aided by a number of popular academic books and by further blows to the mainstream economic edifice in the wake of the 2008 financial crisis, the revival of the behavioural approach in economics ushered in a ‘behavioural turn’ within a wide range of policy fields. The majority of these changes affected policy delivery in the Global North, epitomized by the ‘Behavioural Task Force’ in the UK, the ‘The Social and Behavioral Sciences Team’ in the US, and recently also documented at the level of the European Union (Lourenço et al. 2016). Less attention has so far been paid to the extent to which behavioural thinking has also travelled to the Global South, in particular with a view to development policy. Here, behavioural economics teamed up with a fellow traveller, experimentalism, emerging from a parallel movement informed by advances in the neighbouring discipline of experimental economics. Saugato Datta and Sendhil Mullainathan have recently commented on this double move, referring to a ‘behavioural revolution’ that has been accompanied by a similarly far-reaching ‘evaluation revolution’ (Datta and Mullainathan 2014, p. 8). This chapter has its particular focus on the rise of the behavioural approach and experimentalism in the field of development policy. After briefly mapping the journey of behavioural thought into the subdiscipline of development economics in the first section, the chapter reconstructs how this particular scholarly knowledge has been translated into concrete policy interventions (section 2). Section 3 then engages critically with the behavioural turn in development, discussing the conception of the human subject informing both the scholarly and the more applied literature, the status of behavioural economics as a challenger of the orthodox neoclassical treatment of poverty in the Global South, the mobilization of classical imaginations of progress and modernization, and the extent to which the behavioural turn can be seen as another example of neoliberal ‘third way’ policies that are uninterested in questions of social inequality and the underlying reasons for poverty.

242

M4716-STRAßHEIM_9781785367847_t.indd 242

25/04/2019 13:32

Behavioural economics and development policy  243

2.  BEHAVIOURAL ECONOMICS AND DEVELOPMENT When behaviourism started to become influential in the social sciences more generally during the 1950s and 1960s, it did not leave economics untouched (Simon 1978, 2008). However, attempts to move away from narrow definitions of economic behaviour met resistance by the custodians of more abstract mathematical modelling and opponents to realism in economics. It was only when it came into contact with advances in cognitive psychology and brain science that economic behaviourism received a new breath of life.1 In the early 1970s psychologists Daniel Kahneman and Amos Tversky co-authored a number of articles that criticized the rational-agent model (e.g. Tversky and Kahneman 1974). These insights caught the interest of the economist Richard Thaler. Subsequent joint research between Kahneman, Thaler and others then established behavioural economics as a discipline finally taken seriously by economists. The insights from Kahneman, Tversky, Thaler and others would be unproblematic for standard economic theory as long as these deviations were small and idiosyncratic, that is, if there was reason to assume that they would on average cancel themselves out. The problem, however, is the universalist claim that there are systematic biases built into people’s choices which prevent utility maximization. At a time of mounting scepticism about the political applicability of mainstream economic thinking the behavioural approach was quickly able to fill a void. An interesting case is the realm of development where concepts from behavioural economics have increasingly been applied to questions of poverty and risk management. In this context, it is important to briefly reflect on how ‘poverty’ is conceptualized in behavioural economics. In traditional ‘neoclassical’ economics there is nothing special about ‘the poor’. They may have difficult lives, but they try to make ends meet under these circumstances, thus behaving just as rationally as other people. In fact, having to literally live with scarcity and to think carefully of how to put available resources to use, the poor are almost natural homines economici. Or as Theodore W. Schultz put it in his 1979 Nobel Prize Lecture: ‘The major mistake has been the presumption that standard economic theory is inadequate for understanding low income countries and that a separate economic theory is needed’ (Schultz 1979, p. 1), continuing with a view to the rural poor in the Global South that smallholders ‘the world over’ are ‘calculating economic agents’ and ‘fine-tuning entrepreneurs, tuning so subtly that many experts fail to recognize how efficient they are’ (Schultz 1979, p. 3). According to this logic the poor are driven to allocate land, labour and capital in a way that guarantees an efficient outcome, at least given the circumstances (Banerjee 2001, p. 129; Duflo 2006, pp. 367‒371). This position was modified in the 1980s and 1990s, when it was recognized in the wake of the ongoing softening of rationalist assumptions that market failures directly linked to poverty resulted in inefficient outcomes. It was argued that imperfect information, moral hazard, limited liability or adverse selection are particularly salient for people living in poverty, with the effect that the market mechanism produces inefficient outcomes ‘even if everybody is perfectly rational’ (Duflo 2005, p. 22). Abhijit Banerjee (2001) pointed to two explanations for this in the standard economic literature that partly contradict each other. According to the first the poor are different because they are desperate and have nothing to lose. For instance, they may lack the incentive and discipline to repay a loan, lenders therefore thinking twice before giving them

M4716-STRAßHEIM_9781785367847_t.indd 243

25/04/2019 13:32

244  Handbook of behavioural change and public policy money. At the end of the day the poor may not be given the same opportunities as richer people, trapping them in poverty. The second explanation links poverty with vulnerability. According to this logic this results in the poor being overly cautious and risk averse, as losses cost them dearly, leading to a vicious cycle of underinvestment. Although there are ways to reconcile both moments of poverty in the neoclassical literature, both have obvious contradictions, given that vulnerability is almost the literal opposite of having too little to lose. It is also obvious that both moments may speak to different segments of the population (i.e. dividing the ‘poor’ from the ‘very poor’) (Banerjee 2001, pp. 131, 139; Duflo 2006, pp. 371‒372). But such ambivalences notwithstanding, these explanations deal with poverty-related structural constraints on the decision-making of otherwise rational individuals (incomplete information, distorted incentives; Duflo 2006, p. 376). Behavioural economists criticize this position. They argue that institutional arrangements in the households, families and communities of the poor are too complex to be reduced to mere questions of better information and incentives, or policies to introduce owner-occupational land tenancy and private property rights (Duflo 2006, p. 371). Instead a different approach is suggested, scholars arguing that poverty has implications that affect how people think and act in different ways than the ‘non-poor’. Poverty creates particular psychological burdens, leading to a negative ‘feedback loop’ that results in disadvantageous economic behaviour and ‘behavioural bottlenecks’ (Datta and Mullainathan 2014; Haushofer and Fehr 2014, p. 866; Mullainathan and Shafir 2013). The first of these ‘burdens’ is loss aversion. Tversky and Kahneman (1974) have pointed out that people ‘suffer’ from an asymmetry in their preferences as responses to losses are consistently much more intense than responses to corresponding gains. Subsequently labelled ‘loss aversion’ and put under the label of ‘prospect theory’ this has particular implications in situations of poverty. Poor people, it is argued, are very hesitant to take the risk of trading off their meagre current incomes against potentially higher ones in the future, leading to a lower adoption rate of new technologies and lower human capital investments. All this apparently decreases future income further, resulting in the vicious circle and poverty trap mentioned above (Haushofer and Fehr 2014, p. 862). For instance, in the context of the Global South it is assumed that vulnerable farmers are very reluctant to change their behaviour, sticking to established ways of doing things instead. In the case of shocks this may result in a willingness to ‘toughen it up’ and to rely on informal ways to cope with risks rather than changing behaviour (Fafchamps 2009, p. 12). In the academic literature this position is regularly connected with the question of property rights. Starting with the assumption that the poor often live in contexts where well-defined property rights are absent, it is maintained that they have to invest large amounts of resources to secure what they regard as their land and house (Mullainathan 2005, p. 20). There are obvious links of this debate with Hernando de Soto’s property titling campaigns (see de Soto 2000; Mitchell 2005) and the marketization and formalization of land use in the Global South more generally. A second ‘behavioural weakness’ that appears to be exacerbated in situations of poverty is what is referred to as ‘self-control problem’ or procrastination. This is connected with the idea that people discount time. Discounting refers to the propensity to value the present higher than the future, leading to a tendency to enjoy benefits now and postpone costs to the future. In its hyperbolic variant discounting is time-inconsistent, for instance when trade-offs between the present and the near-future are high, and those between

M4716-STRAßHEIM_9781785367847_t.indd 244

25/04/2019 13:32

Behavioural economics and development policy  245 the near and the far future low. This means that people are impatient in the present and patient in the future (Duflo 2005, p. 14; Mullainathan and Thaler 2001, p. 1096; OECD 2012, p. 12). The argument is that while the time and risk preferences of the poor may not differ from those better off, they are more likely to suffer from liquidity constraints, for instance, because of the more limited access to credit markets. This may result in an even greater bias for the present, a short-termism that lets them forego more sustainable decisions. For instance, it may imply that income earned after selling the harvest is spent right away and/or that decisions to buy fertilizer or other inputs are shifted to an indeterminate future. Another variant of this position is the assumption that poor people tend to delay unpleasant activities in the present even if they have large returns in the future (Duflo 2005, p. 9). Operating at the margin of the market economy, these risks are seen as being particularly painful to the poor in the Global South. A classical vicious circle of low income and lack of savings, acerbated by shocks, results in a preference for solutions with more reliable, but lower average returns and a corresponding unwillingness to take on additional risks by borrowing and making long-term investments due to this uncertainty. In sum, ‘it is reasonable to think that the poorer someone is, the more he [sic] dislikes taking risks … [preferring] a safe strategy with a low return to a riskier strategy with a higher return’ (Duflo 2006, p. 372). Accordingly, it is not so much that poor people have an intrinsically higher degree of ‘behavioural bottlenecks’, but a life in conditions of poverty that exacerbates the problem. Poverty-related stress is believed to make people unhappy, depressed and anxious, inducing a ‘shift from goal-directed to habitual behavior’ (Haushofer and Fehr 2014, pp. 863‒866). According to the findings of behavioural experiments providing the background to these claims, this is generally true whether you live in the Global North or Global South. But it is the larger extent of poverty in ‘southern’ countries and perceived higher vulnerability of the poor that lends itself to a particular urgency in the eyes of behavioural development economists. ‘Behavioral economics’, argue Karlan and Appel clearly having the Global South in their minds (Karlan and Appel 2011, p. 20), ‘reveals that … poor people make mistakes that end up making them poorer, sicker, and less happy.… Identifying and correcting these mistakes is a prerequisite for solving global poverty.’ It is with regard to the Global South in particular therefore that protagonists demand a paradigm shift from an emphasis on improving institutions to solve ‘problems between people’ to a focus on ‘problems within a person’ (Mullainathan 2005, p. 33; emphasis removed) and the ‘cognitive, motivational and even sociological limits on action’ (Anand and Lea 2011, p. 284).

3.  DESIGNING FOR DEVELOPMENT Policy templates travel through networks that connect ideas with people. It is a key insight of recent research on ‘fast policies’ (Peck and Theodore 2015) and the performativity of economic knowledge (Berndt and Boeckler 2012) that academic scholars play a key role in translating economic and psychological ideas into policy practice. In the case of development a number of scholars connected with the behavioural and experimental turn in economics stand out, such as Esther Duflo and Abhijit Banerjee (MIT), Sendhil Mullainathan and Michael Kremer (Harvard), Eldar Shafir (Princeton) and Dean Karlan (Yale). All these scholars are involved, often in overlapping functions, in institutions and

M4716-STRAßHEIM_9781785367847_t.indd 245

25/04/2019 13:32

246  Handbook of behavioural change and public policy think tanks that reach out to the world of the development industry, such as the Abdul Latif Jameel Poverty Action Lab (J-PAL) at MIT (founded in 2003 by Duflo, Banerjee and Mullainathan), ideas42 (a think tank established by Mullainathan, Shafir and others at Harvard with support from IFC), or Innovations for Poverty Action set up by Karlan. There are also southern nodes in the behavioural development economics network, for instance, the University of Cape Town in South Africa with the African subsidiary of J-PAL and the Research Unit in Behavioural Economics and Neuroeconomics (RUBEN) or the Busara Center for Behavioral Economics in Nairobi, Kenya, recently founded by Johannes Haushofer who has links to the behavioural economist Ernst Fehr (University of Zurich) and J-PAL. There are also close connections with the international donor community and multinational development institutions such as the World Bank. As pointed out above, behavioural ideas about ‘why the poor stay poor’ have quickly entered the policy realm, eventually also being translated into a means to design interventions in the Global South (Datta and Mullainathan 2014, p. 8). There is no better proof that behavioural economics has successfully completed its journey into the field of development policy than the 2015 World Development Report (WDR). Titled ‘Mind, Society and Behavior’ the report is an enthusiastic plea for more efficient development policies and interventions by ‘paying attention to how humans think (the processes of mind) and how history and context2 shape thinking (the influence of society)’ (World Bank 2015, p. 2). A ‘science of design for development’ emerged (Datta and Mullainathan 2014, p. 8) targeting human choice and action (behaviour) and transforming the way development policies are formulated and implemented. There are three qualities of this ‘behavioural revolution’ in development policy (Datta and Mullainathan 2014, p. 10), in particular, that are worth highlighting. First, protagonists of the behavioural turn in development distance themselves both from neoclassical market-oriented policies as well as from traditional development aid and large-scale state interventions. Both market and state are found wanting. Against this, scholars claim to occupy the middle ground (see, for instance, Banerjee and Duflo’s positioning in the ‘Easterly-Sachs-debate’; Banerjee and Duflo 2011, pp. 3‒16). This resonates with the suggestion of asymmetric or libertarian paternalism in the behavioural economic literature as an institutional frame that is capable of intervening politically with as much state as necessary and as much free market as possible. Both terms depict policies that are ‘smart’, that is, policies that help those who are less sophisticated cognitively ‘while imposing little or no harm on those who are fully rational’ (Camerer et al. 2003, p. 1212; see also Thaler and Sunstein 2008, p. 249). Libertarian paternalist interventions turn into means to change behaviour, being capable of curing the behavioural defects that are ultimately made responsible for poverty and underdevelopment. However, given that behaviour can be expected ‘to be adaptive’ there is hope: ‘market players can “learn” more efficient behavior’ (Swiss Agency for Development and Cooperation 2008, p. 8). The market is capable of healing the behavioural deficiencies by disentangling the rural poor from the bonds of traditional cultural and social conditions and by enabling them to take the initiative into their own hands (Anderson and Stamoulis 2006, p. 24). Second, development practitioners who translate behavioural thinking into concrete interventions are fully aware that much more is needed for successful behaviour change than waiting for the self-healing forces of the market. This refers to the practical side of the behavioural turn in development. The template for the formulation of concrete

M4716-STRAßHEIM_9781785367847_t.indd 246

25/04/2019 13:32

Behavioural economics and development policy  247 policy interventions is provided by Thaler and Sunstein’s (2008) bestselling book Nudge: Improving Decisions about Health, Wealth, and Happiness. Nudging is about the construction and management of ‘choice architectures’ in order to channel the behaviour of ‘humans’ into a direction that is deemed socially beneficial. Nudges include the framing of information, anchoring, simplification of products and procedures, and also reminders and commitment devices (World Bank 2015, p. 36), most frequently addressing the kind of self-control problems mentioned above. Commitment devices for savings, for instance, see to it that consumers voluntarily stop withdrawing money from their accounts until a certain target level is reached. In rural contexts where poor smallholders are assumed to need incentives to adopt riskier agricultural practices, simple text messages remind recipients not to forget to buy subsidized fertilizer in a given time window, or there is the supply of critical market or production data in a way that smallholders are almost forced to calculate and to entertain the idea of whether or not to take more risks. ‘Battles of self-control’ (Datta et al. 2014, p. 16) are also waged at the workplace in order to increase labour productivity. Datta et al. (2014) refer to an unpublished study of an Indian dataentry company where workers ‘voluntarily’ committed themselves to self-chosen dataentry targets, a sort of ‘negative bonus contract’ that arguably increased self-responsibility and commitment. The workers are reported to have enthusiastically adopted a payment arrangement that penalizes them when failing to hit the targets, apparently making this a promising intervention in situations ‘where workers don’t work as hard as they themselves would like to’ (Datta and Mullainathan 2014, p. 24). Third, the behavioural revolution cannot be separated from the turn towards measurement and valuation. This refers to the productive articulation of behavioural economic thinking with experimentalism,3 making it ‘possible to measure whether a given programme or policy works’ (Datta and Mullainathan 2014, p. 8; see also Deaton 2010). This apparently allows protagonists to base interventions on hard facts about what works and what does not, or in the deceptively simple words of Banerjee and Duflo: ‘We need evidence’ (2011, p. 4). A crucial step has been the development of the randomized field experiment. Randomization is used to neutralize differences between groups. Subjects are assigned randomly to either a control or experimental group, under the assumption that variations with regard to unidentified factors will be distributed evenly across the groups (Guala 2005, pp. 62ff). Although the underlying principle is the classical economic notion of ‘ceteris paribus’ (all other things being held equal), randomization travelled into economics from the medical world where the so-called randomized controlled trial (RCT) has long been an established procedure in the context of clinical investigation. Randomization plays a crucial role in the spread of behavioural thinking into the policy realm. While this includes examples in the Global North, it has been in the context of development in the Global South that this technology has really taken off (Bédécarrats et al. 2015, pp. 5‒6). All this has profound political implications. This crucially concerns the rationale of hard evidence, linking the method with the wider trend towards evidence-based policy delivery (Datta and Mullainathan 2014, p. 31).

M4716-STRAßHEIM_9781785367847_t.indd 247

25/04/2019 13:32

248  Handbook of behavioural change and public policy

4. QUESTIONING THE BEHAVIOURAL AND EXPERIMENTAL TURN By focusing on the behavioural dimensions of poverty, interventions that target human choice and behaviour at the individual level are legitimized, turning ‘highly cost-effective behavioral interventions’ (World Bank 2015, p. 13) into the new gold standard for development policies. In what follows, I turn to four wider tendencies that demand more careful scholarly attention. 4.1  Econs versus Humans? The key idea allowing the articulation of behavioural economic theory with the political sphere of ‘big-D’ Development (Hart 2001, p. 650)4 in the form of libertarian paternalism is a deceptively simple model of human cognition. Taken from cognitive psychology (‘dual process model’), the argument is that judgments can ideally be produced in two ways, ‘a rapid, associative, automatic, and effortless intuitive process (sometimes called System 1), and a slower, rule-governed, deliberate and effortful process (System 2)’ (Kahneman 2002, p. 8; see also Kahneman 2011, pp. 20‒24). System 1 is automatic and unconscious. System 2 is rule-based, rational and explicit. It ‘monitors’ system 1 and is able to rationalize ideas and feelings that were generated by system 1. It is also able to correct or replace erroneous intuitive judgments. However, this does not happen all the time. Since system 2 has its limits, system 1 often prevails, leading to the kind of ‘behavioural bottlenecks’ and ‘anomalies’ mentioned above (Kahneman 2002, p. 8). While this model is in principle assumed to have universal traction, when translated into the policy realm the world of the two cognitive systems becomes the world of different types of people. On the one hand are the experts who are (almost) rational, on the other ordinary people who are mainly steered by emotions, affect, and rules of thumb, and are locked in suboptimal outcomes. In their popular, bestselling books both Thaler and Sunstein as well as Kahneman attach catchy labels to both sides. On the one hand are ideal type ‘econs’, fully rational and modelled after the famous homo economicus. On the other side are imperfect ‘humans’, driven by system 1 with system 2 only delivering spurious checks and balances (Kahneman 2011, p. 413; Thaler and Sunstein 2008, p. 7). The connections with the ‘problem of poverty’ are obvious: The poor ‘human’ lives mainly in system 1, the cognitive world that is automatic and unconscious. Decisions of the non-poor ‘econ’ are more likely to be checked by rational and explicit system 2. It is impossible in this logic to force people to behave against their (imperfect) nature. This provides a crucial step in the discourse. In a situation in which it appears to be economically efficient for the rural poor to adopt a more entrepreneurial strategy, but in which there are systematic cognitive biases that prevent them from doing so, there is a need for incentives to change behaviour ‘voluntarily’. Such a view also lends a reassuring legitimacy to those who deliver the nudges and experiments. As development experts they ought to know better. They are ‘econs’ not unlike the expert billiard player in Thaler’s adaption of a famous example developed by Milton Friedman and L.J. Savage. An expert player can be expected to choose the best shot in any situation. Her performance would not deviate much from the optimum as suggested by a mathematical formula. The expert billiard player is almost perfectly rational,

M4716-STRAßHEIM_9781785367847_t.indd 248

25/04/2019 13:32

Behavioural economics and development policy  249 resulting in the most efficient, the best solution. Intermediate or novice players, on the other hand, operate with different models. Both use heuristics and rules of thumb and both are ‘rational’ in the context of these frames. They do the best they can. But at the same time their performance generates suboptimal outcomes. ‘How does [economic| behaviour relate to billiard behaviour?’, concludes Thaler (Thaler 1980, p. 58; my emphasis), ‘Again there will be various classes of consumers. Some will be experts (PhDs in Economics?), others will be novices (children?) … these shoppers are doing the best they can.’ It is interesting to see that the authors of the 2015 WDR acknowledge the obvious contradictions in this reasoning. There is a full chapter devoted to the insight that development practitioners may eventually be as ‘human’ as those subject to their interventions. The proposed antidote against own imperfections is to ‘engage in more systematic efforts to understand the mindsets of those they are trying to help’. The experts should ‘eat their own dog food’ (World Bank 2015, p. 190). They appear to be nothing without their technologies, programme templates and nudges. However, given that these are devices that only they can master, practitioners are reassured of their expert status. 4.2  A Challenge to Mainstream ‘Neoclassical’ Economics? The identity of behavioural economics as a serious challenger to mainstream economics rests on the argument that there are systematic biases built into people’s choices. As a positive intellectual project (that is, describing and predicting what people actually do) behavioural economics has indeed played an important role in further breaking the spell of the perfect rationality hypothesis. However, such an argument neglects the fact that the translation of its key insights into policy interventions transformed behavioural economics into a normative endeavour. After all, what behavioural engineering and libertarian paternalism are about is to change how people behave. And, as the equation of the dual process model with two types of people aptly illustrates, the ultimate benchmark remains largely unchanged. By advancing entrepreneurial, risky and calculating behaviour as a norm, the perfect rationality assumption re-enters the stage through the backdoor. Behavioural economists share the normative view that rational maximization is what people should do. The gap between both perspectives is therefore not nearly as wide as we are made to believe. Although there are differences within the subdiscipline to this end (see Simon 2008), behavioural economics can therefore be regarded as complementary to neoclassical thinking, and this surely not only in the realm of development (Datta and Mullainathan 2014, pp. 7‒8). Protagonists continue to conceptualize the poor as means-ends-oriented, weakening the assumption that they are all-knowing and perfect calculators only to some extent. A good example to illustrate this is the application of advanced technological inputs in agriculture such as fertilizer, a question that is a recurring theme in the academic and project-related literature (e.g. Duflo et al. 2011). ‘Non-rational’ smallholders (who stubbornly resist to use inorganic fertilizer) are to be transformed into active individuals who rationally calculate their choices (the benefits and costs of buying fertilizer) and – importantly – take full responsibility for the decisions taken (such as the uncertainty that fertilizer use may not have the desired result or even unwanted side-effects). In a context in which smallholders are fully aware what is expected from them, what it means to be reasonable and responsible, refusal to do so is a difficult decision indeed. Displacing the

M4716-STRAßHEIM_9781785367847_t.indd 249

25/04/2019 13:32

250  Handbook of behavioural change and public policy market with the individual market subject as target of policy interventions therefore strengthens the far-reaching normative aspirations of the orthodox economic project. In so doing the emerging behavioural mainstream actually provides a means to stabilize the orthodoxy during turbulent times, translating it into a utopian yardstick to measure concrete economic behaviour and as a behavioural norm performing economic realities. One could even go a step further and argue that behavioural economics is an integral part of the economic mainstream, helping it to maintain its resilience in turbulent times.5 This resilience owes a lot to the ability to internalize the repeated challenges against its integrity. In his ‘governmentality lectures’ Michel Foucault pointed to a decisive shift associated with the emergence of neoliberal rationalities in this context. The individual was no longer addressed as a partner in exchange, but as entrepreneurial subject. In so doing, the homo economicus turned into the ‘entrepreneur of himself’, as Foucault (2008, p. 226) famously put it. The ‘neoliberal’ homo economicus exerts its normalizing force indirectly as a normative benchmark whereby to classify behaviour across a population as ‘economic’ and ‘non-economic’ (or normal and not normal). In so doing, the idea of economic behaviour as optimal allocation of scarce resources to alternative ends turned into a norm that exerts a subtler force, capable of governing individual conduct indirectly (Newheiser 2016, p. 6). The recent emergence of a reformulated behavioural approach in economics, and the methodological challenge of experimentalism accompanying it, constitutes a further step in this ongoing transformation. On the one hand, it may be interpreted almost as a step back towards a disciplinary attention to the individual subject. This appears to be particularly marked in the context of rural poverty in the Global South where we ultimately observe a more direct interventionist logic that aims at bringing individual behaviour in line with a market logic. On the other hand, the encounter between economic behaviourism and cognitive psychology resulted in a further rearticulation of the notion of economic man or woman. This includes the gradual transformation of subjects in the Global North and increasingly also in the Global South into human capital under the gaze of socio-technologies such as human resource management (Brown 2015, p. 32; see also Berndt 2013). In the wake of this rearticulation homo economicus has undoubtedly become more psychologically complex (Hargreaves-Heap 2008, p. 3). However, there is more to this: The ‘new’ homo economicus no longer pretends to have autonomous sovereignty. Having no agency of her own, she is the relational effect of distributed cognitive and calculative processes. And here the already mentioned market devices, the various nudges and experimental templates, play a crucial role. 4.3  Imaginations of Progress and Modernity The increasing influence of behaviourist and experimentalist thought is also linked to the rhetorical connection of geographical imaginations of backwardness and marginality with market failure. This can be illustrated with the various policy scripts and templates that have given form to market-oriented development projects across the Global South (e.g. Making Markets Work for the Poor – M4P; Value Chain Development – VCD). A M4P policy document on the maize market in Bangladesh (Gibson 2006), for instance, sets out to find answers to the question of why the Bangladeshi maize market fails the poor. The author, a development consultant, starts with a number of systemic constraints

M4716-STRAßHEIM_9781785367847_t.indd 250

25/04/2019 13:32

Behavioural economics and development policy  251 against more productive agricultural practices, for instance, fragmented landownership, poor physical infrastructure, weak transportation, high wastage rates, or inadequate financial services (Gibson 2006, p. 19). However, according to the study these are only the ‘immediate causes’, underlying these problems are more far-reaching ‘systemic’ ones: Market players are found to be ‘especially slow to learn about new ideas and opportunities’, traditional structures such as informal institutions and prevalent social norms are seen as being ‘insufficient in an era of high intensity farming’, and government support is regarded as ineffective (Gibson 2006, pp. 21‒23; see also M4P 2008, p. 12). Traditional ownership patterns are regularly represented as a major disincentive to investment and risk-taking (Ferrand et al. 2004, p. 3). These representations are telling examples of how individual behaviour is problematized as the main cause of poverty and underdevelopment, transforming this problem into an opportunity for all sorts of behavioural engineers and their paternalist programmes (see Roy 2010, p. 23). The behavioural and experimental turn in development is driven by a desire to transform poor risk-averse smallholders, trapped in traditional agricultural practices, into ‘responsibilized’ entrepreneurial farmers who readily take risks and never shy away from adopting the latest technology (Sheppard and Leitner 2010, p. 190). In what is a variant of well-known imaginations of linear progress and modernization, the well-functioning, efficient market is the benchmark against which the ‘underdeveloped’, poor South is given shape. There is a tendency in the literature to mobilize classical cultural stereotypes and tropes of modernization when turning to the external environment. Participants in the debate that are more closely linked with development practice are particularly outspoken in this regard. A WDR 2015 concept note mobilizes ‘mental models rooted in particular cultures’ (World Bank 2014). Other documents draw a clear line between modern market economies in the North and more traditional arrangements in the South. ‘Rationality violations are less prevalent in an environment that more closely resembles the market-type of setting that is the basis of neoclassical models’, argues Cecchi (2010, p. 5). And according to Anderson and Stamoulis (2006, p. 17) it is a lack of exposure to the discipline of the market that sees to it that ‘behavioural anomalies may be even less anomalous (i.e. they may be much closer to the norm) than observed in the USA and Europe’ (Anderson and Stamoulis 2006, p. 17). Translated into the relatively simple world of policy implementation, this creates a dual world all too familiar to those who have engaged with classical modernization theories: On the one side are ‘the poor’, reduced to ‘indigenous’, ‘local’ and ‘traditional’ knowledge populating a world characterized by small-scale and traditional agriculture. On the other side we have ‘the non-poor’ trained and educated, involved in large-scale production using sophisticated production methods. On the one side are poor peasants and smallholders, on the other entrepreneurial farmers (Swiss Agency for Development and Cooperation 2008, p. 39; Ferrand et al. 2004, p. 10). ‘The very poor engage with markets in less structured and informal ways’, argue the authors of a USAID discussion paper (Fowler and Brand 2011, p. 3), contrasting this with the less poor ‘where greater points of leverage exist in more structured and formal opportunities’. By positing ‘risk aversion’ as a universal character trait of smallholders, heterogeneous economic practices are reduced to a single explanatory variable (Watts 2013, p. 16). In order for this representation to work there has to be a strict distinction between a traditional self-sufficient world of subsistence and a dynamic modern realm of market

M4716-STRAßHEIM_9781785367847_t.indd 251

25/04/2019 13:32

252  Handbook of behavioural change and public policy exchange. Their deep and necessary entanglement notwithstanding, the construction and subsequent reproduction of ‘two apparently contradictory but unspecified production systems’ (Watts 2013, p. 15) and the deeply asymmetric way these two systems are valued provides the ground on which the practitioners of behavioural engineering apply their paternalist therapies to the behavioural ills afflicting the rural poor.6 4.4  Ruling Experts: Rendering Development a Technical Problem Policy practitioners and academic economists take issue with the tendency of both market-fundamentalist and state interventionist approaches to focus on the ‘big questions’. It is against this that the case for a seemingly less ambitious, micro-level and technocratic approach acquires its legitimization (Banerjee and Duflo 2011, pp. 3, 9), thereby rendering development technical, that is, ‘extracting from the messiness of the social world, with all the processes that run through it, a set of relations that can be formulated as a diagram in which problem (a) plus intervention (b) will produce (c), a beneficial result’ (Li 2007, p. 265). Wider societal issues are translated into technical problems that can be corrected with the help of behavioural engineering and experimentalism (see also Reddy 2012). In so doing, the protagonists of ‘development design’ programmes take part in the emergence of a politics of ‘forced choice’ that erases any notion of the spatial and social contexts of decision-making. A case in point is a recent paper by Johannes Haushofer and Ernst Fehr. Three potential ways to escape from the poverty trap are identified: ‘The first is to target poverty directly, the second is to target its psychological consequences, and the third is to target the economic behaviours that result from them’ (Haushofer and Fehr 2014, p. 866). Yet the authors subsequently deal only with possibilities 2 and 3, the question of the underlying causes of poverty is not even considered in passing. Every notion of the wider social and geographical conditions of poverty is erased. In the context of attempts to integrate people and places into global markets and value chains, this concerns the relative position in networks of production, distribution and consumption. It is obvious that some person’s or group’s articulation with global markets may be another’s disarticulation and marginalization. This starts with the carefully constructed settings in which the nudges of behavioural engineering acquire their particular force, or with randomized experiments arbitrarily separating those who are treated from those who are not (for recent critiques see Bédécarrats et al. 2015; Morvant-Roux et al. 2014; Webber 2015). And it ends with the caprices of global value chains and markets, in one moment providing opportunities of linking and upgrading, in other moments unceremoniously severing ties with consumers in the Global North or in the urban centres of the Global South. The technocratic rationale and the neglect of wider contextual circumstances may at least partly explain that there is also often little softness or ‘liberty’ in the paternalism prescribed for those who are subject to these interventions. We are confronted with only thinly-veiled pretensions to engage in outright ‘behavioural engineering’ (Bolton and Ockenfels 2012), and the ‘breaking’ of inefficient habits or (OECD 2012, p. 45). This is a reminder that markets and market subjects have to be actively produced in a highly selective process in which ‘market forces are imposed on some but not others’ (Hall et al. 2013, p. 14) and that marketization constitutes individualized rational actors against a backdrop of collective failure.

M4716-STRAßHEIM_9781785367847_t.indd 252

25/04/2019 13:32

Behavioural economics and development policy  253

5. CONCLUSION Behavioural economics has made an impressive career during the last decades, at first moving from a marginal position at the crossroads of psychology and economics into the economic mainstream and more recently transforming into a powerful policy script that has profoundly rearticulated what may be called neoliberal marketization. In this chapter I focused on development policy and framed the double behavioural and experimental turn in this field as a particular conjuncture that is characterized by a readjusted assemblage of economic knowledge (e.g. about the market, about the ‘southern poor’, about the role of the state), social technologies (e.g. experiments, design of choice architectures, evidence-based policies) and material devices (e.g. concrete nudges, game sheets, evaluation metrics). While certainly not negating the potentially positive way in which these ideas may be put to use, I wrote this chapter from a sceptical perspective, above all concerning the translation of behavioural thinking into the policy realm. There clearly are wider implications that go beyond the sphere of development in the Global South. In the context of development failure is associated with ‘the poor smallholder’ or the ‘urban “slum” dweller’. In different policy contexts this role may be played by ‘single mothers’, ‘welfare recipients’ and so on. It is crucial therefore to look carefully how these processes interact with registers of social difference, such as gender, ethnicity or class.

NOTES 1. The developments in cognitive psychology that revitalized behavioural economics were explicitly directed against behaviourism in the tradition of Watson and Skinner, although there were parallels with the ‘behavioralist or engineering approach of twentieth-century U.S. psychology’ (Heukelom 2014, p. 97). As I will explain in more detail below, behaviourist impulses have become more strongly visible in the wake of the translation of behavioural economic ideas into the realm of social policy. However, instances such as the uncritical treatment of ‘priming’ (an intervention that steers behaviour with the help of stimuli of which the targeted subjects are not aware) in the influential books by Thaler and Sunstein (2008, pp. 69‒70) and Kahneman (2011, pp. 52‒58) are an indication that the distance to behaviourism is not always as large as assumed. 2. Taking context into account in order to meaningfully understand the world is what geographers, sociologists, anthropologists and qualitative social science more generally have been doing for a century. As will be discussed more carefully in the final section of this chapter, what is mobilized is a very truncated notion of socio-spatial context shorn of any idea of inequality and power asymmetries. 3. Experimental economics emerged during the 1950s as a combination of the experimental method used in psychology and new advances in economic theory, above all the rise of game theory that defined new standards for mathematical rigor in economics. Behavioural economists similarly adopted experiments in dialogue with insights from psychological research as key methodological instruments to test the extent to which actual behaviour deviates from the perfect rationality assumption. This reinvigorated the experimental tradition in economics that was equally sidelined when economics established itself as a strictly non-empirical, deductive discipline in the course of the 1950s and 1960s. Although one should not overlook the differences, for instance with a view to the exact role of experiments in policy interventions (e.g. they may be used to evaluate a policy measure or to change behaviour directly) it is justified in the light of this entangled history to speak of a broader intellectual project that emerges as a serious challenger to standard economic theory (Guala 2008; Leonard 2008, pp. 7‒8; Simon 2008, p. 612). 4. Geographer Gillian Hart (2001) made the useful distinction between ‘big D’ Development and ‘little d’ development, defining the former ‘as a post-second world war project of intervention in the “third world” that emerged in the context of decolonization and the cold war’ and contrasting the notion with the development of capitalism more generally (‘little d’ development) ‘as a geographically uneven, profoundly contradictory set of historical processes’.

M4716-STRAßHEIM_9781785367847_t.indd 253

25/04/2019 13:32

254  Handbook of behavioural change and public policy 5. The term ‘neoclassical economics’ is less a self-description emerging from the inside of what was to become the economic mainstream than a label coined by sceptical scholars such as Thorstein Veblen who pointed to the differences between ‘historical and Marxist schools’ and ‘modernized classical’, that is, ‘neo-classical’ views (Veblen 1900, p. 261). What exactly is part of neoclassical economics and what is not has always been a matter of dispute. Important differences notwithstanding, what unified the economic mainstream has been a belief in marginalism, methodological individualism, opportunity costs and the virtuous effect of market exchange (see, for instance, Aspromourgos 2008; Brennan and Moehler 2010). 6. This is not to say that the script of behavioural programming is inscribed on docile peasant bodies, passively performing the subjectivities it is naming. Nudging and experimenting are never complete and always prone to failure. They may be readily adopted, they may meet outright resistance, and there may be instances of what James Scott has termed ‘calculated conformity’ (Scott 1985). These are important questions that can only be answered empirically and go beyond the scope of this chapter.

REFERENCES Anand, Paul and Stephen Lea (2011), ‘The psychology and behavioural economics of poverty’, Journal of Economic Psychology, 32 (2), 284–293. Anderson, C. Leigh and Kostas Stamoulis (2006), Applying behavioural economics to international development policy. United Nations University, World Institute for Development Economics Research (UNU-WIDER), Research Paper No. 2006/24. Aspromourgos, Tony (2008), ‘Neoclassical’, in Steven N. Durlauf and Lawrence E. Blume (eds), The New Palgrave Dictionary of Economics, Basingstoke: Palgrave Macmillan (online version). Banerjee, Abhijit (2001), ‘The two poverties’, Nordic Journal of Political Economy, 26, 129‒141. Banerjee, Abhijit and Esther Duflo (2011), Poor Economics: A Radical Rethinking of the Way to Fight Global Poverty, New York: Public Affairs. Berndt, Christian (2013), ‘Assembling market b/orders: violence, dispossession, and economic development in Ciudad Juárez, Mexico’, Environment and Planning A, 45 (11), 2646‒2662. Berndt, Christian and Marc Boeckler (2012), ‘Geographies of marketization’, in Trevor J. Barnes, Jamie Peck and Eric Sheppard (eds), The New Companion to Economic Geography, Oxford: Wiley-Blackwell, pp. 199‒212. Bédécarrats, Florent, Guérin, Isabelle and François Roubaud (2015), The gold standard for randomised evaluations: from discussion of method to political economy, Paris: UMR 225 IRD – Paris-Dauphine , Working Paper WP 2015-01. Bolton, Gary E. and Axel Ockenfels (2012), ‘Behavioral economic engineering’, Journal of Economic Psychology, 33 (3): 665‒676. Brennan, Geoffrey and Michael Moehler (2010), ‘Neoclassical economics’, in Mark Bevir (ed.), Encyclopedia of Political Theory, Thousand Oaks: Sage, pp. 947‒951. Brown, Wendy (2015), Undoing the Demos: Neoliberalism’s Stealth Revolution, New York: Zone Books. Camerer, Colin, Issacharoff, Samuel, Loewenstein, George, O’Donoghue, Ted and Matthew Rabin (2003), ‘Regulation for conservatives: behavioral economics and the case for “asymmetric paternalism”’, University of Pennsylvania Law Review, 151, 1211‒1254. Cecchi, Francesco (2010), Does Market Experience Shape Behaviour: Experimental Evidence From Rural Ethiopia, Wageningen: Ministry of Foreign Affairs and Wageningen UR. Datta, Saugato and Sendhil Mullainathan, S. (2014), ‘Behavioral design: a new approach to development policy’, Review of Income and Wealth, 60 (1), 7‒35. de Soto, Hernando (2000), The Mystery of Capital: Why Capitalism Triumphs in the West and Fails Everywhere Else, New York: Basic Books. Deaton, Angus (2010), ‘Instruments, randomization, and learning about development’, Journal of Economic Literature, 48 (2): 424–455. Duflo, Esther (2005), Field Experiments in Development Economics, BREAD Policy Paper No. 012, December 2005, Massachusetts Institute of Technology. (ipl.econ.duke.edu/bread/papers/policy/p012.pdf; 5 May 2012). Duflo, Esther (2006), ‘Poor but rational?’, in Abhijit V. Banerjee, Roland Benabou and Dilip Mookherjee (eds), Understanding Poverty, Oxford and New York: Oxford University Press, pp. 367‒378. Duflo, Esther, Kremer, Michael and Jonathan Robinson (2011), ‘Nudging farmers to use fertilizer: theory and experimental evidence from Kenya’, American Economic Review, 101 (6): 2350‒2390. Fafchamps, Marcel (2009), ‘Vulnerability, risk management, and agricultural development’, paper presented at the AERC Conference on Agriculture and Development, Mombasa, Kenya, 28–30 May. Ferrand, David, Gibson, Alan and Hugh Scott (2004), ‘Making Markets Work for the Poor’: An Objective and an Approach for Governments and Development Agencies, Woodmead: ComMark Trust.

M4716-STRAßHEIM_9781785367847_t.indd 254

25/04/2019 13:32

Behavioural economics and development policy  255 Foucault, Michel (2008), The Birth of Biopolitics: Lectures at the Collège de France, 1978–79, Basingstoke: Palgrave Macmillan. Fowler, Ben and Margie Brand (2011), Pathways out of Poverty: Applying Key Principles of the Value Chain Approach to Reach the Very Poor, USAID Discussion Paper microReport 173. Gibson, Alan (2006), Enhancing the Supply-side of the Maize Market, Durham: The Springfield Centre for Business in Development. Guala, Francesco (2005), The Methodology of Experimental Economics, New York: Cambridge University Press. Guala, Francesco (2008), ‘Experimental economics, history of’, in Steven N. Durlauf and Lawrence E. Blume (eds), The New Palgrave Dictionary of Economics, Basingstoke: Palgrave Macmillan (online version). Hall, Stuart, Massey, Doreen and Michael Rustin (2013), ‘After neoliberalism: analysing the present’, Soundings, 53, 8‒22. Hargreaves-Heap, Shaun (2008), ‘Economic man’, in Steven N. Durlauf and Lawrence E. Blume (eds), The New Palgrave Dictionary of Economics, Basingstoke: Palgrave Macmillan (online version). Hart, Gilian (2001), ‘Development critiques in the 1990s: culs de sac and promising paths’, Progress in Human Geography, 25 (4), 649‒658. Haushofer, Johannes and Ernst Fehr (2014), ‘On the psychology of poverty’, Science, 344 (6186), 862‒867. Heukelom, Floris (2014), Behavioral Economics: A History, Cambridge: University of Cambridge Press. Kahneman, Daniel (2002), Daniel Kahneman – Autobiography, Stockholm: The Royal Swedish Academy of Science. Kahneman, Daniel (2011), Thinking, Fast and Slow, New York: Farrar, Straus and Giroux. Karlan, Dean and Jacob Appel (2011), More than Good Intentions: How a New Economics is Helping to Solve Global Poverty, New York: Dutton. Leonard, Robert (2008), ‘Game theory in economics, origins of’, in Steven N. Durlauf and Lawrence E. Blume (eds), The New Palgrave Dictionary of Economics, Basingstoke: Palgrave Macmillan (online version). Li, Tania Murray (2007), ‘Practices of assemblage and community forest management’, Economy and Society, 36 (2): 263‒293. Lourenço, Joana Sousa, Ciriolo, Emanuele, Rodrigues Vieira de Almeida, Rafael and Troussard, Xavier (2016), Behavioural Insights Applied to Policy, European Report 2016, Brussels: European Commission. M4P (2008), Making Value Chains Work Better for the Poor: A Toolbook for Practitioners of Value Chain Analysis, Version 3. Making Markets Work Better for the Poor (M4P) Project, UK Department for International Development (DFID), Phnom Pehn: Agricultural Development International. Mitchell, Timothy (2005), ‘The work of economics: how a discipline makes its world’, European Journal of Sociology, 45 (2): 297‒320. Morvant-Roux, Solène, Guérin, Isabelle, Roesch, Marc and Jean-Yves Moisseron (2014), ‘Adding value to randomization with qualitative analysis: the case of microcredit in rural Morocco’, World Development, 56, 302‒312. Mullainathan, Sendhil (2005), ‘Development economics through the lens of psychology’, in: François Bourguignon and Boris Pleskovic (eds), Annual World Bank Conference in Development Economics 2005: Lessons of Experience, New York: The World Bank and Oxford University Press, pp. 45‒70. Mullainathan, Sendhil and Eldar Shafir (2013), Scarcity: Why Having Too Little Means So Much, New York: Times Book. Mullainathan, Sendhil and Richard H. Thaler (2001), ‘Behavioral economics’, in: Neil J. Smelser and Paul B. Baltes (eds), International Encyclopedia of the Social & Behavioral Sciences, Oxford: Pergamon, pp. 1094‒1100. Newheiser, David (2016), ‘Foucault, Gary Becker and the critique of neoliberalism’, Theory, Culture & Society, 33 (5), 3‒21. OECD (2012), Farmer Behaviour, Agricultural Management and Climate Change, Paris: Organisation for Economic Co-operation and Development. Peck, Jamie and Nik Theodore (2015), Fast Policy: Experimental Statecraft at the Thresholds of Neoliberalism, Minneapolis: University of Minnesota Press. Reddy, Sanjay G. (2012), ‘Randomise this! On poor economics’, Review of Agrarian Studies, 2 (2), 60–73. Roy, Ananya (2010), Poverty Capital: Microfinance and the Making of Development, New York: Routledge. Schultz, Theodor W. (1979), Theodore W. Schultz – Prize Lecture: The Economics of Being Poor, Stockholm: The Royal Swedish Academy of Science. Scott, James C. (1985), Weapons of the Weak: Everyday Forms of Peasant Resistance, New Haven: Yale University Press. Sheppard, Eric and Helga Leitner (2010), ‘Quo vadis neoliberalism? The remaking of global capitalist governance after the Washington Consensus’, Geoforum, 41 (2), 185–194. Simon, Herbert (1978), ‘Herbert A. Simon – Prize Lecture: Rational decision-making in business organizations’, The Royal Swedish Academy of Science, Stockholm. Simon, Herbert (2008), ‘Behavioural economics’, in Steven N. Durlauf and Lawrence E. Blume (eds), The New Palgrave Dictionary of Economics, Basingstoke: Palgrave Macmillan (online version).

M4716-STRAßHEIM_9781785367847_t.indd 255

25/04/2019 13:32

256  Handbook of behavioural change and public policy Swiss Agency for Development and Cooperation (2008), Perspectives on the Making Markets Work for the Poor (M4P) approach, Bern: Swiss Agency for Development and Cooperation. Thaler, Richard (1980), ‘Toward a positive theory of consumer choice’, Journal of Economic Behavior & Organization, 1 (1), 39‒60. Thaler, Richard H. and Cass R. Sunstein (2008), Nudge: Improving Decisions About Health, Wealth, And Happiness, New Haven: Yale University Press. Tversky, Amos and Daniel Kahneman (1974), ‘Judgment under uncertainty: heuristics and biases’, Science, 185 (4157), 1124‒1131. Veblen, Thorstein (1900), ‘The preconceptions of economic science – part 3’, The Quarterly Journal of Economics, 14 (2), 240–269. Watts, Michael J. (2013), Silent Violence: Food, Famine, and Peasantry in Northern Nigeria, Athens, USA: University of Georgia Press. Webber, Sophie (2015), ‘Randomising development: geography, economics and the search for scientific rigour’, Tijdschrift voor economische en sociale geografie, 106 (1), 36‒52. World Bank (2014), WDR 2015: Mind and Society – Concept Note. Retrieved from http://econ.worldbank.org/ WBSITE/EXTERNAL/EXTDEC/EXTRESEARCH/EXTWDRS/ETNWDR2013/0,,contentMDK:23488 439~pagePK:8261309~piPK:8258028~theSitePK:8258025,00.html (accessed 5 August 2014). World Bank (2015), World Development Report 2015: Mind, Society and Behavior, Washington: The World Bank.

M4716-STRAßHEIM_9781785367847_t.indd 256

25/04/2019 13:32

18.  Steering the behaviour of young people: the EU’s policy approach to promote employment Jale Tosun and Felix Hörisch

1. INTRODUCTION The question of how to steer the behaviour of individuals to attain a given goal has been at the heart of the public policy literature on policy design for several years now (see, e.g. Schneider and Ingram 1990). The corresponding literature describes and examines the choice and effects of policy instruments (see, e.g. Howlett 1991, 2014). In the last few years, the concept of Libertarian Paternalism – a behavioural economicsinformed policy concept closely associated with the work of Thaler and Sunstein (2003, 2008) – has captured considerable attention. Thaler and Sunstein (2003, p. 179) define libertarian paternalism as ‘an approach that preserves freedom of choice but that authorises both private and public institutions to steer people in directions that will promote their welfare’. While the option to affirmatively choose to behave otherwise is the ‘libertarian’ aspect of the definition, the steering of people in certain directions is the ‘paternalistic’ aspect thereof. This approach focuses on ‘choice architecture’, which is the ways in which the behaviours of individuals are ‘nudged’ in particular – typically welfare promoting – directions (Brown 2012, p. 306). Thaler and Sunstein (2008) propose six methods for devising a choice architecture that can potentially promote welfare: incentives, understanding mapping, defaults, feedback, expecting errors and the structuring of complex choices. Thaler and Sunstein (2003, 2008) discuss nudging as a method by which choices can be designed and presented in a manner that will help individuals act in their own interest without restricting their choices. Of course, there are also other scholars that hold a more critical position on nudging and libertarian paternalism, regarding them as a means for governments to introduce behaviour-change interventions that might be hidden for those towards whom they are targeted (Oliver 2015, p. 712). From this, it follows that there exists a fundamental difference between policy instruments that draw on libertarian paternalism and ones those are deliberative; the latter arguing that those whose behaviour will be influenced should be aware of such influence. This distinction is also recognized and discussed in detail by Grüne-Yanoff and Hertwig (2016). According to these authors, government policy can consist of ‘established’ tools to bring about behaviour change such as regulation and (financial) incentives or disincentives (so-called boost policies) or ‘new’ tools that draw on the insights yielded by psychology and behavioural economics on the drivers of behaviour change (so-called nudge policies). These two fundamentally different policy approaches also have diverging assumptions in their analytical frameworks regarding behaviour change. Boost policies are rooted in the ‘simple heuristics’ research programme, associated with the work of Gigerenzer and various collaborators (see, e.g. Hutchinson and Gigerenzer 257

M4716-STRAßHEIM_9781785367847_t.indd 257

25/04/2019 13:32

258  Handbook of behavioural change and public policy 2005). This research programme postulates that policies should strive to boost people’s decision-making competences by targeting their skills and knowledge, their feasible set of decision tools, and/or by structuring their decision-making environment. The research programme underlying the nudging policies corresponds to that of ‘heuristics and biases’ as put forth by the work of Kahneman, Tversky and their various collaborators (see, e.g. Kahneman et al. 1982). Drawing on this research programme, policies aiming to bring about behaviour change should capitalize on co-opting systematic biases and thereby ‘nudge individual behaviour toward a different, more beneficial outcome’ (Grüne-Yanoff and Hertwig 2016, p. 152). While acknowledging these two different policy approaches to bring about behaviour change, in this chapter we concentrate on deliberative strategies – or ‘boost policies’ to pick up the terminology used by Grüne-Yanoff and Hertwig (2016) – adopted primarily by policy-makers. Our decision is motivated by the prevalence of both instruments in policy-making. While in the last few years nudging policies have increasingly been adopted (see, e.g. Thies et al. 2018) – particularly in the UK (see, e.g. John et al. 2009) – the predominant type of policy instruments used still corresponds to boost policies. This is the reason we consider it a worthwhile attempt to improve our knowledge on the designs and effects of these policy tools. While the attempt to change the behaviour of individuals is inherent to any public policy, not all policy instruments chosen are effective for increasing the consistency between the actual behaviour and the behaviour preferred by policy-makers (Weaver 2015). Especially in situations where the solution to a policy problem depends not only on the behaviour changes of one target group, but on changes in the behaviour of multiple target groups, choosing the correct policy response can be challenging for decision-makers. In this contribution we examine an issue that requires the adoption of policy instruments for multiple target groups: youth unemployment. The selection of this problem area is motivated by the recent episode of youth unemployment in the European Union (EU). Youth unemployment – like unemployment in general – has many undesirable economic, social and political consequences for individuals and societies alike. It is associated with a multitude of economic challenges, a general lack of social support, and often leaves young people more vulnerable to physical and mental health problems (see, e.g. Brandt and Hank 2014). Here, we are interested in the policy instruments adopted by EU policy-makers in response to this problem. We compare the perceptions of young people in Germany (aged 18‒35) on the causes of unemployment and their dispositions to change their behaviour to gain employment. We chose Germany because it is the European country regarded a leader in promoting youth employment (Shore and Tosun 2019), and has a vocational education and training system in place that is widely perceived to be successful and worth to be adopted (Wieland 2015). In this way, we conduct a preliminary assessment of the fit between the policy instruments chosen and the perceptions of their target groups. Concentrating on the views of young people is appropriate since despite the multitude of target groups that should be addressed by policies aiming to combat youth unemployment, the EU’s approach mostly – albeit not exclusively – concerns the supply-side and therefore, seeks to change the skills and behaviour of young people themselves (Lahusen et al. 2013; Thies et al. 2018).

M4716-STRAßHEIM_9781785367847_t.indd 258

25/04/2019 13:32

Steering the behaviour of young people  259 What are the main perceived barriers to entering and staying in employment among young Germans? How well suited are the EU level policy measures in overcoming these barriers? These are the two research questions guiding this chapter. The remainder of the chapter is organized as follows: In the next section, we provide an overview of the EU’s policy approach to promote youth employment since 1997. Subsequently, we present information based on empirical data from the CUPESSE survey1 on what young Germans between the ages of 18 and 35 believe to be the causes of youth unemployment and what they would be willing to do to get a (better) job. Then, we provide an assessment of the policy measures adopted at the EU level to bring young people into employment following the recommendations by Weaver (2015) on how to influence behavioural change by public policies. Finally, we conclude by summarizing the main findings on what the main perceived barriers to promoting ­employment among young Germans are and how well suited the EU level policy measures implemented by the German government are in overcoming these identified barriers.

2.  OVERVIEW OF THE EU’S POLICY APPROACH O’Reilly et al. (2015) argue that there are several characteristics that make the current episode of youth unemployment rather unique. The first is that labour market flexibility makes it difficult for young people to benefit from stable employment trajectories. Young people are usually among the first affected by downsizing and restructuring measures aiming to improve competitiveness. Second, according to the authors, there is a growing awareness regarding the problem of over-qualification and skills mismatch. The third characteristic refers to ‘more extensive, selective, and diversified’ youth migration within the EU. Fourth, and in line with Warmuth et al. (2015), O’Reilly et al. (2015) point to the importance of the role of families and the formation and transmission of employmentrelated values. The authors conclude that these characteristics of current youth unemployment have resulted in a series of policy measures reflecting the EU’s commitment to overcoming youth unemployment, which are summarized in Table 18.1. In contrast to ‘classic’ EU employment policy, which predominantly rests on the Open Method of Coordination (OMC) (see, e.g. Heidenreich and Bischoff 2008; Borrás and Radaelli 2015), over the period 2014–2020 a battery of EU-funded tools and harmonized policy instruments were and continue to be available for member states to stimulate youth employment (Tosun et al. 2016a; Tosun 2017; Tosun et al. 2017). The Luxembourg Job Summit in November 1997 can be regarded as the starting point of the EU youth employment policy. This was when the European Employment Strategy (EES) together with the OMC were adopted, which established the so-called Luxembourg process (1997‒2004): an annual coordinating and monitoring cycle for national employment policies based on the member states’ commitment to establishing a set of common objectives and targets (European Parliament 2016). In parallel to the Luxembourg process, the Lisbon Strategy was in place between 2000 and 2010 (for an overview, see Borrás and Radaelli 2011). In its context, the EES was reviewed in 2002 and re-launched in 2005. Revisions included the introduction of a

M4716-STRAßHEIM_9781785367847_t.indd 259

25/04/2019 13:32

260  Handbook of behavioural change and public policy Table 18.1  Overview on EU’s policy approach to foster youth employment Event / Policy programme

Policy measures

Luxembourg Job Summit (1997)

Adoption of European Employment Strategy (EES) as well as the Open Method of Coordination (OMC) Inclusion of a title on employment in the EU treaties Revisions of the EES including the introduction of a multiannual time framework and the integration of the employment guidelines into the broad economic policy guidelines. First time that young people were mentioned as target group for employment policy at EU level Promotion of mobility, non-formal learning, intercultural dialogue and the inclusion of all young people, regardless of their educational, social and cultural background Sets several targets, including the increase of labour market participation of people aged 20 to 64 to 75% by 2020 and reducing the proportion of early school leavers to 10% (from 15%), and increase the share of 30‒34-yearolds having completed tertiary or equivalent education to at least 40% (instead of 31%) Initiative for improving the prospects of young people in finding a job Promoting the creation of jobs by stimulating demand for labour; restoration of the dynamics of labour markets, including the reform of the labour markets by promoting internal flexibility, expanding lifelong learning and active labour market policies; more effective control at the EU level through stronger coordination and multilateral surveillance of employment policies Calls on the member states to adopt measures which ensure that young people get a ‘good quality’ offer for a job, an apprenticeship, a traineeship, or continued education within four months of them leaving education or becoming unemployed Fostering youth employment especially in regions where youth unemployment rate exceeds 25%

Treaty of Amsterdam (1997) Lisbon Strategy (in place between   2000 and 2010)

Youth in Action (2007) Europe 2020 strategy (2010)

‘Youth on the Move’-initiative (2010) Employment Package (2012)

Youth Guarantee (2013)

Youth Employment Initiative (2013)

Source:  Own overview.

multiannual time framework and the integration of employment guidelines into broad economic policy guidelines (European Parliament 2016). Most importantly, the updated EES for the first time declared young people and older low-skilled workers as target groups. This resulted in the launch of the programme ‘Youth in Action’ in 2007, which promoted mobility, informal learning, intercultural dialogue and the inclusion of all young people, regardless of their educational, social or cultural background. Given that the EU employment policy was long characterized by general policy recommendations, it is noteworthy that these actions for youth unemployment are concrete. Moreover, they are significantly financially supported by EU funds. In February 2012,

M4716-STRAßHEIM_9781785367847_t.indd 260

25/04/2019 13:32

Steering the behaviour of young people  261 unused funds from the EU Structural Funds were mobilized for the programme period between 2007 and 2013 in order to improve the employment situation of young people in the member states with the highest youth unemployment levels at the time. The next milestone came in the form of the passage of the Europe 2020 strategy in 2010 (2010‒2020; for an overview, see Chung et al. 2012). This strategy sets several targets to be completed by 2020, including the increase of labour market participation of people aged 20 to 64 to 75 per cent; to lift at least 20 million people out of the risk of poverty and exclusion; and – most importantly for our research interests – reducing the proportion of early school leavers to 10 per cent (from 15 per cent), and increase the share of 30‒34-year-olds having completed a tertiary education or its equivalent to at least 40 per cent (instead of 31 per cent; European Parliament 2016). An outcome of this strategy was the adoption of the ‘Youth on the Move’ initiative in 2010, which for the first time dealt exclusively with the prospects of young people for finding a job. One component of this effort was the Youth Opportunities initiative, which from 2012 to 2013 took advantage of the financial means offered by the ESS to improve vocational training opportunities and to promote youth entrepreneurs and social entrepreneurship. The second component contains actions to promote the labour market mobility of young people, for example, by means of the job mobility portal EURES (EURopean Employment Services), which supports young people in finding employment, training or internships in other EU countries. In April 2012 the European Commission launched a number of employment-promoting measures to combat the soaring unemployment rate: the so-called Employment Package. The measures address three main areas: First, promoting creation of jobs by stimulating demand for labour, increasing the potential of job-intensive sectors, and providing funds from the ESS. Second, restoring the dynamics of labour markets, including reforming labour markets by promoting internal flexibility, expanding lifelong learning and active labour market policies, creating opportunities, and presenting special employment packages for young people. Finally, the package seeks more effective control at the EU level through stronger coordination and multilateral surveillance of employment policies, effective participation of the social partners and the use of relevant financial instruments (European Parliament 2016). Launched in 2010 by the European Parliament, but only officially endorsed by the Council of the EU in June 2013, the Youth Guarantee marks a major milestone in the EU’s political commitment to support young people in overcoming the obstacles with which they are confronted on their journey from education to employment (Tosun 2017). The Youth Guarantee calls on the member states to adopt measures to ensure young people get a ‘good quality’ offer for a job, an apprenticeship, a traineeship or continued education within four months of leaving education or becoming unemployed. As with any EU policy measure, the member states are responsible for the implementation of the Youth Guarantee, and this can lead to variation in the policy designs adopted and the institutional arrangements created (Tosun 2017). As a policy measure, the Youth Employment Initiative (YEI) was adopted by the European Council on 7‒8 February 2013 (European Commission 2015a). The EU heads of states agreed to set aside about 6.4 billion Euro for the period 2014‒2020 for regions where youth unemployment rates exceed 25 per cent. 3.2 billion euro of the YEI budget comes from a specific EU budget line dedicated to youth employment, while the other half

M4716-STRAßHEIM_9781785367847_t.indd 261

25/04/2019 13:32

262  Handbook of behavioural change and public policy of the budget comes from the EU structural and investment funds, national allocations (European Commission 2015b, p. 2). The YEI’s target group are young people not in employment, education or training (NEETs) aged below 25 years, and if the respective member state considers it relevant, also those aged below 30 years (European Commission 2015a). All in all, the EU’s youth employment policy has developed in a rather remarkable fashion. While initially the member states had sole legislative competence in this area, the policy measures adopted over the last few years indicate that the member states have transferred legal competences to the EU level. In return for this, policy measures are accompanied by funding mechanisms like the ESS, and more specifically, the YEI. The policy arsenal is constrained to employability, activation and labour market mobility. According to Lahusen et al. (2013, p. 304), this policy approach ‘takes labour markets as a given factor to which young people have to be adapted’ and establishes the ‘expectation that young people have to be flexible when education, training and jobs search are at stake’. In light of this policy approach, in the subsequent section, we will also concentrate on the perspectives of young people and discuss to what extent the policies adopted are likely to bring about behavioural change.

3.  PERCEPTIONS OF THE TARGET GROUP In this section, we present empirical information on, first, what young Germans between the ages of 18 and 35 believe to be the causes of the current episode of youth unemployment, and second, what they would be willing to do to get a (better) job. The perceived causes of youth unemployment are important to the extent that the EU’s policy approach mostly addresses three dimensions (employability, activation and labour market mobility) and for a policy instrument to bring about the intended changes in behaviour, the target group needs to accept the instrument (see, e.g. Comte et al. 2000). Even more directly related to behavioural changes is the second dimension of the empirical information we will present, as the questions asked explicitly address the respondents’ disposition to change their behaviour. For the purpose of this study, we use original data collected between February and April 2016 with the assistance of the polling firm YouGov Germany in the context of the CUPESSE project (Tosun et al. 2018). For the survey, we used YouGov Germany’s online panel and a turbo sampling method.2 Table 18.2 gives an overview of the level of agreement for seven statements regarding the causes of youth unemployment. The respondents could strongly/somewhat agree or disagree with these statements. When inspecting the table, we make several interesting observations. First, young people in Germany tend to somewhat or strongly disagree with the statement that the economic situation is the main reason for youth unemployment in the country (about 52 per cent); it is the only survey question with which more respondents disagreed than agreed. This brings us to the second observation: for all other response options, the majority of the panel members surveyed, somewhat agreed with the statements, although, there is some variation in how strong the difference between somewhat agreeing and disagreeing

M4716-STRAßHEIM_9781785367847_t.indd 262

25/04/2019 13:32

Steering the behaviour of young people  263 Table 18.2 Responses to the survey question: There are several reasons why some young adults find it difficult to get a job in Germany. Please indicate the extent to which you agree or disagree with these following statements Response categories

Absolute frequency

Per cent

Cumulated per cent

a.

Current economic situation Strongly disagree Somewhat disagree Somewhat agree Strongly agree

295 1407 1273 304

9.00 42.91 38.82 9.27

9.00 51.91 90.73 100.00

b.

Skills mismatch Strongly disagree Somewhat disagree Somewhat agree Strongly agree

102 578 1724 875

3.11 17.63 52.58 26.68

3.11 20.74 73.32 100.00

c.

Insufficient information Strongly disagree Somewhat disagree Somewhat agree Strongly agree

288 1154 1301 536

8.78 35.19 39.68 16.35

8.78 43.98 83.65 100.00

d.

No chance to demonstrate their skills and abilities 111 Strongly disagree 715 Somewhat disagree 1457 Somewhat agree Strongly agree 996

3.39 21.81 44.43 30.38

3.39 25.19 69.62 100.00

e.

No opportunity to get the training needed 155 Strongly disagree 938 Somewhat disagree 1497 Somewhat agree 689 Strongly agree

4.73 28.61 45.65 21.01

4.73 33.33 78.99 100.00

f.

No right careers advice before leaving school/university 120 Strongly disagree Somewhat disagree 716 Somewhat agree 1476 967 Strongly agree

3.66 21.84 45.01 29.49

3.66 25.50 70.51 100.00

g.

No right employment support Strongly disagree Somewhat disagree Somewhat agree Strongly agree

5.25 27.11 41.72 25.92

5.25 32.36 74.08 100.00

172 889 1368 850

Source:  www.cupesse.eu (N=3279); Tosun et al. (2016b).

M4716-STRAßHEIM_9781785367847_t.indd 263

25/04/2019 13:32

264  Handbook of behavioural change and public policy was. We observed the strongest support for the statement that young people’s skills do not match with what employers expect (response category b), followed by the statements that young job applicants are not given a chance to demonstrate their abilities (response category d), and that young people do not get good career advice before leaving school or university (response category f). By marked contrast, while most young people somewhat agreed that they do not get enough detailed information about job opportunities, agreement with this statement is the least strong (response category c). Finally, young Germans mostly agreed that there are no opportunities to procure the training needed to get a good job (response category e), and that people who are unemployed do not get the right employment support in finding a job (response category g). All in all, the majority of young Germans were found to agree with statements that are addressed by the policy approach of the EU. More specifically, the Youth Guarantee comprises a number of measures that directly map onto the response options and agreements of respondents. The Youth Guarantee is about improving the educational and vocational training of young people and better aligning their skills with employer demands (response categories b and e). It also calls on the member states to improve the quantity and quality of internships, which should help young jobseekers in demonstrating their skills and abilities (response category d). Furthermore, it seeks to provide better support to young jobseekers in terms of learning about job opportunities and finding a job (response categories c and g), a task fulfilled by public employment services in most member states (see Tosun 2017). However, what is not directly addressed by the Youth Guarantee is career counselling for young people before leaving school or university (response category f). Yet, it should be noted that career guidance is already provided in Germany. We now turn to the question about what young Germans would be willing to do to get a (better) job. For this question there are six response categories, presented in Table 18.3. This second empirical picture is much more nuanced than the first. We see that the majority of respondents are willing to move within Germany to get a (better) job (response category a), while only about 24 per cent indicate a willingness to move to a different country (response category b). Young Germans are particularly supportive of the idea to learn (completely) new skills (response categories c and d) if this helps them to find a (better) job. Finally, they may consider lowering their expectations about pay and working conditions, however, fewer respondents give a clear yes answer to response category e (yes: 18 per cent) than response category f (yes: 22 per cent). Summing up, respondents in Germany are willing to move within the country and to learn new skills or even get retrained in order to get a (better) job. The majority of the survey participants rule out moving to a different country. Concerning the respondents’ expectations regarding payment and working conditions, the empirical picture is less clear, since most of them indicated that they would maybe lower their expectations. Again, the response categories can be translated to EU level policy measures. Increasing young people’s mobility lies at the heart of the Youth on the Move programme and is expected to be stimulated by the online portal EURES (response categories a and b). Training and skills are targeted by several measures, but primarily by the Youth Guarantee (response categories c and d). The two questions on expectations are not addressed by the EU youth employment policies directly, but EU funds can be used to provide financial incentives to employers for hiring young people. Employment that is created by such subsidies tends

M4716-STRAßHEIM_9781785367847_t.indd 264

25/04/2019 13:32

Steering the behaviour of young people  265 Table 18.3 Responses to the survey question: Some people are willing to make big changes in their lives to get a job. What changes would you be willing to make to get a job or to get a new or a better job? Response categories

Absolute frequency

Per cent

Cumulated per cent

a. Move within country No Maybe Yes

689 1259 1331

21.01 38.40 40.59

21.01 59.41 100.00

b. Move to different country No Maybe Yes

1298 1178 803

39.59 35.93 24.49

39.59 75.51 100.00

c. Learn new skills (e.g. foreign language) No 107 Maybe 682 2490 Yes

3.26 20.80 75.94

3.26 24.06 100.00

d. Learn completely new skills/retrain No Maybe Yes

253 1081 1945

7.72 32.97 59.32

7.72 40.68 100.00

e. Lower expectations about earnings No Maybe Yes

1115 1588 576

34.00 48.43 17.57

34.00 82.43 100.00

25.25 52.91 21.84

25.25 78.16 100.00

f. Lower expectations about conditions/responsibilities No 828 1735 Maybe Yes 716 Source:  www.cupesse.eu (N=3279); Tosun et al. (2016b).

to entail lower earnings, so there is a connection with response category e. The working conditions, however, are not directly addressed by the EU policies. In the last step, we examine to which actors the German respondents assign responsibility in terms of addressing youth unemployment. Respondents indicate that businesses and employers have the greatest power to address youth unemployment (74 per cent), followed by education and training institutions (62 per cent). Interestingly, policy-makers – at different levels of government – are less frequently mentioned for promoting youth employment than employers and institutions of education. The prominence young Germans give to education and training institutions is well in line with the responses they gave to previous questions. With regard to business and employers, the panel members surveyed agree that these are the most influential actors in bringing young people into employment, but they do not perceive the economic situation to be the reason for youth unemployment in Germany. This is an interesting finding, suggesting that the economic conditions are

M4716-STRAßHEIM_9781785367847_t.indd 265

25/04/2019 13:32

266  Handbook of behavioural change and public policy comparatively favourable in Germany, which seems to be noticed by the respondents. The EU is credited by only 28 per cent of the respondents to have the competence to address this issue.

4. ASSESSMENT OF THE POLICY APPROACH EMBRACED BY THE EU In this section, we offer an assessment of the policy measures adopted at the EU level to bring young people into employment. We follow the 11 recommendations formulated by Weaver (2015) for ‘getting people to behave’ in order to assess whether the existing policies are in line with behavioural insights as subsumed by Weaver. The first recommendation invites policy-makers to think about the causes of the problem and why the target group may not comply with the policies adopted. We can confirm that EU policy-makers have invested in learning about the causes of youth unemployment and have devised policy instruments based on these insights. In this regard, an important source of information is Eurofound, which regularly produces reports that characterize the different facets of this issue and discuss the possibilities of solving it. For example, it was a report published by Eurofound (2012) that introduced the notion of NEETs to the EU. Young people who are NEETs were first recognized by the British government in 1988 (see Furlong 2006), and as a result of the input provided by Eurofound, it diffused to the EU level. The EU policies also address the services offered by education and training institutions, which shows that their target groups are not only young people. Businesses and employers are only indirectly targeted by the EU by offering them financial incentives for hiring young people (see Tosun et al. 2017). Nonetheless, we can confirm that EU policy-makers have demonstrated an approach that is in line with Weaver’s suggestions. The second recommendation concerns linking strategies for increased compliance to the underlying reason for non-compliance. This recommendation also seems to be taken into account by EU policy-makers since they adopted a range of strategies to facilitate the school-to-work transition of young Europeans. In this context, it does not appear to be too problematic that the policy measures adopted concentrate on employability, activation and mobility. In the case of Germany, the importance of education, skills and work experience as well as the possibility to demonstrate these were acknowledged by the survey respondents. The situation is different regarding mobility, however, since young Germans indicated a reluctance to move to a different country. The EU does not seem to have realized the low mobility of young people (in some member states) yet – a point which we will revisit shortly. Weaver’s third recommendation is that policy-makers should bear in mind that target populations are not homogenous and that strategies need to take this into account. Inspecting the Youth Guarantee Implementation Plans the individual member states had to submit, we can confirm that policy-makers demonstrate awareness to different groups of young people and their respective characteristics. In most cases, the policy measures adopted in response to the Youth Guarantee are tailor-made for specific groups of young people. For example, young jobseekers who have completed only basic education are provided different services than those who have completed more advanced education or even those who have not completed any education at all (see, e.g. Hall et al. 2015).

M4716-STRAßHEIM_9781785367847_t.indd 266

25/04/2019 13:32

Steering the behaviour of young people  267 Therefore, perhaps the most remarkable feature of the Youth Guarantee is its awareness of heterogeneity among young jobseekers. Thus, we can confirm that the EU’s approach is in line with this recommendation. The fourth recommendation concerns ways to make complying with government policies easier and more cost effective. While this recommendation is plausible, the cost argument does not apply to the specific case of the Youth Guarantee since young jobseekers do not incur costs on this programme. What does apply, however, is the recommendation about facilitating compliance by means of inducing young people to register as customers and benefit from labour market integration services offered by public employment services or other implementing bodies like ministries or municipalities (but see Shore and Tosun 2019). An alternative or complementary strategy is to expand and diversify entry points to labour market integration services, for example, through social media or proactive work with schools. The guidelines for the implementation of the Youth Guarantee also include outreach strategy recommendations, suggesting that service delivery should be facilitated by ‘taking out labour market integration services of their standard settings, and then tailoring and providing them “closer” to the young people i.e., in local community, school, one-stop-shop or mobile settings’ (Hall et al. 2015, p. 2). In this context, one-stop-shops appear to be particularly useful for providing the young unemployed people the support they need. From this, it follows that we can also consider this dimension to be addressed by the EU policy measures. The fifth suggestion is to prioritize the barriers that are most important in causing noncompliance. This is the first dimension where the current approach embraced by the EU appears to fall short – at least for Germany. As we have seen in the previous sections, young Germans are rather unwilling to move to another country in order to find a (better) job. In other words, immobility can be perceived to pose a major barrier to the labour market integration of young Germans. Despite the launch of the programme Youth on the Move and the online portal EURES (see Cahuc et al. 2013), no increased efforts have been made to enhance the geographical mobility of German youth. Of course, Germany benefits from one of the lowest levels of youth unemployment in the EU, especially in comparison to Southern European countries, and therefore, the responses given can simply be a result of insufficient problem pressure. Nevertheless, this represents an area that warrants further attention by German policy-makers and their approach to implementing the EU policies. Sixth, Weaver recommends using mass media and social media campaigns to achieve policy objectives, while also recognizing their limitations. The European Parliament (2014a, p. 3) added a question about the Youth Guarantee to the Flash Eurobarometer on European Youth and concluded that this initiative ‘suffers from a lack of visibility among its core target audience’. In response, authorities in some member states decided to make publicity for the Youth Guarantee on social media (Tosun 2017). Consequently, this recommendation is taken into consideration by both EU and national policy-makers. Weaver’s (2015, p. 813) seventh recommendation reads as follows: ‘Analyze how much compliance is “good enough” – and by whom – for the policy to be a success, and devote more resources to those “mission-critical” populations.’ Of this recommendation, only the second part about resource allocation is applicable to the case at hand. For the EU employment policies to be successful, two conditions must be met. First, youth unemployment levels should overall decrease; second, youth unemployment levels, particularly in countries where it is critically high, should decrease. To achieve this, the YEI allocates funds

M4716-STRAßHEIM_9781785367847_t.indd 267

25/04/2019 13:32

268  Handbook of behavioural change and public policy to those member states which are most severely affected by youth unemployment (Tosun 2017), corresponding to an approach that devotes more resources to ‘mission-critical’ populations. The eighth recommendation by Weaver is to look for so-called leverage points where desired behaviour can be promoted with minimum cost and resistance. This recommendation is followed to the extent that the public employment services – which in most countries are the main organizations responsible for implementing Youth Guarantee – strive to reach out to both, easily and harder accessible young people. The ones that can be reached easily can be regarded as leverage points of the implementation activities. The ninth recommendation advises policy-makers to be attentive to unanticipated behavioural effects of the policy measures they adopt. This lesson has not yet been taken up by EU policy-makers since youth employment policies are a relatively new policy area and little research on its behavioural effects exists (see Thies et al. 2018). Therefore, at this point, policy-makers are more concerned about making the policy approach work rather than assessing the unintended behavioural effects. To be sure, the beginning of the policy was not smooth, as demonstrated by the YEI. In the first year, of the 20 member states eligible for funding, only 11 had submitted Operational Programmes covering the YEI-related actions. Especially, the member states most in need were confronted with serious problems in finding the sufficient funds to launch YEI-compatible programmes. As a result, on 17 July 2014 the Members of European Parliament criticized the YEI, and on 17 July adopted a resolution (522 votes in favour, 112 votes against and 22 abstentions) calling on the European Commission and member states to modify the rules in order to improve the absorption of the funds (European Parliament 2014b). The next recommendation reads as follows: ‘Guard against reversals and erosion in improvements in policy compliance’ (Weaver 2015, p. 814). While the EU policies promoting youth employment are relatively recent, the EU has already developed a system for monitoring their implementation. This is done by regularly monitoring the national Youth Guarantee Implementation plans as well as by assessing additional reports prepared by various organizations involved in specific aspects of the policy implementation (see, e.g. Hall et al. 2015; Tosun 2017). Weaver’s final recommendation concerns the temporal dimension and multiple phases of behaviour change strategies. In the case of the EU policies, the measures adopted and implemented so far are planned for a limited period of time, that is, until 2020. The date can, in principle, be extended, but this depends on the policies’ perceived success. Consequently, the EU policy measures cannot be conceived as a long-term behaviour change strategy. In a similar vein, the policies adopted do not suggest that the policy measures take into account multiple phases of behaviour change. All in all, the EU’s policy regime on youth unemployment is mostly in line with the recommendations formulated by Weaver. Areas for improvement include prioritizing the barriers that are most important in causing non-compliance and developing strategies with longer time horizons and multiple phases. Also not addressed so far are the chances of unintended behavioural effects, but this is not surprising considering the policies were adopted only recently and under strong political pressure. Therefore, currently, priority is given to putting the Youth Guarantee into action and bringing about any behavioural changes, regardless of whether they are intended or unintended.

M4716-STRAßHEIM_9781785367847_t.indd 268

25/04/2019 13:32

Steering the behaviour of young people  269

5.  CONCLUDING REMARKS What are the main perceived barriers to promoting employment among young Germans? How well suited are the EU level policy measures – as implemented by the German government – in overcoming these barriers? These two research questions guided the analysis carried out in this chapter. Building on behavioural science-based approaches to public policies (see Thaler and Sunstein 2003, 2008) but taking a more deliberative perspective (see Grüne-Yanoff and Hertwig 2016) and applying it to the EU’s policy approaches to foster youth employment (see Thies et al. 2018), we argue that there is a substantial behavioural component in policy analysis, design and implementation that is decisive for the failure or success of the policies implemented (on this, see also, e.g. Rapp et al. 2018). Drawing on original data, we showed that young Germans are willing to move within the country, learn new skills, or even get retrained in order to get a (better) job. The majority of them, however, are unwilling to move to a different country. While the policy measures also aim to increase the geographic mobility of young people, it appears promising to invest more resources to attain this goal. More broadly, we discussed the policy measures adopted at the EU level in light of the recommendations formulated by Weaver (2015) to bring about behavioural changes. Remarkably, despite the EU’s limited experience with youth employment policies, the policy approach embraced is mostly in line with Weaver’s recommendations. Despite this positive assessment, we must not forget the policy approach is limited, as it only concentrates on three major dimensions: employability, activation and mobility (Lahusen et al. 2013; Tosun et al. 2017). It does not include social protection for young people. Moreover, the policy measures adopted only marginally address the demand-side of youth labour. To be fair, however, addressing the demand-side would have been difficult, due to different national labour market institutions and specificities in industrial relations. While this study provides some instructive insights, it suffers from at least three limitations. First, our analysis only concentrates on EU-level policy measures and does not offer a systematic assessment of relevant national policy measures. We are well aware that the national level is more important for adopting and implementing policies – especially for targeting the demand-side of youth labour. Second, we strived to combine survey data assessing the policy measures adopted by EU policy-makers and implemented by national policy-makers and bureaucrats. The data used was not specifically collected for this purpose and therefore, we were limited in the possibility to systematically integrate the different data sources. Notwithstanding these two limitations, the survey questions used did offer some useful insights that could not have arisen otherwise. Third, our analysis only allows for drawing conclusions with regard to deliberative strategies of steering the behaviour of target groups. We do not provide insights that would pertain to the small but growing literature on the concepts of nudging and libertarian paternalism (see Thaler and Sunstein 2003, 2008). This stems from the research interest of our study, namely policies that aim to facilitate the school-to-work transition of young people in Europe. Nevertheless, by giving first insights into the motivation of the target group of young people and systematically comparing these with the policies implemented by the EU to foster youth employment and the recommendations by Weaver (2015), we hope to contribute to the literature on how appropriate policies can alter behaviour. We are confident that the limitations of our approach will inspire future research on this topic.

M4716-STRAßHEIM_9781785367847_t.indd 269

25/04/2019 13:32

270  Handbook of behavioural change and public policy Indeed, in the past few years there has been a growing interest in exploring the behavioural foundations of public policy (see, e.g. Shafir 2013), which is an encouraging development. What is now needed is to establish a genuine dialogue between behavioural sciences and policy studies.

NOTES 1. The abbreviation CUPESSE stands for Cultural Pathways to Economic Self-Sufficiency and Entrepre­ neurship (www.cupesse.eu), which received funding by the European Commission under FP7 Grant Agreement (No. 613257). 2. The main characteristic of this method is that members of YouGov Germany’s panel were invited to participate in a survey and then were assigned to a specific survey once they accepted the invitation. The overall number of panel members of this age group is 19 000; 3989 panel members started the survey, which indicates a response rate of approximately 21 per cent. Of those who started taking the survey, about 7 per cent (N = 274) did not complete the survey and were not included in the final sample (N = 3 279).

REFERENCES Borrás, S. and C.M. Radaelli (2011), ‘The politics of governance architectures: creation, change and effects of the EU Lisbon Strategy’, Journal of European Public Policy, 18 (4), 463‒484. Borrás, S. and C.M. Radaelli (2015), ‘Open method of co-ordination for democracy? Standards and purposes’, Journal of European Public Policy, 22 (1), 129‒144. Brandt, M. and K. Hank (2014), ‘Scars that will not disappear: long-term associations between early and later life unemployment under different welfare regimes’, Journal of Social Policy, 43 (4), 727‒743. Brown, P. (2012), ‘A nudge in the right direction? Towards a sociological engagement with libertarian paternalism’, Social Policy and Society, 11 (3), 305‒317. Cahuc, P., S. Carcillo, U. Rinne and K.F. Zimmermann (2013), ‘Youth unemployment in old Europe: the polar cases of France and Germany’, IZA Journal of European Labor Studies, 2 (1), 1. Chung, H., S. Bekker and H. Houwing (2012), ‘Young people and the post-recession labour market in the context of Europe 2020’, Transfer: European Review of Labour and Research, 18 (3), 301‒317. Comte, S., M. Wardman and G. Whelan (2000), ‘Drivers’ acceptance of automatic speed limiters: implications for policy and implementation’, Transport Policy, 7 (4), 259‒267. European Commission (2015a), ‘Addressing youth unemployment in the EU’, accessed 12 April 2015 at http:// ec.europa.eu/social/main.jsp?catId=1036. European Commission (2015b), ‘Accelerated pre-financing for the Youth Employment Initiative: questions and answers’, accessed 27 February 2017 at http://europa.eu/rapid/press-release_MEMO-15-4101_en.htm. European Parliament (2014a), Flash Eurobarometer of the European Parliament (EP EB395) – European Youth in 2014: Analytical Synthesis, Brussels and Strasbourg: European Parliament. European Parliament (2014b), ‘European Parliament resolution of 17 July 2014 on Youth Employment (2014/2713(RSP))’, accessed 27 February 2017 at http://www.europarl.europa.eu/sides/getDoc.do?type=TA&​ reference=P8-TA-2014-0010&language=EN. European Parliament (2016), ‘Employment Policy’, accessed 27 February 2017 at http://www.europarl.europa. eu/ftu/pdf/en/FTU_5.10.3.pdf. Eurofound (2012), NEETs: Young people not in employment education or training: characteristics, costs and policy responses in Europe, Luxembourg: Publications Office of the European Union. Furlong, A. (2006), ‘Not a very NEET solution representing problematic labour market transitions among early school-leavers’, Work, Employment & Society, 20 (3), 553‒569. Grüne-Yanoff, T. and R. Hertwig (2016), ‘Nudge versus boost: how coherent are policy and theory?’, Minds and Machines, 26 (1), 149–183. Hall, A.-M., H. Metcalfe and P. Irving (2015), PES practices for the outreach and activation of NEETs: a contribution of the Network of Public Employment Services, Brussels: European Commission. Heidenreich, M. and G. Bischoff (2008), ‘The open method of co-ordination: a way to the Europeanization of social and employment policies?’, Journal of Common Market Studies, 46 (3), 497‒532.

M4716-STRAßHEIM_9781785367847_t.indd 270

25/04/2019 13:32

Steering the behaviour of young people  271 Howlett, M. (1991), ‘Policy instruments, policy styles, and policy implementations, national approaches to theories of instrument choice’, Policy Studies Journal, 19 (2), 1‒21. Howlett, M. (2014), ‘From the ‘old’ to the ‘new’ policy design: design thinking beyond markets and collaborative governance, Policy Sciences, 47 (3), 187‒207. Hutchinson, J.M. and G. Gigerenzer (2005), ‘Simple heuristics and rules of thumb: where psychologists and behavioural biologists might meet’, Behavioural Processes, 69 (2), 97‒124. John, P., G. Smith and G. Stoker (2009), ‘Nudge nudge, think think: two strategies for changing civic behaviour’, The Political Quarterly, 80 (3), 361‒370. Kahneman, D., P. Slovic and A. Tversky (eds) (1982), Judgment Under Uncertainty: Heuristics and Biases, New York: Cambridge University Press. Lahusen, C., N. Schulz and P.R. Graziano (2013), ‘Promoting social Europe? The development of European youth unemployment policies’, International Journal of Social Welfare, 22 (3), 300‒309. Oliver, A. (2015), ‘Nudging, shoving, and budging: behavioural economic-informed policy’, Public Adminis­ tration, 93 (3), 700‒714. O’Reilly, J., W. Eichhorst, A. Gabos, K. Hadjivassiliou, L. Kurakova, D. Lain, J. Lesckhe, S. McGuinness, T. Nazio, R. Ortlieb, H. Russell and P. Villa (2015), ‘Five characteristics of youth unemployment in Europe: flexibility, education, migration, family legacies and EU policy’, Sage Open, 5 (1), 1‒19. Rapp, C., J. Shore and J. Tosun (2018), ‘Not so risky business? How social policies shape the perceived feasibility of self-employment’, Journal of European Social Policy, 28 (2), 143-160. Schneider, A. and H. Ingram (1990), ‘Behavioral assumptions of policy tools’, The Journal of Politics, 52 (2), 510‒529. Shafir, E. (ed.) (2013), The Behavioral Foundations of Public Policy, Princeton: University Press. Shore, J. and J. Tosun (2019), ‘Assessing youth labour market services: young people’s perceptions and evaluations of service delivery in Germany’, Public Policy and Administration, 34 (1), 22-41. Thaler, R.H. and C.R. Sunstein (2003), ‘Libertarian paternalism’, American Economic Review, 93 (2), 175‒179. Thaler, R.H. and C.R. Sunstein (2008), Nudge: Improving Decisions about Health, Wealth and Happiness, New Haven, CT: Yale University Press. Thies, L, Z. Cemalcilar and J. Tosun (2018). ‘Nudging’ the youth guarantee hit its target: lessons from behavioural public policy. Policy Brief VI of the CUPESSE Project, accessed 14 January 2019 at http://cupesse.eu/ fileadmin/cupesse/downloads/policy-briefs/Policy-Brief-VI.pdf. Tosun, J. (2017), ‘Promoting youth employment through multi-organizational governance’, Public Money & Management Volume, 37 (1), 39‒46. Tosun, J., M. Unt and E. Wadensjö (2017), ‘Youth-oriented active labour market policies: explaining policy effort in the Nordic and the Baltic states’, Social Policy & Administration, 51 (4), 598–616. Tosun, J., C. Jensen, S. Speckesser and J. O’Reilly (2016a), ‘The absorption of structural and investment funds and youth unemployment: an empirical test’, in Peter Berkowitz, John Bachtler, Tatjana Muravska and Sally Hardy (eds), EU Cohesion Policy: Reassessing Performance and Direction, London: Routledge, 151‒168. Tosun, J., F. Hörisch, B. Schuck, J. Shore and R. Strohmeyer (2016b), CUPESSE Two-Generation Survey: Germany, Heidelberg/Mannheim: Heidelberg University/ MZES. Tosun, J., J.L. Arco-Tirado, M. Caserta, Z. Cemalcilar, M. Freitag, F. Hörisch, C. Jensen, B. Kittel, L. Littvay, M. Lukes, W.A. Maloney, M. Mühlböck, E. Rainsford, C. Rapp, B. Schuck, J. Shore, N. Steiber, N. Sümer, P. Tsakoglou, M. Vancea and F. Vegetti (2018), ‘Perceived economic self-sufficiency: A country- and generation-comparative approach’, European Political Science, doi: 10.1057/s41304-018-0186-3. Warmuth, J.R., B. Kittel, N. Steiber and M. Mühlböck (2015), ‘Cultural pathways to economic self-sufficiency and entrepreneurship. an overview of theoretical perspectives on micromechanisms’, CUPESSE Working Paper No. 1, CUPESSE Working Paper Series, Heidelberg. Weaver, R.K. (2015), ‘Getting people to behave: research lessons for policy makers’, Public Administration Review, 75 (6), 806‒816. Wieland, C. (2015), ‘Germany’s dual vocational-training system: possibilities for and limitations to transferability’, Local Economy, 30 (5), 577‒583.

M4716-STRAßHEIM_9781785367847_t.indd 271

25/04/2019 13:32

19.  The governance of behavioural taxation: moralization and the new modes of tax collection Sebastian Botzem

1. INTRODUCTION: BEHAVIOURAL GOVERNANCE AND TAXATION POLICY1 Fiscal measures, most notably taxes, are a cornerstone of public policy-making in any jurisdiction. Moreover, they are a core characteristic of statehood. Taxation is a primary tool to collect resources and to alter individual and organizational behaviour. At the same time, levying taxes is an indication of collective appropriateness: Undesired behaviour and lifestyles are made costlier while tax-deductibility legitimates collectively preferred choices and intends to stimulate desired conduct. Today, in most jurisdictions, taxes are the predominant instrument to ensure that public entities have sufficient financial means available. In addition to taxation policies, tax collection itself is becoming the object of academic and practical engagement. This chapter will therefore focus on behavioural aspects of tax collection, most notably, the rise of field experiments to improve resource collection, and discuss wider political implications of these developments. Taxes are the basis of (re-)distributive policies and essential to maintain public services. As OECD-Data indicates, in most industrialized countries, the tax quota at GDP has remained remarkably stable for the last two decades, ranging mostly between 30 to 50 per cent of GDP (OECD 2016a). However, in the age of globalization, capital and mobile factors of production are taxed less relative to individual incomes generated from waged labour. To attract capital – mostly corporations but also high net worth individuals – a fierce tax competition between jurisdictions takes place. Commonly, increased marketorientation of government activities has led to praising efficient resource allocation and lean government. After ridding states from many of their asset holdings and part of their infrastructure, privatization as an almost universally accepted paradigm can be observed with respect to government entities as well. In 2014, the British Behavioural Insights Team (BIT), once part of Downing Street’s Cabinet office, was privatized and became a limited company. It engaged in consulting governments around the world on how to introduce behavioural change policies based on scientific experiments, most notably randomized control trials. BIT and similar nudge units, often set up as organizational entities within the state bureaucracy, aim at identifying superior choice architectures (Thaler et al. 2013) that include fiscal as well as non-fiscal measures. Currently, nudging advocates are expanding the realm of behavioural governance to core responsibilities of public administration, such as tax collection. In comparison to other policy fields such as consumer or health policies, behavioural change policies in tax collection are relatively recent phenomena, making up a nascent field of activity (Hallsworth 2014). With respect to theoretical implications, these experiments are indicative of a process of reinterpretation of the relationship between the individual and 272

M4716-STRAßHEIM_9781785367847_t.indd 272

25/04/2019 13:32

The governance of behavioural taxation  273 the state, placing emphasis on moral conduct and social norms thereby bringing to the fore the moralizing character of official communications in the field of taxation policy (see Nonhoff and Vogelmann 2016). This chapter discusses the relevance of behavioural change governance in taxation policies in four steps: A first brief section introduces taxation policies and the novel role of behavioural governance with regard to tax collection. The second part gives account of the challenges arising from incomplete tax payments and tax morale, and outlines how these issues are debated in behavioural economics. A third section outlines the activities of nudge units and their contributions to taxation policies. The final section concludes with discussing the insights of behavioural governance and taxation and interprets findings with respect to the apparent reconfiguration of state-citizen relations.

2.  THE CHALLENGES OF TAX COMPETITION Taxes constitute an individual’s or a legal entity’s financial obligation towards the general public and the state. Fundamentally, taxes epitomize an important aspect of the relation between the citizenry with its rights and obligations on the one hand, and the state as the manager of collective resources on the other hand. Needless to say, this relation is complex and contested as it involves a transfer of funds from citizens to the state and the fiduciary management of these resources by public actors. Distributive conflicts unfold regarding the appropriate level of taxation, favouring selected stakeholder groups over others and legitimizing existing trade-off decisions. Similarly, in academia, the debate about ‘optimal’ taxation is equally contested and has given rise to an entire subfield of economics.2 In the era of neoliberalism, in which states are conceded only limited capacities to act in a comprehensive manner, the scope of public engagement is put equally in question as are redistributive policies (Gamble 2006; Mudge 2008). The trend of fiscal restraint is particularly relevant for taxation policies and contributes to the rise of behaviourally informed instruments that indicate promising and novel solutions to old problems concerning the effective collection of taxes. Improving tax collection is often seen as one of the few viable strategies to increase public resources without raising taxes which is deemed politically costly and difficult to implement under the conditions of global tax competition (see Dietsch and Rixen 2016). Increased global capital mobility is said to confront taxation policy-makers with three interrelated constraints: First, international competition has resulted in reductions on capital taxes; second, domestic economic problems call for a reduction on labour and capital taxation; and, third, budgetary imperatives limit the scope for tax policy change (Swank and Steinmo 2002, p. 651). While the set of political choices is limited, ‘policymakers in contemporary democratic polities have faced intensifying pressure to reform tax policy to promote economic efficiency. They have found, however, little room for maneuver’ (Swank and Steinmo 2002, p. 651). In this context, improving tax compliance, essentially intending to enhance fiscal revenues through novel means of tax collection, but without altering legislation, becomes a seemingly low cost political strategy to pursue. This opens possibilities for behaviourally informed policy innovations that draw on direct communication with citizens appealing to the ethical aspects of behaviour. As will be shown in the later part of this chapter, the monetary effects of nudging tax compliance

M4716-STRAßHEIM_9781785367847_t.indd 273

25/04/2019 13:32

274  Handbook of behavioural change and public policy are minimal. This makes moralization a cornerstone of tax collection altering the statetaxpayer relationship. Taxes direct individuals’ behaviour in multiple ways. The most immediate effect on daily activities unfolds when taxes increase the price for consumption of selected goods. Indirect taxes, such as sales taxes or value added taxes, increase product prices and affect available resources. In addition, direct taxes on certain products or activities equally impact consumption choices of citizens. Such levies can be collected by the state to actively steer consumer behaviour, even though the framing of a ‘tax nudge’ is a relatively recent phenomenon (Hilton at al. 2014).3 Actively guiding behaviour through taxes, for instance on alcohol, tobacco or gambling, is a centuries old practice on which public budgets continue to rely. However, recent discussions suggest ‘that individuals are inattentive to some types of incentives’ such as taxes because they are too complex and non-transparent (Chetty et al. 2009). Therefore, while taxing the consumption of undesired products and unwanted consumption continues to play a role in securing public finances, so far, no final consensus has been reached about the concrete impact of taxes on guiding behaviour. While specific levies might be effective in crowding out potentially dangerous or unhealthy products and lifestyles, they are not (yet) core instruments of nudging. In part, because ‘changing people’s economic incentives by imposing a tax is not a “nudge” in the classic definition’ (BIT 2016). While direct and indirect taxation have not yet been widely explored as integral parts of more recent attempts to inform behavioural change policies, nudges – predominantly explored in natural field experiments – gain traction with tax authorities when it comes to collecting taxes. Experiment-based policy tools are applied to narrow the tax gap and to improve tax compliance. In practice, behavioural change policies are sought to discourage deviant behaviour of taxpayers that do not follow their legally prescribed duties to pay taxes.

3. TAX COMPLIANCE, TAX MORALE AND BEHAVIOURAL ECONOMICS Endeavours to improve tax compliance are as old as tax collection itself. The level of compliance (or non-compliance) of tax payments is an important yard stick for the effectiveness of tax collection and mirrors the citizenry’s willingness to pay taxes. In Europe, for instance, the sovereign debt crisis and the dire state of public finances after the financial crisis of 2007, underline the importance of effective tax collection precisely because public authorities are chronically underfunded (Streeck 2014). To secure needed finances, recent work on tax policies investigates the behavioural aspect of tax compliance and the issue of tax morale more broadly (Ritsatos 2014). Tax non-compliance is often labelled a ‘tax gap’. The US-American Internal Revenue Service (IRS) distinguishes between gross and net tax gap: ‘The gross tax gap is the amount of true tax liability that is not paid voluntarily and timely … . The net tax gap is the gross tax gap less tax that will be subsequently collected’ either through voluntary payment or after administrative and enforcement activities (IRS 2016, p. 1). For the tax year 2008‒2010, the IRS estimates a remaining (net) tax gap of about 16.3 per cent of the entire US tax liability, about US $406 billion. For the given year, US $52 billion were enforced and late payments. Adding these to the net tax gap leads to a gross tax gap of

M4716-STRAßHEIM_9781785367847_t.indd 274

25/04/2019 13:32

The governance of behavioural taxation  275 US $458 billion (of which 85 per cent were underreported, 8 per cent were underpaid and 7 per cent not filled). The underreported individual income tax makes up US $291 billion, of which at least US $125 billion are individual business income (43 per cent) (IRS 2016, p. 3). US figures appear to be rather high by international comparison, but comparing tax gaps between countries is highly complicated because of methodological differences and the size of the shadow economy which ranges from 10 to 65 per cent of GDP in different countries (Schneider et al. 2010). International comparative data is difficult to obtain because of classification and measurement differences. Figures for the UK for instance, identify a tax gap of about 6.5 per cent of all tax liabilities (HM Revenue & Customs 2016). The tax gap indicates unlawful behaviour; not paying taxes constitutes an illicit practice in form of misrepresenting tax liabilities, dishonest reporting or exaggerating deductions. Such practices are tax evading activities and represent a breach of legal obligations. Tax avoidance, in contrast, is much more difficult to assess and to detect, let alone prosecuted once it has been identified as illicit practice. Recent revelations such as Lux Leaks – essentially benefitting large multinational companies based in Luxembourg with the help of Luxemburg tax authorities – (European Commission 2016) and the Panama Papers scandal, which documented the systematic establishment of shell corporations to divert financial flows, evade taxes and international sanctions (European Parliament 2016), underscore that tax avoidance is utterly complex. Tax avoidance usually relies on ­sophisticated – and costly – financial services which centre around identifying tax loopholes and consulting clients and/or arranging their business affairs to take advantage of imprecise tax regulation. Tax avoiding individuals and organizations tend to pursue innovative, often questionable financial transactions to reduce tax liabilities to the lowest possible level. Despite being a widespread practice of rich individuals and corporations, tax avoidance is ethically questionable because it can be taken advantage of predominantly by wealthy and resourceful actors. It also undermines tax morale more generally even if being in conformity with legal prescriptions. More recently, however, many jurisdictions have become concerned with both tax avoidance and tax evasion, partly because the two are closely related and because governments are deprived of fiscal resources. Therefore, in addition to activities to limit the attractiveness of tax havens (or so called secrecy jurisdictions) the OECD is seeking to develop activities to detect and discourage tax avoidance strategies by multinational corporations (see BEPS project: Base Erosion and Profit Shifting, OECD 2016b). In academia, tax compliance has been a central topic ever since Allingham and Sandmo (1972) published their seminal piece on income tax evasion. Following a rationalist perspective, the authors seek to understand the conditions under which people pay taxes instead of evading their duties. They apply a model of rationalist behaviour which interprets tax evasion as a form of portfolio choice derived from the probability to be debunked (see also Sandmo 2005). Consequently, tax audits became an issue of concern for many of the deterrence-oriented approaches to enhance tax collection (Hallsworth 2014). The limits of rationalist and neoclassical approaches that followed Allingham and Sandmo’s footsteps were outlined by Alm et al. (1992), drawing on experimental studies that suggest complementary explanations for why people might pay taxes based on prospect theory (Kahneman and Tversky 1979). Rather than seeking answers to why people do pay taxes despite opportunities to avoid and evade obligations in rationalist frameworks, alternative

M4716-STRAßHEIM_9781785367847_t.indd 275

25/04/2019 13:32

276  Handbook of behavioural change and public policy explanations are rooted in behavioural economic theory. Over the last decades, issues such as tax morale, trust and cultural determinants of tax compliance have emerged, blending dominant neoclassical accounts with methods and definitions from behavioural sciences focussing specifically on the ethical dimensions of agents’ behaviour (Ritsatos 2014, p. 245). Ethical considerations therefore complemented rationalist reasoning by bringing in behavioural insights from psychology and social psychology (see for instance Kahneman et al. 1986). With respect to taxation, implications of ethical decision-making are relatively straightforward: If taxpayers believe evading taxes to be unethical, then levels of compliance will rise. Furthermore, theoretical analysis of moral intensity may lead to identification of individual ethical stimuli. Such identification will have policy implications for tax authorities, providing the appropriate indications for tax maximization policy, while reinforcing the formation of ethical sentiments to taxpayers. (Ritsatos 2014, p. 249)

Tax morale became a leading concept in academic research interested in the social norms underlying tax compliance (Luttmer and Singhal 2014) and seeking policy solutions ‘besides coercion that will help to reduce tax fraud’ (Torgler 2011, p. 4). Numerous studies were undertaken to link tax compliance to cultural determinants (Alm and Torgler 2006) and to consider institutional factors impacting tax morale (Torgler 2003). With respect to concrete interactions between citizenry and public authorities in their capacity as tax collectors, behaviourally informed approaches argue that a ‘psychological contract between taxpayers and tax authorities exists’ (Feld and Frey 2002, p. 90), which compels tax authorities to take positive actions to support this contractual relationship (Feld and Frey 2002, p. 91). Similarly, the ‘slippery slope framework’ (Kirchler et al. 2008) suggests that the relationship between taxpayers and tax authorities is characterized by an interplay of voluntary honesty (function of trust build-up within the context of the psychological contract) and enforced honesty (compliance attained by powerful enforcement mechanisms): ‘Tax proceeds may increase from the minimum point, either by empowering the enforcement mechanism, or by increasing the embedded trust in authorities’ (Ritsatos 2014, p. 256). Growing ethical considerations of taxpayers’ behaviour and the psychologization of tax collection underline the desire to leave behind the theoretical limitations of rationalist accounts. Moreover, behavioural economics have a long tradition in drawing on psychological and socio-psychological methods and concepts in their empirical research. Laboratory experiments have been relevant first attempts to better understand details of taxpayers’ behaviour but come with well-known limitations of laboratory experiments such as lack of external validity and lack of respondents’ representativeness (Hallsworth 2014, p. 667). In addition, recent findings indicate that tax behaviour varies according to different personal characteristics or previous experience, calling into question onesize-fits-all approaches to improve tax compliance (Brockmann et al. 2015; Choo et al. 2016). Therefore, to allow for a more encompassing impact measurement, natural field experiments with randomly allocated treatments, often conducted in cooperation with national tax authorities, are flourishing (Slemrod and Weber 2012).

M4716-STRAßHEIM_9781785367847_t.indd 276

25/04/2019 13:32

The governance of behavioural taxation  277

4. ADVANCES IN TAX COLLECTION: THE RISE OF RANDOMIZED CONTROL TRIALS Since the 2000s, field experiments have gained importance as novel tools to investigate conditions of taxpayers’ non-compliance and to improve tax collection more generally. However, scope for action is limited as legal prescriptions make it unlikely to modify tax rates or to adapt tax bases as part of an experimental setting. Nevertheless, there is room for alternative treatments in tax policy: Although tax rates and bases are unlikely to be randomized in the field, for other tax-system instruments policy randomization is possible and, more important, has become a reality. The researcher must overcome tax authorities’ understandable reluctance to randomize tax rates or bases, so that randomized tax experiments have heretofore mostly concerned tax-system instruments such as communication with taxpayers. (Slemrod 2016, p. 13)

Such ‘communication’ with taxpayers mostly takes place in natural field experiments (Hallsworth 2014). Around the world, a number of randomized control trials (RCTs) have been conducted, often using tax letters to induce behavioural change that can subsequently be measured. Such studies have been carried out in Europe (for instance in Austria, the UK, Germany, Switzerland) North America (the US and Canada) and Latin America (Argentina, Chile, Peru and Guatemala) (Kettle et al. 2016, p. 4). In developing countries, RCTs using text reminders to increase tax compliance are often conducted with support of the World Bank and, interestingly, in partnership with the UK-based Behavioural Insights Team (BIT) (World Bank 2016).4 While proponents of RCT-studies see evidence that communication with taxpayers affects tax behaviour (Kettle et al. 2016, p. 4), concrete findings vary. Nevertheless, one of the merits of RCTs is that they allow investigating the behaviour of selected actor groups (e.g. taxpayers, law makers, tax advisors, tax authorities, etc.) at specific stages within the taxation process and to focus on selected interactions (Pickhardt and Prinz 2014). Randomized experiments address different aspects of tax collection. In Denmark, for instance, a study showed that while in general, the tax evasion rate is quite small, ‘threat-of-audit letters’ have a measurable impact on self-reported income (Kleven et al. 2010). In contrast, research on church tax from Germany demonstrates that, contrary to rationalist assumptions, intrinsically motivated compliance plays an important role guiding taxpayers’ behaviour, indicting the relevance of duty-to-comply preferences (Dwenger et al. 2016). Increasingly, large-scale experiments with tax letters are carried out using RCTs to measure the impact of selected written messages. These studies focus on identifying, which formulations in tax letters yield the most effective results. For instance, a natural field experiment in Argentina, conducted in cooperation with the Inter-American Development Bank, found that, with respect to property taxes, a deterrence message leads to measurably higher probability to comply (almost 5 percentage points) compared to the control group (Castro and Scartascini 2015).5 Other treatments within the same experiment included reciprocity and peer-effects. Here, no significant stimulation of behavioural change on behalf of letter recipients could be observed, leading the authors to conclude that messages, which draw on moral reasoning, have limitations. For Argentina, they infer that people who ‘do not comply may have lower trust in the government, which may translate into lower effectiveness of the messages (i.e., those who do not trust the

M4716-STRAßHEIM_9781785367847_t.indd 277

25/04/2019 13:32

278  Handbook of behavioural change and public policy government may not trust the messages they receive about the levels of tax evasion and public works)’ (Castro and Scartascini 2015, p. 76). Behavioural interventions that use RCTs on a large scale to remind tardy taxpayers to meet their obligations are on the rise in many countries. In essence, these experiments test which type of communication works most effectively with recipients, usually comparing deterrent messages with communications that make reference to social norms. The latter contain assumptions about the appropriateness of the expected behaviour (doing the ‘right thing’) and suggest that these expectations can be met at the individual level through adhering to existing social norms. In its positive and negative form, moralization becomes a core aspect of tax authorities’ communication with taxpayers. In Guatemala, a trial conducted by the World Bank and the UK Behavioural Insights Team used four adapted letters in addition to one standard letter. Three other types of letters contained ‘additional persuasive messages: one included a social norms message, another highlighted non-declaration as a deliberate choice, and the third emphasized national pride’ (Kettle et al. 2016, p. 4). The two most effective letters were the deterrent message (deliberate choice) and the social norms approach: The deliberate choice/deterrent message threatens an increase in the ‘moral cost’ of non-compliance (by signalling that evasion is a deliberate choice), while the social norms message increases the ‘moral benefit’ of compliance (by providing information on positive effects for public goods) (Kettle et al. 2016, p. 7). The Guatemala trial, as it was conducted, resulted in an effective financial benefit of US $288 301 for Guatemalan tax authorities that were paid as a direct response within a few weeks after having received a letter. Had the most effective letter (the deliberate choice/deterrent massage) been used to address everyone with overdue taxes, according to the authors of the study, it would have resulted in more than US $750 000, a profit presenting a ‘36 times return on investment for the tax authority’ (Kettle et al. 2016, p. 20). Outlining the positive financial return of natural field experiments is a frequent argument to promote the application of RCT studies, specifically by policy practitioners, such as the London based Behavioural Insights Team, who occasionally conducts studies in cooperation with academic institutions. To enhance tax collection in the UK, a large study was carried out by BIT and Her Majesty’s Revenue & Customs (HM R&C) aiming to close the gap between declared taxes and taxes effectively paid. The underlying assumption was that ‘individuals which are not liquidity-constrained will trade off an explicit financial liability and a “moral” cost’ (Hallsworth et al. 2014, p. 7, original emphasis). Moral cost is thought of as a combination of norms and public goods concerns. The relative importance of different aspects of moral cost is therefore the object of empirical investigations, in this case a natural field experiment. In the study, different sentences are introduced into HM R&C-letters sent out to individuals with outstanding tax payments. Distinct phrases are included into the letter on the basis of randomized selection, each encouraging payment according to specific assumptions. The field experiment’s aim is to use selected phrases in tax letters to empirically assess how to most effectively reduce the psychological distance between the taxpayer and the tax authority. Ultimately, the goal is to make the social norm of appropriate behaviour, in this case paying outstanding taxes, more salient (Hallsworth et al. 2014, p. 10). In total, over 100 000 reminder letters were sent out in 2011 to incentivize recipients to follow obligations. Findings support the general effectiveness of sending mail to negligent

M4716-STRAßHEIM_9781785367847_t.indd 278

25/04/2019 13:32

The governance of behavioural taxation  279 taxpayers: Within less than a month after receipt, those citizens receiving a letter from HM R&C were four times more likely to pay their tax bill compared to those who did not receive a letter (Hallsworth et al. 2014, p. 16f.). Letters with a social norm reference included the most effective message, but public goods framing also yielded positive results. Measuring whether respondents paid the outstanding tax in the three weeks following the letter, the social norm letter (message included read: ‘Nine out of ten people in the UK pay their tax on time. You are currently in the very small minority of people who have not paid us yet’ (Hallsworth et al. 2014, p. 14) produced the largest effect, amounting to 5.1 percentage points increase in payment rates (Hallsworth et al. 2014, p. 18). The experiment generated an additional £1.9 million in tax revenues within three weeks. Had the social norm letter been sent to every taxpayer, this wording would have generated an accelerated revenue of £11.3 million, according to estimations of HM R&C. With regard to the total outstanding UK tax gap in 2011, £33 billion, advancing £11.3 million makes up about 0.034 per cent of the total outstanding UK tax gap (HM Revenue & Customs 2016). While absolute numbers are noteworthy, with regard to the total tax gap, the advancement of outstanding payments seems rather negligible. Additional positive effects can also be assumed for tax administrations as they receive the money with little additional effort. Despite using the advanced taxes as an argument to sell nudging via tax letters to international organizations and public authorities, additional financial resources are not the primary motivation to nudge taxpayers. As they are currently designed, natural field experiments mostly do not generate additional resources but advance later payments. Nudging practitioners therefore see the merits of behavioural change policies in leaner public administrations and in giving clear guidance to citizens: Now it’s important to know that we’re talking about bringing forward revenue rather than additional revenue. Most of the time the tax authority will get to most of the people eventually. What you’re really doing in this area, in a tax system where ultimately almost everybody pays eventually, is you’re making the processes much more efficient. You’re making them better for citizens. They don’t get chased up. They don’t get taken to court. And you’re saving costs to the administration as well. You’re not having to spend money on people constantly ringing people up, writing extra letters, and taking people to court. (Owain Service 2015, interview with ‘The Psych Report’, Nesterak 2015)

As additional resources are not the primary concern when sending out tax letters, the focus rests on reinterpreting the relationship of citizens and public authority. Using tax letters as personalized communication allows to directly address citizens and to convince them to behave according to established social norms; to remind them of their obligations of the ‘psychological contract’. It is made clear to taxpayers that paying taxes is a moral obligation. Improving tax morale is therefore no longer just a normative academic agenda dominant in behavioural economics. The state is using moralization as an approach to address its citizenry, invoking an imagined contract between taxpayers and tax authorities. Preliminary findings suggest that deterrence messages seem to yield more convincing results in the Global South, while in the industrialized world, social norms appear to bring forward more positive reactions (that is a higher proportion of rule compliance). These comparisons indicate that RCTs can be used to document the effectiveness of a wide variety of messages, ranging from positive and supportive formulations to negative and moralizing ones. However, methodological considerations provide little guidance on

M4716-STRAßHEIM_9781785367847_t.indd 279

25/04/2019 13:32

280  Handbook of behavioural change and public policy how to make use of field experiments. Consequently, nudging experts, tax collectors and academics have reached consensus that some balance needs to be struck between incentives to increase the ‘trust in authorities’ on the one hand and the need to substantiate and enforce the ‘power of authorities’ to influence tax behaviour in a desired way on the other hand. RCTs can be used to identify the most suitable combination ‘of deterrent fears and civic obligations as motivation to comply’ (Hallsworth 2014, p. 671), shining light on the possibly not-so-subtle motivations of public agents.

5.  CONCLUSION: ADVENT OF THE MORALIZING TAX STATE Intentions to close the tax gap are as old as taxation itself. Taxing individuals and economic entities has always been contested as it entails the forced collection and subsequent redistribution of material resources. While identifying new sources of taxation is neither particularly high on the agenda of economists nor political decision-makers, closing the tax gap is a major concern. During the last quarter of the twentieth century, the rise of behavioural economics has contributed substantially to redefining tax collection: the relationship between citizens and tax authorities is increasingly reinterpreted as a personal responsibility, a psychological contract as Feld and Frey (2002) have put it. While these innovations have acknowledged the limits of rational choice approaches, the focus on social-psychological aspects of deviant behaviour of taxpayers drives the moralizing discourse of public authorities, indicating the ambiguities inherent in libertarian paternalism (see Thaler and Sunstein 2003). Recent laboratory and natural field experiments have brought about ever more sophisticated studies of how to nudge citizens to pay outstanding tax liabilities. The ‘discovery’ of social norms as a decisive factor for individuals’ behaviour comes as little surprise to those thinking about the social world in complex categories. However, the big picture gives rise to more critical considerations as it sheds light on who is not being addressed through nudges. The focus on individual misconduct underrates the systematic imbalances the modern tax state is faced with. In particular, how behaviourally informed policies can address larger issues of tax collection is unclear. If field experiments and RCTs are not going to be significantly altered to address the tax gaps of corporations and high-networth-individuals that use professional services to limit tax liabilities, the legitimacy of tax authorities might be significantly undermined. Despite attempts to increasingly focus on firms (Hallsworth 2014, Kosonen and Ropponen 2015), so far, deviant behaviour is problematized primarily at the individual level (see also Zuidhof, Chapter 12 in this volume). This leads to a serious underestimation of corporate tax evasion. And, perhaps more pressing, the issue of tax avoidance is not even considered in current approaches of behavioural economics. These shortcomings are echoed by the fact that additional financial resources are of limited importance when merits of RCTs are discussed. The focus of natural field experiments rests on directly contacting and disciplining citizens. Public authorities are advancing their communication with taxpayers asking them to act responsibly and to adhere to a ‘contract’ that exists between taxpayers and tax authorities. However, invoking the image of lawful contractual behaviour might result in a boomerang effect when tax policies are more broadly are considered: Reminding those citizens identified as

M4716-STRAßHEIM_9781785367847_t.indd 280

25/04/2019 13:32

The governance of behavioural taxation  281 tardy taxpayers of their obligation to pay outstanding taxes appears to be a rather trivial problem for public finances compared to the larger picture of substantially underfunded public entities. Nudging becomes even more questionable when globally organized tax avoidance is considered. In fact, why should individuals feel compelled to adhere to ethical considerations uttered by public authorities when the same institutions privilege rich individuals and large commercial entities by allowing them to pay little or virtually no taxes? Moralization cuts both ways: If it is only used to urge tardy taxpayers to comply while resourceful actors are offered concessions in the form of reduced taxes, loopholes or are turned on a blind eye, the legitimacy of the state and its capacity to act might be further undermined. Abandoning the principle of fair and equal treatment might backfire and further weaken public authority in the long run. Further research based on natural field experiments might therefore benefit substantially if academics, nudging experts, public authorities and development agencies not only considered short-term gains in advanced payments and bureaucratic effectiveness. Otherwise, unintended consequences of nudging might pose more severe challenges than randomization and external validity suggest. This is particularly pressing in the current era of the neoliberal state, which increasingly relies on disciplining its citizens and tends to encourage simple cost-benefit consideration and short-term pecuniary gains of public and private actors alike. Instead of increasingly turning to discipline individual citizens, tax authorities need to be aware of the risk of alienating its citizenry further if individuals find themselves to be the object of randomized control trials to marginally improve the state’s tax base while high net worth individuals and large corporations evade and avoid paying taxes wherever possible. If nudging is to play a positive role in tax collection, its proponents should advise a cautious application of field experiments and make clear that it is merely an add-on to and not a substitute for a strong, viable tax state.

NOTES 1. I thank Judith Hesselmann for excellent research assistance and Britta Plote for language editing. 2. The idea of optimal taxation has been prominently established by Peter Diamond and James Mirrlees (1971a, 1971b) who aimed at identifying a taxation regime which distorts economic activity as little as possible – at least in theory acknowledging a gap between academic research and practical policy advice (see also Weinzierl 2014). 3. More recently, nudging activities that aim to engage with taxpayers in a more playful manner have also received increasing attention: A number of jurisdictions have introduced so called tax lotteries to incentivise taxpayers to fulfil their obligations on time. The basic idea is that on-time payments qualify to participate in a public lottery that promises high prizes at a low probability of winning (EU 2016, p. 27ff.). 4. The World Bank uses RCTs to engage in a variety of policy domains, such as health and fertility issues or conditional cash transfers (Macours et al. 2008; see also Fernald et al. 2008; Paxson and Schady 2010). 5. Text mailed to taxpayers read as follows: ‘Did you know that if you do not pay the CVP [Property Tax or Tasa de Conservación de la Via Pública] on time for a debt of AR $1,000 you will have to disburse AR $268 in arrears at the end of the year and the Municipality can take administrative and legal action?’ (Castro and Scartascini 2015, p. 72).

M4716-STRAßHEIM_9781785367847_t.indd 281

25/04/2019 13:32

282  Handbook of behavioural change and public policy

REFERENCES Allingham, Michael G. and Agnar Sandmo (1972), ‘Income tax evasion: a theoretical analysis’, Journal of Public Economics, 1 (3‒4), 323‒338. Alm, James and Benno Torgler (2006), ‘Cultural differences and tax morale in the United States and Europe’, Journal of Economic Psychology, 27 (2), 224‒246. Alm, James, G.H. McClelland and W.D. Schulze (1992), ‘‘Why do people pay taxes?’, Journal of Public Economics, 48 (1), 21–38. BIT (2016), ‘Sugar tax: how will it affect behaviour?’, accessed 10 December 2016 at http://www.behaviouralin​ sights.co.uk/health/behaviour-change-and-the-new-sugar-tax/. Brockmann, Hilke; Philipp Genschel and Laura Seelkopf (2015), ‘Happy taxation: increasing tax compliance through positive rewards?’, Journal of Public Policy, 36 (3), 381‒406. Castro, Lucio and Carlos Scartascini (2015), ‘Tax compliance and enforcement in the pampas. Evidence from a field experiment’, Journal of Economic Behavior & Organization, 116 (C), 65–82. Chetty, Raj; Adam Looney and Kory Krof (2009), ‘Salience and taxation: theory and evidence’, American Economic Review, 99 (4), 1145–1177. Choo, C.Y. Lawrence, Miguel Fonseca and Gareth D. Myles (2016), ‘Do students behave like real taxpayers in the lab? Evidence from a real effort tax compliance experiment’, Journal of Economic Behavior & Organization, 124 (C), 102‒114. Diamond, Peter A. and James A. Mirrlees (1971a), ‘Optimal taxation and public production I: production efficiency’, The American Economic Review, 61 (1), 8–27. Diamond, Peter A. and James A. Mirrlees (1971b), ‘Optimal taxation and public production II: Tax rules’, The American Economic Review, 61 (3), 261–78. Dietsch, Peter and Thomas Rixen (eds) (2016), Global Tax Governance. What is Wrong with It and How to Fix It, Colchester: ECPR Press. Dwenger, Nadja, Henrik Kleven, Imran Rasul and Johannes Rincke (2016), ‘Extrinsic and intrinsic motivations for tax compliance: evidence from a field experiment in Germany’, American Economic Journal: Economic Policy, 8 (3), 203‒232. European Commission (2016), ‘Anti tax avoidance package’, accessed 10 December 2016 at http://ec.europa.eu/ taxation_customs/business/company-tax/anti-tax-avoidance-package_en. European Parliament (2016), ‘Tax evasion: MEPs back automatic exchange of bank data to track account owners’, accessed 13 December 2016 at http://www.europarl.europa.eu/news/en/news-room/20161117IPR51550/ tax-evasion-meps-back-automatic-exchange-of-bank-data-to-track-account-owners. European Union (EU) (2016), ‘Behavioural insights applied to policy: European Report 2016’, accessed at 8 December 2016 at http://publications.jrc.ec.europa.eu/repository/bitstream/JRC100146/kjna27726enn_new. pdf. Feld, Lars P. and Bruno S. Frey (2002), ‘Trust breeds trust: how taxpayers are treated’, CESifo Working Paper Series No. 322; Zurich IEER Working Paper No. 98. Fernald, Lia C.H., Paul J. Gertler and Lynnette M. Neufeld (2008), ‘Role of cash in conditional cash transfer programmes for child health, growth, and development: an analysis of Mexico’s Oportunidades’, The Lancet, 371 (9615), 828–837. Gamble, Andrew (2006), ‘Two faces of neo-liberalism’, in Richard Robison (ed.), The Neoliberal Revolution. Forging the Market State, Basingstoke, Hampshire: Palgrave Macmillan UK, 20‒35. Hallsworth, Michael (2014), ‘The use of field experiments to increase tax compliance’, Oxford Review of Economic Policy, 30 (4), 658–679. Hilton, Denis, Laetitia Charalambides; Christophe Demarque; Laurent Waroquier and Charles Raux (2014), ‘A tax can nudge: the impact of an environmentally motivated bonus/malus fiscal system on transport preferences’, Journal of Economic Psychology, 42 (C), 17‒27. HM Revenue & Customs (2016), ‘Measuring tax gaps 2016 edition. Tax gap estimates for 2014‒15’, accessed 8 December 2016 at https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/561312/ HMRC-measuring-tax-gaps-2016.pdf#page=4&zoom=auto,-82,392. IRS (2016), ‘Tax gap estimates for tax years 2008–2010’, Internal Revenue Service, accessed 10 December 2016 at https://www.irs.gov/PUP/newsroom/tax%20gap%20estimates%20for%202008%20through%202010.pdf. Kahneman, Daniel and Amos Tversky (1979), ‘Prospect theory: an analysis of decision under risk’, Econometrica, 47 (2), 263‒292. Kahneman, Daniel; Jack L. Knetsch and Richard H. Thaler (1986), ‘Fairness and the assumptions of economics’, The Journal of Business, 59 (4), 285‒300. Kettle, Stewart, Marco Hernandez, Simon Ruda and Michael Sanders (2016), ‘Behavioral interventions in tax compliance: evidence from Guatemala’, Policy Research Working Papers, Nr. 7690, World Bank Group, June 2016.

M4716-STRAßHEIM_9781785367847_t.indd 282

25/04/2019 13:32

The governance of behavioural taxation  283 Kirchler, Erich, Erik Hoelzl and Ingrid Wahl (2008), ‘Enforced versus voluntary tax compliance: the “slippery slope” framework’, Journal of Economic Psychology, 29 (2), 210‒225. Kleven, Henrik J., Martin B. Knudsen, Claus T. Kreiner, Søren Pedersen and Emmanuel Saez (2010), ‘Unwilling or unable to cheat? Evidence From a randomized tax audit experiment in Denmark’, NBER Working Paper 15769, February 2010. Kosonen, Tuomas and Olli Ropponen (2015), ‘The role of information in tax compliance: evidence from a natural field experiment’, Economics Letters, 129 (C), 18–21. Luttmer, Erzo F.P. and Monica Singhal (2014), ‘Tax morale’, Journal of Economic Perspectives, 28 (4), 149–168. Macours, Karen, Norbert Schady and Renos Vakis (2008), ‘Cash transfers, behavioral changes, and cognitive development in early childhood: evidence from a randomized experiment’, IDP Working Papers Series No. IDP-WP-301, February 2008. Mudge, Stephanie Lee (2008), ‘What Is neo-liberalism?’, Socio-Economic Review, 6 (4), 703‒731. Nesterak, Evan (2015), ‘The Behavioral Insights Team 5 years on: a conversation with Owain Service’, The Psych Report, 3 November 2015, accessed 10 December 2016 at http://thepsychreport.com/politics/ the-behavioral-insights-team-5-years-on-a-conversation-with-owain-service/. Nonhoff, Martin and Frieder Vogelmann (2016), ‘Paying for identity. the formation of differentiated collectives through taxes’ InIIS Working Papers Series No. 41/16, University of Bremen, Institute for Intercultural and International Studies. OECD (2016a), ‘Revenue statistics – OECD countries: comparative tables’, accessed 13 December 2016 at https://stats.oecd.org/Index.aspx?DataSetCode=REV. OECD (2016b), ‘Base erosion and profit shifting’, accessed 20 December 2016 as http://www.oecd.org/tax/beps/. Paxson, Christina and Norbert Schady (2010), ‘Does money matter? The effects of cash transfers on child development in rural Ecuador’, Economic Development and Cultural Change, 59 (1), 187‒229. Pickhardt, Michael and Aloys Prinz (2014), ‘Behavioral dynamics of tax evasion – a survey’, Journal of Economic Psychology, 40 (C), 1–19. Ritsatos, Titos (2014), ‘Tax evasion and compliance: from the neo classical paradigm to behavioural economics, a review’, Journal of Accounting & Organizational Change, 10 (2), 244–262. Sandmo, Agnar (2005), ‘The theory of tax evasion: a retrospective view’, National Tax Journal, 58 (4), 643‒663. Schneider, Friedrich, Andreas Bruehn and Claudio E. Montenegro (2010), ‘Shadow economies all over the world: new estimates for 162 countries from 1999 to 2007’, Policy Research Working Papers 5356, The World Bank, July 2010. Slemrod, Joel B. (2016), ‘Tax compliance and enforcement: new research and its policy implications’, Ross School of Business Paper No. 1302, 16 January 2016. Slemrod, Joel B. and Caroline Weber (2012), ‘Evidence of the invisible: toward a credibility revolution in the empirical analysis of tax evasion and the informal economy’, International Tax and Public Finance, 19 (1), 25–53. Streeck, Wolfgang (2014), Buying Time: The Delated Crisis of Democratic Capitalism, London, New York: Verso. Swank, Duane and Sven Steinmo (2002), ‘The new political economy of taxation in advanced capitalist democracies’, American Journal of Political Science, 46 (3), 642‒655. Thaler, Richard H. and Cass Sunstein (2003), ‘Libertarian paternalism’, The American Economic Review, 93 (2), 175–179. Thaler, Richard H., Cass R. Sunstein and John P. Balz (2013), ‘Choice architecture’, in Eldar Shafir (ed.), The Behavioral Foundations of Public Policy, Princeton, New Jersey: Princeton University Press, 428–439. World Bank (2016), ‘Behavioral initiatives: current projects’, accessed 13 December 2017 at http://www.world​ bank.org/en/topic/poverty/brief/behavioral-initiatives-what-we-do#learners. Torgler, Benno (2003), ‘Tax morale, rule-governed behaviour and trust’, Constitutional Political Economy, 14 (2), 119‒140. Torgler, Benno (2011), ‘Tax morale and compliance: review of evidence for case studies in Europe’, Policy Research Working Papers 5922, the World Bank, December 2011. Weinzierl, Matthew (2014), ‘The promise of positive optimal taxation: normative diversity and a role for equal sacrifice’, Journal of Public Economics, 118 (C), 128‒142.

M4716-STRAßHEIM_9781785367847_t.indd 283

25/04/2019 13:32

PART IV THE GOVERNANCE OF BEHAVIOUR: NORMATIVE IDEALS, CRITICAL PERSPECTIVES AND POLITICAL CONSEQUENCES

M4716-STRAßHEIM_9781785367847_t.indd 284

25/04/2019 13:32

20.  Which nudges do people like? A national survey Cass R. Sunstein1

1. INTRODUCTION Nudges are interventions that preserve freedom of choice but that nonetheless influence people’s decisions. A reminder is a nudge;2 so is a warning. By steering people in certain directions, a GPS nudges; a default rule nudges. Disclosure of relevant information (about the risks of smoking or the costs of borrowing) counts as a nudge. Save More Tomorrow plans, allowing employees to sign up to give some portion of their future earnings to pension programmes, are nudges.3 A recommendation is a nudge. A criminal penalty, a civil fine, and a subsidy are not nudges, because they impose significant material incentives on people’s choices.4 Some people have raised ethical objections to nudges.5 On one view, some nudges compromise individual dignity. Acknowledging that they might prove helpful, Jeremy Waldron writes: ‘I wish, though, that I could be made a better chooser rather than having someone on high take advantage (even for my own benefit) of my current thoughtlessness and my shabby intuitions’6 On another view, some nudges are more coercive than they seem, precisely because of their potentially large impact; because of inertia, people might stick with default rules, which can in that sense operate like mandates.7 Some people contend that it is important to ‘boost’ people’s own capacities, by increasing their knowledge and their agency, rather than to nudge them.8 There are also pervasive questions about those who design nudges. Should they really be trusted?9 Who will nudge, or control, them? These and other questions raise serious problems, which deserve extended treatment.10 Ethical issues are normative ones, of course, but there is also a positive question: Which nudges do people endorse, and which do they reject? Is it possible to develop principles by which to organize people’s judgments? To be sure, people’s responses to survey questions cannot be considered authoritative. If the issue is how to resolve those questions in principle, empirical findings about people’s answers are not decisive. In any case, their answers might not reflect their considered judgments. When people have time to think, and when they are informed, they might respond differently from how they do on surveys. Moreover, intuitions and considered judgments might go wrong.11 Behavioural scientists would emphasize a related point: People’s answers to ethical questions, or questions about moral approval or disapproval, might well depend on how such questions are framed; slight differences in framing can yield dramatically different answers.12 But surveys certainly provide important clues about what people are likely to think. In a democracy, their responses undoubtedly matter in practice, if only because public officials are responsive to citizens. Such officials are likely to hesitate before proceeding with nudges that strike large numbers of citizens as troublesome or unethical. In addition, it is unnecessary to make strong claims about the wisdom of crowds, especially on normative issues, in order to believe that an ethical judgment, on the part of those who might 285

M4716-STRAßHEIM_9781785367847_t.indd 285

25/04/2019 13:32

286  Handbook of behavioural change and public policy be subject to nudges, deserves respectful attention. Public officials should be humble and attentive to the views of others, and if strong majorities favour or oppose nudges, then their views ought to be taken into account. The goal of this chapter is to report on a nationally representative survey involving thirty-four nudges.13 The survey involved 563 Americans, with a margin of error of plus or minus 4.1 per cent. From their responses, two dominant principles emerge. First, Americans reject nudges that promote what they see as illicit ends (such as religious favouritism). Second, Americans reject nudges that they view as inconsistent with the interests or values of most choosers. By contrast, there is widespread support for nudges that are taken to have legitimate ends and to be consistent with the interests and the values of most choosers. It follows that numerous nudges – default rules, warnings, and tendentious public education campaigns – are likely to attract bipartisan support, so long as people approve of their ends, and think that they are consistent with choosers’ values and interests. Notably, Americans do not see nudges as unacceptably manipulative, with the single (and highly exotic) exceptions of subliminal advertising (which, surprisingly, receives substantial minority support in the context of efforts to combat smoking and overeating). As we will see, political divisions sometimes affect the level of support, because Democrats tend to be somewhat more favourably disposed towards health and safety nudges than Republicans. And in cases that raise strong partisan differences, such divisions will map onto nudges as well.14 But across a wide range, clear majorities of Democrats and Republicans (and also Independents) are in full agreement about what they support – and what they reject.

2.  POPULAR NUDGES In recent years, the US federal government has adopted or promoted a large number of nudges.15 Three of the most prominent include (1) mandatory calorie labels at chain restaurants;16 (2) mandatory graphic warnings on cigarette packages17 (struck down by a federal court of appeals18); and (3) automatic enrolment in savings plans, subject to opt out.19 The nationally representative sample found substantial majority support for all three policies, including support for (3) regardless of whether it consists of federal ‘encouragement’ of such enrolment or a federal mandate for automatic enrolment, imposed on large employers. About 87 per cent of Americans favoured calorie labels20 and 74 per cent favoured graphic warnings (see Table 20.1). Both policies had strong majority support from Democrats, Republicans and Independents. Overall, 80 per cent and 71 per cent respectively approved of and encouraged mandatory enrolment in savings plans. Here as well, all three groups showed strong majority support.21 Three educational campaigns also attracted widespread approval (see Table 20.2). Respondents were overwhelmingly supportive of a public education campaign from the US federal government to combat childhood obesity (82 per cent approval, again with strong support from Democrats, Republicans, and Independents). They were highly supportive of a public education campaign from the federal government designed to combat distracted driving, with graphic stories and images (85 per cent approval). About 75 per cent of people favoured a federal education campaign to encourage people not

M4716-STRAßHEIM_9781785367847_t.indd 286

25/04/2019 13:32

Which nudges do people like?  287 Table 20.1  American attitudes towards prominent recent nudges Calorie labels

Total support  (in percentages) Democrats Independents Republicans

Graphic warnings Federal (cigarettes) encouragement: auto-enrolment

Federal mandate: auto-enrolment

87/13

74/26

80/20

71/29

92/8 88/12 77/23

77/23 74/26 68/32

88/12 75/25 73/27

78/22 67/33 62/38

Table 20.2  American attitudes towards five educational campaigns

Total  support (in percentages) Democrats Independents Republicans

Childhood obesity

Distracted driving

Sexual orientation discrimination

Movie theatres

Animal Obesity welfare (arguably society manipulative)

82/18

85/15

75/25

53/47

52/48

57/43

90/11 81/19 70/30

88/12 84/16 80/20

85/15 75/25 57/43

61/39 51/49 41/59

59/41 55/45 34/66

61/40 60/40 47/53

Note:  Percentages may not total 100 due to rounding.

to discriminate on the basis of sexual orientation, though here there was a noteworthy division across party lines (85 per cent of Democrats, 57 per cent of Republicans, and 75 per cent of Independents). Three other educational campaigns attracted majority support, but at significantly lower levels, and with only minority approval from Republicans. About 53 per cent of Americans favoured a federal requirement that movie theatres run public education messages to discourage people from smoking and overeating. Democrats showed higher approval ratings than Republicans (61 per cent as opposed to 41 per cent, with Independents at 51 per cent). By a very small majority (52 per cent), Americans supported a public education campaign, by the federal government itself, to encourage people to give money to the Animal Welfare Society of America (a hypothetical organization) (59 per cent of Democrats, 34 per cent of Republicans, and 55 per cent of Independents; party was a statistically significant factor). This latter finding seems surprising; it could not easily be predicted that respondents would want their government to design a campaign to promote donations to an animal welfare society. About 57 per cent of people supported an aggressive public education campaign from the federal government to combat obesity, showing obese children struggling to exercise, and also showing interviews with obese adults, who are saying such things as, ‘My biggest regret in life is that I have not managed to control my weight’, and ‘To me, obesity is like a terrible curse.’ This question was designed to test people’s reactions to a tendentious and

M4716-STRAßHEIM_9781785367847_t.indd 287

25/04/2019 13:32

288  Handbook of behavioural change and public policy Table 20.3  American attitudes towards environmental and public health nudges

Total  support (in percentages) Democrats Independents Republicans

GMO labels

Salt labels

Healthy food placement

Traffic Lights

Organ donor Encouragement: Mandate: choice Green energy Green energy

86/14

73/27

56/44

64/36

70/30

72/28

67/33

89/11 87/13 80/20

79/21 72/28 61/39

63/37 57/43 43/57

71/29 61/39 57/43

75/25 69/31 62/38

82/18 66/34 61/39

79/21 63/37 51/49

arguably manipulative campaign, which might have been expected to receive widespread disapproval, as it did not. Here there was a significant disparity between Democrats (61 per cent approval) and Independents (60 per cent approval) on the one hand and Republicans on the other (47 per cent approval); the difference between Democrats’ and Republicans’ views was statistically significant. Most Americans were also supportive of efforts to use choice architecture to promote public health and environmental protection (see Table 20.3). In recent years, there has been considerable discussion of ‘traffic lights’ systems for food, which would use the familiar red, yellow, and green to demarcate health rankings.22 In the United States, the national government has shown no official interest in these initiatives, but with respondents in the nationally representative survey, the idea attracted strong support (64 per cent). There was also majority approval of automatic use of ‘green’ energy providers, subject to opt out23 – perhaps surprisingly, with support for automatic use of green energy whether it consisted of federal ‘encouragement’ (72 per cent) or instead a federal mandate on large electricity providers (67 per cent). In these cases, there were significant differences across partisan lines, but majorities of Democrats, Republicans, and Independents were all supportive. Most respondents were in favour of requiring companies to disclose whether the food they sell contains genetically modified organisms (GMOs) (86 per cent approval).24 There was strong majority support (73 per cent) for a mandatory warning label on products that have unusually high levels of salt, as in, ‘This product has been found to contain unusually high levels of salt, which may be harmful to your health.’ Perhaps surprisingly, most respondents (but not most Republicans) approved of a state requirement that grocery stores put their most healthy foods in prominent, visible locations (56 per cent approval; 63 per cent from Democrats, 43 per cent from Republicans, 57 per cent from Independents). Respondents also supported a state requirement that people must say, when they obtain their drivers’ license, whether they want to be organ donors (70 per cent approval; 75 per cent from Democrats, 62 per cent from Republicans, 69 per cent from Independents).25 For all of these policies, the differences between Democrats and Republicans were statistically significant. Five other forms of choice architecture, expected to be more controversial, also obtained majority support. The first would list the name of the incumbent politician first on every ballot (see Table 3.4). It might be expected that this pro-incumbent nudge would be widely rejected, because respondents might not want the voting process to be skewed in favour of incumbents. But a bare majority (53 per cent) approved of this approach,

M4716-STRAßHEIM_9781785367847_t.indd 288

25/04/2019 13:32

Which nudges do people like?  289 Table 20.4  American attitudes towards some potentially provocative nudges

Total support   (in percentages) Democrats Independents Republicans

Listing incumbent politician first

Automatic voter registration

Husband’s last name

Mandatory manufacturing label: labour violations

Mandatory manufacturing label: aiding terrorists

53/47

53/47

58/42

60/40

54/46

58/42 51/49 47/53

63/37 50/50 39/61

61/40 56/44 57/43

67/33 57/43 50/50

56/44 49/51 58/42

Note:  Percentages may not total 100 due to rounding.

perhaps because most people believed that it would promote clarity, perhaps because they did not see the risk of bias from order effects. There was also majority approval (53 per cent) for the approach, recently adopted in Oregon, of automatically registering eligible citizens as voters, subject to opt-out.26 Interestingly, most Republicans (61 per cent) rejected this approach. One reason might be that they believe that people who do not take the time to register to vote ought not to be counted as voters. Another reason is that they might believe that Oregon’s approach would favour Democrats. By a modest majority, most people (58 per cent) also approved of an approach by which women’s last names would automatically be changed to that of their husband, subject to opt-out. This approach obtained majority support from Democrats, Republicans, and Independents. This result is especially noteworthy in view of the fact that an approach to this effect would almost certainly be unconstitutional as a form of sex discrimination, even if it tracked behaviour and preferences.27 We might expect a difference between men and women on this question, but notably, 58 per cent of both groups approved of this approach. Finally, there was majority support for a federal labelling requirement for products that come from companies that have repeatedly violated the nation’s labour laws (such as laws requiring occupational safety or forbidding discrimination), as in, ‘This product is made by a company that has repeatedly violated the nation’s labor laws.’ About 60 per cent of participants supported that policy, with a significant difference between Democrats (67 per cent approval) and Republicans (50 per cent approval). There was also majority support for federally required labels on products that come from countries that have recently harboured terrorists, as in, ‘This product comes from a nation that was recently found to harbor terrorists.’ This approach attracted 54 per cent approval – 56 per cent from Democrats, 58 per cent from Republicans, and 49 per cent from Independents.

3.  UNPOPULAR NUDGES By contrast, twelve nudges were widely disapproved. Of these, seven involved uses of default rules. Two of these defaults were designed so as to be not merely provocative but also highly offensive, and strong majorities took them exactly as they were designed.

M4716-STRAßHEIM_9781785367847_t.indd 289

25/04/2019 13:32

290  Handbook of behavioural change and public policy Table 20.5  Unpopular defaults Democrat Christian registration on census Total support   (in percentages) Democrats Independents Republicans

Wife’s last name

Red Cross

Animal Welfare Society

United Way

Carbon emissions charge

26/74

21/79

24/76

27/73

26/74

24/76

36/64

32/68 26/74 16/84

22/78 17/83 27/73

28/72 23/77 18/82

30/70 28/72 20/80

30/70 25/75 20/80

26/74 25/75 17/83

43/57 34/66 25/75

Under the first, a state would assume that people want to register as Democrats, subject to opt out if people explicitly say that they want to register as Republicans or Independent (see Table 20.5). The overwhelming majority of people rejected this approach (26 per cent total approval; 32 per cent of Democrats, 16 per cent of Republicans, and 26 per cent of Independents, with statistically significant differences between Democrats and Republicans). The second was a state law assuming that people are Christian, for purposes of the census, unless they specifically state otherwise. (This initiative can be seen as a nudge insofar as it is a default rule used for public purposes, and it might well steer behaviour, if only indirectly.) Here too there was widespread disapproval (21 per cent overall approval; 22 per cent of Democrats, 27 per cent of Republicans, 17 per cent of Independents). The third unpopular default rule involved a state law assuming that upon marriage, husbands would automatically change their last names to that of their wives, subject to opt out (24 per cent total approval; 28 per cent of Democrats, 18 per cent of Republicans and 23 per cent of Independents). Interestingly, there was no gender disparity; 24 per cent of both men and women approved. With the fourth, the federal government would assume, on tax returns, that people want to donate $50 to the Red Cross, subject to opt out if people explicitly say that they do not want to make that donation (27 per cent approval; 30 per cent of Democrats, 20 per cent of Republicans, 28 per cent of Independents). The fifth was identical but substituted the Animal Welfare Society for the Red Cross. Not surprisingly, that question also produced widespread disapproval (26 per cent approval; 30 per cent of Democrats, 20 per cent of Republicans, and 25 per cent of Independents). With the sixth, state government assumed that state employees would give $20 per month to the United Way, subject to opt out. It might be expected that because state government and state employees were involved, approval rates would grow. But they did not (24 per cent approval; 26 per cent of Democrats, 17 per cent of Republicans, and 25 per cent of Independents). With the seventh, a majority (64 per cent) disapproved of a federal requirement that airlines charge people, with their airline tickets, a specific amount to offset their carbon emissions (about $10 per ticket), subject to opt out if passengers said that they did not want to pay. The five other unpopular nudges involved information and education (see Table 20.6). With the first, a newly elected president adopted a public education campaign designed to convince people that criticism of his decisions is unpatriotic and potentially damaging to national security. There was overwhelming disapproval of this campaign (23 per cent approval; 24 per cent of Democrats, 21 per cent of Republicans, 22 per cent

M4716-STRAßHEIM_9781785367847_t.indd 290

25/04/2019 13:32

Which nudges do people like?  291 Table 20.6  Unpopular education campaigns and disclosure Unpatriotic Stay-at-homecriticism mothers

Total support   (in percentages) Democrats Independents Republicans

Subliminal advertising

Mandatory manufacturing label: Communism

Transgender

23/77

33/67

41/59

44/56

41/59

24/76 22/78 21/79

33/67 34/67 31/69

47/53 35/65 42/58

47/53 42/58 43/57

49/51 38/62 29/71

Note:  Percentages may not total 100 due to rounding.

of Independents). With the second, the federal government adopted a public education campaign designed to convince mothers to stay home to take care of their young children. Over two-thirds of respondents rejected this nudge (33 per cent approval; 33 per cent of Democrats, 31 per cent of Republicans, 34 per cent of Independents). The third involved a government requirement that movie theatres run subliminal advertisements to discourage smoking and overeating. Here too, there was majority disapproval (41 per cent approval; 47 per cent of Democrats, 42 per cent of Republicans, 35 per cent of Independents). It is noteworthy and surprising, however, that over two-fifths of people actually supported this requirement. With the fourth, the federal government would require all products that come from a Communist country (such as China or Cuba) to be sold with the label, ‘Made in whole or in part under Communism.’ Slightly over half of respondents disapproved of this requirement (44 per cent approval; 47 per cent of Democrats, 43 per cent of Republicans, 42 per cent of Independents). With the fifth, a majority (59 per cent) also rejected a public education campaign from the federal government, informing people that it is possible for people to change their gender from male to female or from female to male, and encouraging people to consider that possibility ‘if that is really what they want to do.’ It is perhaps surprising that this somewhat adventurous campaign was endorsed by 41 per cent of respondents; note that approval rates differed between Democrats (49 per cent) and Republicans (29 per cent; Independents, 38 per cent).

4.  WHY ARE SOME NUDGES UNPOPULAR? What separates the approved nudges from the rejected ones? Two principles seem to dominate the cases. First, people reject nudges that are taken to have illicit motivations. In a democracy, it is illegitimate to attempt to convince people that criticism of a public official is unpatriotic. At least in the United States, nudges that favour a particular religion or party will meet with widespread disapproval, even among people of that very religion or party.28 In this light, there should be no puzzle that people disapprove of an attempt, by a president of the United States, to nudge citizens to believe that it is unpatriotic to criticize him. This simple principle justifies a prediction: Whenever people do not like the

M4716-STRAßHEIM_9781785367847_t.indd 291

25/04/2019 13:32

292  Handbook of behavioural change and public policy motivations of the choice architect, they will disapprove of the nudge. To be sure, that prediction might not exactly seem surprising, but it suggests that people will not oppose (for example) default rules and warnings as such; everything will turn on what they are nudging people towards.29 When there are partisan differences in judgments about nudges, it is often because of partisan disagreement about whether the relevant motivations are legitimate. Second, people oppose nudges that are inconsistent with the interests or values of most choosers. The most illuminating finding here is that while most people support automatic name change for women, they reject automatic name change for men. The evident reason is that the former tracks people’s interests and values (at least in general), while the latter countermands them.30 Any default rule, of course, is likely to harm at least some people; some people will want, for good reason, to opt out, and some people who want to opt out will not do so, perhaps because of inertia and procrastination. By itself, however, that fact is not enough to produce public opprobrium. Recall that there is majority approval for automatic enrolment in pension plans and green energy, and that people are also favourably disposed towards automatic voter registration (and also automatic namechanging by women). The size of the group of the people who are harmed by a default rule undoubtedly matters. If the group is large (but not a majority), people might reject a default rule and favour active choosing instead. By contrast, strong majorities of people reject automatic charitable donations of several different kinds. The apparent concern is that as a result of inertia, procrastination, or inattention, people might find themselves giving money to a charity even though they do not wish to do so. We might therefore complement the second principle with a third and narrower one, which can be seen as a corollary: Most people reject automatic enrolment in charitable giving programmes, at least if they are operated by public institutions. The case of carbon offsets can be understood in similar terms.31 We do not yet know the exact boundaries of apparent public scepticism about default rules that would give away people’s money without their active consent,32 but there is no doubt that such scepticism exists. We have seen that people generally favour disclosures that, in their view, bear on health and safety (salt content, GMOs). At the same time, the results leave open the question whether and when people will favour mandatory disclosures that involve political issues associated with production of a product rather than the health and environmental effects of the product itself. Americans seem closely divided on that question. With repeated violations of the nation’s labour laws, and nations that harbour terrorism, such disclosure achieved majority support – but not with products coming from Communist nations. People might well demand a certain threshold of egregiousness, in terms of the behaviour of those who produce a good or service, before they will want to require disclosure of that behaviour. On this question, partisan differences are to be expected, because people will disagree about whether the relevant threshold has been met, and about what it exactly is. It is tempting, and not inconsistent with the data, to suggest that people’s reactions to nudges also show the influence of a fourth principle: People reject nudges that they regard as unambiguously manipulative.33 The subliminal advertising finding can be taken as support for this principle. But what counts as unambiguous manipulation? Most people are in favour of graphic warning labels on cigarettes; they like default rules (if consistent with people’s values and interests); a majority favours a mandatory cafeteria design to promote healthy eating; people approve of a graphic campaign to discourage distracted driving;

M4716-STRAßHEIM_9781785367847_t.indd 292

25/04/2019 13:32

Which nudges do people like?  293 with respect to obesity, a majority favours a somewhat tendentious public education, one that could plausibly be characterized as manipulative. No one likes manipulation in the abstract, but there do not appear to be many cases in which people are willing to reject nudges as unacceptably manipulative, at least if they have legitimate ends and are taken to be in the interest of most choosers. There is a final question: What is the role of partisan differences? Democrats and Republicans will sometimes disagree, of course, about whether the goals of a particular nudge are illicit, and they will also disagree, on occasion, about whether a nudge is consistent with the interests or values of choosers.34 For example, those who disapprove of abortion will be especially likely to support nudges that are designed to discourage abortion; those who do not disapprove of abortion will be unlikely to support such nudges. But there is another and more general division as well. Even when majorities of Democrats, Republicans, and Independents support a particular initiative, the level of support is sometimes higher within one group than within another.35 Even if the underlying end is broadly shared – as it is, for example, in the area of public health – many Republicans sometimes seem sceptical of government nudges, taken as such, and will therefore disapprove of them even if they do accept the legitimacy of the end and do not think that the nudge is inconsistent with choosers’ interests or values. Some Republicans, and undoubtedly some Democrats and Independents, support another principle: There should be a rebuttable presumption against nudging, at least if the government can avoid it. It is important to see that the strength of the presumption will vary with the particular issue, with partisan affiliations, and with competing views about the role of government.36 In some of the cases, Republicans are more sceptical of nudges than are Democrats. With calorie labels and childhood obesity campaigns, for example, there are significant differences in the levels of support within the two groups, even though majorities of both are supportive. But in some cases, Republicans are more enthusiastic about nudges than are Democrats.37 The fact that no such cases are found here is an artefact of the particular questions. If the issue involved automatic enrolment in programmes by which high-income earners automatically receive capital gains tax benefits, for example, we can predict, with some confidence, that Republicans would be more supportive than Democrats.38 The potential existence of partisan differences is important, and it complicates the basic story I have presented here. But it does not conflict with the larger lesson: Across a wide range of domains, clear majorities of Democrats, Republicans, and Independents are in agreement about which nudges deserve support – and about which do not. They are also in fundamental agreement about the principles that underlie those judgments, even if those principles occasionally lead to divisions in particular cases.

NOTES   1. I am most grateful to Arevik Avedian for indispensable help and wisdom with the surveys conducted here, and also to Heidi Liu for superb comments and research assistance. Maya Bar-Hillel, Lucia Reisch, and Richard Thaler provided valuable thoughts and suggestions. For support of multiple kinds, thanks to Harvard Law School’s Program on Behavioral Economics and Public Policy.  2. See Raifman et al. (2014).  3. See Thaler (2015).   4. On some of the complexities here, see generally Sunstein (2014).

M4716-STRAßHEIM_9781785367847_t.indd 293

25/04/2019 13:32

294  Handbook of behavioural change and public policy  5.  6.  7.  8.

See generally Rebonato (2012). See Waldron (2014). See generally Rebonato (2012), supra note 4. See generally Grüne-Yanoff and Hertwig (2015). In my view, boosts are best understood as a subcategory of nudges, but it is certainly true that some nudges boost people’s agency, whereas other nudges do not do so. See Sunstein (2016).  9. See id. 10. Some efforts are made in Sunstein (2015 and 2016). 11. See Sunstein (2005). 12. See Frederick (2003). 13. In an earlier paper, I reported on a much smaller study, with five nudges overlapping with those explored here. The results are within the margin of error of plus or minus 4.5 per cent. See Sunstein (2016). Note that the automatic enrolment questions, in that essay, did not specify involvement from the government. A valuable study, with fewer nudges but results similar to those here, is William Hagman et al. (2015). Also valuable, and with illuminating twists, is Tannenbaum et al. (2017). 14. See id. 15. For examples, see generally Sunstein (2013a). 16. 79 Fed. Reg. 71156. 17. 75 Fed. Reg. 69524. 18. On the FDA’s effort to require graphic warnings on packages, see R.J. Reynolds Tobacco Co. v. U.S. Food & Drug Admin., 823 F. Supp. 2d 36 (D.D.C. 2011), aff’d on other grounds, 696 F.3d 1205 (D.C. Cir. 2012). 19. For discussion of relevant laws and policies, see generally Gale et al. (2009). 20. Note that there were statistically significant differences with respect to calorie labels between Republicans (77 per cent approval) and both Democrats (92 per cent approval) and Independents (88 per cent approval). 21. Here as well, there were statistically significant differences between Democrats and Republicans for both policies and between Democrats and Independents with respect to encouragement. (Encouraged: 88 per cent of Democrats, 73 per cent of Republicans, and 75 per cent of Independents. Mandated: 78 per cent of Democrats, 62 per cent of Republicans, and 67 per cent of Independents.) 22. See Thorndike et al. (2014). 23. See Sunstein and Reisch (2014). 24. In my view, this is not a good idea. See Sunstein (2013b). 25. Another study finds that most Americans reject a default rule to be the effect that people would be presumed to be organ donors, subject to opt out. Hagman et al. (2015). 26. See Berman (2015). 27. See Craig v. Boren, 429 U.S. 190, 200-204 (1978). For valuable discussion of the general topic, see Emens (2007). 28. We could, of course, imagine a nation in which favouritism on the basis of religion or party would attract widespread support, and might be seen as analogous to a default rule in which women’s last name changes to that of their husband (approved, it might be recalled, by a majority of respondents here). A default rule in favour of the most popular party, or the dominant religion, might be taken to track people’s preferences and values, and not to be a violation of neutrality at all. But I certainly do not mean to approve of any such default rule. 29. The striking findings of ‘partisan nudge bias’ are fully consistent with this claim. See David Tannenbaum et al. (2017). 30. We could easily imagine a population that would reverse these results. Suppose that one believes that automatically assuming that wives take their husbands’ last names undermines sex equality, and the automatically assuming that husbands take their wives’ last names promotes sex equality. For those who have these beliefs, and are committed to sex equality, reversing the majority’s views might seem attractive. 31. Framing might matter here, and note that in most Sweden, citizens are supportive. See Hagman et al. 2015, supra note. The case of climate compensation is of course different from charitable contributions, because it is designed to ensure that people pay the social cost of their own activities. 32. A natural question is whether people would reject an automatic donation programme from private employers, subject to opt-out. 33. In fact I hoped to provide general support for that principle, but was unable to do so. 34. See the discussion of partisan nudge bias in id. 35. It would of course be easy to design nudges that would show an opposite pattern, as with nudges that influence people in directions that are most favoured by Republicans. 36. Hagman, supra note, offers some interesting findings on this count. For example, those with an individualistic worldview, as such, were more likely to disapprove of nudges. See id. 37. See Tannenbaum et al. (2017), supra note. 38. Id.

M4716-STRAßHEIM_9781785367847_t.indd 294

25/04/2019 13:32

Which nudges do people like?  295

REFERENCES Berman, R. (2015),’Should Voter Registration be Automatic?’, The Atlantic, accessed 7 November 2017 at www. theatlantic.com/politics/archive/2015/03/should-voter-registration-be-automatic/388258/. Emens, E. (2007), ‘Changing Name Changing: Framing Rules and the Future of Marital Names’, University of Chicago Law Review, 74(3), 761. Frederick, S. (2003), ‘Measuring Intergenerational Time Preference: Are Future Lives Valued Less?’, 26(1), 39‒53. Gale, W., Iwry, J., John, D. and L. Walker (eds) (2009), Automatic: Changing the Way America Saves, Washington, DC.: Brookings Institution Press. Grune-Yanoff, T and R. Hertwig (2015), ‘Nudge Versus Boost: How Coherent Are Policy and Theory?’, Minds & Machines, 25, 149‒183. Hagman, W., Andersson, D., Västfjäll, D. and G. Tinghög (2015), ‘Public Views on Policies Involving Nudges‘, Review of Philosophy and Psychology, 6(3), 439‒453. Raifman, J., Landthorn, H., Rokicki, S. and G. Fink (2014), ‘The Impact of Text Message Reminders on Adherence to Antimalarial Treatment in Northern Ghana’, PLOS ONE, 9(10) e109032. Rebonato, R. (2012), Taking Liberties: A Critical Examination of Libertarian Paternalism, London: Palgrave Macmillan. Sunstein, C. (2005), ‘Moral Heuristics’, Behavioral and Brain Sciences, 28(4), 531‒542. Sunstein. C. (2013a), Simpler, New York: Simon & Schuster Paperbacks. Sunstein, C. (2013b), ‘Don’t Mandate Labeling for Gene-Altered Foods’, accessed 7 November 2017 at www. bloomberg.com/news/articles/2013-05-12/don-t-mandate-labeling-for-gene-altered-foods. Sunstein, C. (2014), Why Nudge? The Politics of Libertarian Paternalism, New Haven and London: Yale University Press. Sunstein, C. (2015), ‘Ethical Nudging’, Yale Journal on Regulation, 32(2), 413‒450. Sunstein, C. (2016), The Ethics of Influence – Government in the Age of Behavioral Science, Cambridge: Cambridge University Press. Sunstein, C. and L. Reisch (2014), ‘Automatically Green’, Harvard Environmental Law Review, 38, 127‒158. Tannenbaum , D., Fox, C. and T. Rogers (2017), ‘On the Misplaced Politics of Behavioural Policy Interventions’, Nature Human Behaviour, 1, 1‒7. Thaler, R.A. (2015), Misbehaving: The Making of Behavioural Economics, New York and London: W.W. Norton & Company. Thorndike, A., Riis, J., Sonnenberg, L. and Levy D. (2014), ‘Traffic-Light Labels and Choice Architecture’, American Journal of Preventive Medicine, 46(2), 143‒149. Waldron, J. (2014), ‘It’s All For Your Own Good’, accessed 7 November 2017 at www.nybooks.com/articles/ archives/2014/oct/09/cass-sunstein-its-all-your-own-good/.

M4716-STRAßHEIM_9781785367847_t.indd 295

25/04/2019 13:32

296

M4716-STRAßHEIM_9781785367847_t.indd 296

25/04/2019 13:32

 1.  Mandatory calorie labels  2.  Mandatory labelling: GMOs  3.  Public education campaign: distracted driving  4.  Public education campaign: childhood obesity  5.  Governmentencouraged automatic enrolment: pension plan  6.  Public education campaign: sexual orientation discrimination  7.  Mandatory graphic warnings on cigarettes  8.  Mandatory labels for high salt content  9.  Governmentencouraged automatic enrolment: green energy

Nudge

25%

26% 27% 28%

74%

73%

72%

18%

82%

75%

15%

85%

20%

14%

86%

80%

13%

Disapprove

87%

Approve

All

82%

79%

77%

85%

88%

90%

88%

89%

92%

Approve

18%

21%

23%

15%

12%

11%

12%

11%

 8%

Disapprove

Democrat

61%

61%

68%

57%

73%

70%

80%

80%

77%

Approve

39%

39%

32%

43%

27%

30%

20%

20%

23%

Disapprove

Republican

APPENDIX A: APPROVED AND DISAPPROVED NUDGES

66%

72%

74%

75%

75%

81%

84%

87%

88%

Approve

34%

28%

26%

25%

25%

19%

16%

13%

12%

Disapprove

Independent

1%

1%

1%

1%

1%

5%

1%

Significance betweenparty, all conditions

D/R (p