Uncertain Futures: Imaginaries, Narratives, And Calculation In The Economy 0198820801, 9780198820802

Table of contents :
1: Jens Beckert and Richard Bronk: An Introduction to Uncertain FuturesSection I: The Nature of Expectations in Modern Political Economies2: Robert Boyer: Expectations, Narratives, and Socio-Economic Regimes3: David Tuckett: Conviction Narrative Theory and Understanding Decision-Making in Economics and Finance4: Jenny Andersson: Arctic Futures: Expectations, Interests, Claims, and the Making of Arctic TerritorySection II: The Strange World of Economic Forecasting5: Werner Reichmann: The Interactional Foundations of Economic Forecasting6: Olivier Pilmis: Escaping the Reality Test: How Macroeconomic Forecasters Deal with 'Errors'7: Andrew G. Haldane: Uncertainty in Macroeconomic ModellingSection III: The Role of Narratives and Planning in Central Banking8: Douglas R. Holmes: A Tractable Future: Central Banks in Conversation with their Publics9: Benjamin Braun: Central Bank Planning? Unconventional Monetary Policy and the Price of Bending the Yield CurveSection IV: Constructing Futures in Finance10: Elena Esposito: Predicted Uncertainty: Volatility Calculus and the Indeterminacy of the Future11: Natalia Besedovsky: Uncertain Meanings of Risk: Calculative Practices and Risk Conceptions in Credit Rating AgenciesSection V: Managing Expectations in Innovative Business12: Martin Giraudeau: Processing the Future: Venture Project Evaluation at American Research and Development Corporation (1946-1973)13: Liliana Doganova: Discounting and the Making of the Future: On Uncertainty in Forest Management and Drug Development14: Timur Ergen: The Dilemma between Aligned Expectations and Diversity in Innovation: Evidence from Early Energy Technology Policies

Citation preview

OUP CORRECTED PROOF – FINAL, 24/5/2018, SPi

Uncertain Futures

OUP CORRECTED PROOF – FINAL, 24/5/2018, SPi

OUP CORRECTED PROOF – FINAL, 24/5/2018, SPi

Uncertain Futures Imaginaries, Narratives, and Calculation in the Economy Edited by

Jens Beckert and Richard Bronk

1

OUP CORRECTED PROOF – FINAL, 24/5/2018, SPi

3

Great Clarendon Street, Oxford, OX2 6DP, United Kingdom Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries © Oxford University Press 2018 The moral rights of the authors have been asserted First Edition published in 2018 Impression: 1 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, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: 2017964276 ISBN 978–0–19–882080–2 Printed and bound by CPI Group (UK) Ltd, Croydon, CR0 4YY Links to third party websites are provided by Oxford in good faith and for information only. Oxford disclaims any responsibility for the materials contained in any third party website referenced in this work.

OUP CORRECTED PROOF – FINAL, 24/5/2018, SPi

For Beatrice, Jasper, Justin, and Philip

OUP CORRECTED PROOF – FINAL, 24/5/2018, SPi

OUP CORRECTED PROOF – FINAL, 24/5/2018, SPi

Preface

Uncertain Futures considers how economic actors visualize the future and decide how to act in conditions of radical uncertainty. It starts from the premise that modern dynamic capitalist economies are characterized by relentless innovation and novelty, and argues that investors, policy-makers, and consumers alike face uncertain futures that cannot be reduced to measurable risk. Put simply, their future is indeterminate because it has yet to be created by the innovations and complex interdependent choices they and others will make. The organizing question of this edited volume then becomes how economic actors form expectations and make decisions despite the uncertainty they face. There has recently been a considerable renewal of interest in radical uncertainty well outside normally heterodox areas of economics, in research that builds on the insights of Frank Knight, John Maynard Keynes, F.A. Hayek, and George Shackle. This book seeks to integrate this revived epistemic tradition with exciting new work being done by economic sociologists, anthropologists, political economists, and psychologists to explore the socially and politically contingent nature of expectations in conditions of uncertainty and the role they play in capitalist dynamics. For the most part, economics remains wedded to variants of rational expectations theory, which assume that economic actors will—as a result of competitive pressures—converge on expectations that avoid systematic forecasting errors and accord with the forecasts of the most pertinent economic model. In the major fields of information economics and behavioural economics, the focus has been on bolting amendments onto this theory to deal with important knowledge problems caused by asymmetries of information and the predictable cognitive biases of economic actors. But the central assumption has remained that economic actors face measurable risk rather than radically indeterminate futures, and that this risk can be estimated in the form of objective probability functions. It is becoming increasingly clear that the current microfoundations of standard economic models cannot handle genuinely uncertain futures. To understand decision-making in such conditions requires an entirely new model of economic reasoning.

OUP CORRECTED PROOF – FINAL, 24/5/2018, SPi

Preface

This edited volume helps lay foundations for this new model by showing how economic actors in practice form expectations in conditions of uncertainty. To do so, it draws on ground-breaking research in economic sociology, economics, anthropology, and psychology to present a series of theoreticallygrounded empirical case studies that demonstrate the role of imaginaries, narratives, and calculative technologies—and their various combinations—in helping economic actors form expectations and cope with uncertain futures. The book examines calculative technologies including risk management techniques, finance models, and discounted cash-flow models, as well as other methods of envisaging the future—central bank forward guidance, economic forecasts, business plans, visions of technological futures, and new era stories—that all to varying degrees combine calculation with imaginaries and narrative structure. Considerable attention is given to how calculative models and narratives influence actors’ expectations, coordinate action, and provide the confidence to act, and how they become instruments of power in markets and societies. Since Uncertain Futures focuses mostly on empirical analysis of how economic agents actually use imaginaries, narratives, and calculation to cope with uncertain futures, it is not primarily envisaged as a normative guide to how economic actors ought to analyse uncertain futures and make decisions in the absence of foreknowledge. Nor does it claim to be a detailed practical manual for economists on how to reform their modelling practices. Nevertheless, some chapters (for example, Andrew Haldane’s on agent-based models) explicitly recommend new approaches to modelling the unknown future, and many more contain implicit lessons for policy-makers, business practitioners, and economists alike. These are drawn together in the opening chapter, which develops the theoretical framework for the empirical analyses that follow. For policy-makers and market practitioners, the indeterminacy of the future—and the consequent impossibility in many circumstances of making probabilistic predictions or knowing ex ante what the right explanatory model will be—has the disquieting implication that there is no self-evident anchor in knowledge (or objective probabilities) to govern the imaginaries and narratives shaping the beliefs and behaviour of key actors. Rational analysis still has an important role to play, of course, in stress-testing stories and expectations for long-term plausibility and feasibility; but in the short-run at least, the market or political success of any narrative may depend more on its emotional appeal, the credibility of its author, and the rhetorical techniques it employs than on its ability to capture available information about persistent constraints and emerging patterns. The expectations guiding market players are often the contingent product of novel imaginaries and a market or political battle to establish the supremacy of different narratives. Far from reflecting a knowable future, expectations (and the narratives and models viii

OUP CORRECTED PROOF – FINAL, 24/5/2018, SPi

Preface

that guide them) can be a source of novelty and indeterminacy in the economic system—the vital product of political debate, conversation, and ideational entrepreneurship. Other important lessons flowing from the studies in this book include the dangers of treating contingent attempts to narrate or calculate the future— whether new era stories, economic forecasts, or the output of finance models—as apparently objective representations of that future; and the importance of relying on judgement rather than mechanical prediction in conditions of uncertainty. The book also underlines the need for economic agents to avoid any premature coalescing around an analytical monoculture that reduces the diversity of cognitive inputs to decision-making and, when found wanting, may prove destabilizing to markets. While most standard economic theories assume that all economic agents internalize the same correct model of how the future will unfold—namely that implied by the theory concerned—such homogeneity of analysis may in the real world be a warning sign of shared cognitive myopia. For economic theorists, the main lesson implied by the book is that, when the future is demonstrably indeterminate thanks to widespread innovation and complex interdependencies, those attempting to model decision-making need to analyse the role of shared narratives and fictional expectations in guiding beliefs and behaviour. They should also be attentive to the importance of fashionable theory, political or market power, emotional contagion, and rhetoric in determining which narratives and models will succeed in performing the future and coordinating behaviour. And, finally, any economic model should be prized as much for its ability—in conjunction with others— to diagnose emerging tendencies in dynamic economic conditions as for its mathematical fit with data on past regularities. The past is not necessarily a good guide to uncertain futures. This is an unashamedly interdisciplinary book, aimed at an international audience of academics, policy-makers, and students from a wide range of disciplines. As editors, we have long shared a common interest in the role of imaginaries as both a cause of uncertainty and our main tool for coping with it. From our very different disciplinary and career backgrounds, we have found it exhilarating to work on this project with the thirteen contributing authors who also come from a wide range of disciplines and academic traditions. They bring different theoretical perspectives, methodological approaches, and practical experience to bear on a topic that is central to much of modern social science. As editors, we have had the pleasure of knowing or working with most of the contributors for a number of years. They are also well acquainted with each other’s work—not least because most attended one of two conferences that we organized on the topic. We hope that the book therefore reads like the product ix

OUP CORRECTED PROOF – FINAL, 24/5/2018, SPi

Preface

of a community of scholars who, from their various vantage points, share a fascination with how economic actors manage and exploit uncertain futures. It would be impossible to acknowledge all the debts incurred in the long gestation of this project. We do, though, wish to register our particular gratitude to W. Brian Arthur, Nicholas Barr, Bob Hancké, Abby Innes, Wade Jacoby, Waltraud Schelkle, David Stark, David Tuckett, and the anonymous reviewers for their input to our thinking and, where applicable, their comments on certain parts of the text. We are very grateful to Christine Claus and Emily Niemann of the Max Planck Institute for the Study of Societies for their help in organizing the conferences that inspired this volume and their work on the text. Our thanks extend also to the Institut d’études avancées in Paris for hosting the conference that laid much of the foundation for the project. We would like to thank Clare Kennedy, Jenny King, and Adam Swallow, our editors at Oxford University Press, for their faith in the book, and Santhosh Palani, Lynette Woodward, Yvonne Dixon, and the rest of the editorial team for guiding it through production with great care. Finally, on a personal level, we wish to thank Annelies Fryberger and Vyvian Bronk for their unfailing emotional support and intellectual inspiration. We dedicate this volume to our respective children in the hope that during their lifetimes our societies will learn new and more effective ways to navigate safely the uncertain futures that face them, while building on imaginaries that have a reasonable chance of serving the interests of the many and not only the few. Jens Beckert and Richard Bronk Cologne and London

x

OUP CORRECTED PROOF – FINAL, 24/5/2018, SPi

Contents

1. An Introduction to Uncertain Futures Jens Beckert and Richard Bronk

1

Section I: The Nature of Expectations in Modern Political Economies 2. Expectations, Narratives, and Socio-Economic Regimes Robert Boyer

39

3. Conviction Narrative Theory and Understanding Decision-Making in Economics and Finance David Tuckett

62

4. Arctic Futures: Expectations, Interests, Claims, and the Making of Arctic Territory Jenny Andersson

83

Section II: The Strange World of Economic Forecasting 5. The Interactional Foundations of Economic Forecasting Werner Reichmann 6. Escaping the Reality Test: How Macroeconomic Forecasters Deal With ‘Errors’ Olivier Pilmis 7. Uncertainty in Macroeconomic Modelling Andrew G. Haldane

105

124 144

Section III: The Role of Narratives and Planning in Central Banking 8. A Tractable Future: Central Banks in Conversation with their Publics Douglas R. Holmes

173

9. Central Bank Planning: Unconventional Monetary Policy and the Price of Bending the Yield Curve Benjamin Braun

194

OUP CORRECTED PROOF – FINAL, 24/5/2018, SPi

Contents

Section IV: Constructing Futures in Finance 10. Predicted Uncertainty: Volatility Calculus and the Indeterminacy of the Future Elena Esposito

219

11. Uncertain Meanings of Risk: Calculative Practices and Risk Conceptions in Credit Rating Agencies Natalia Besedovsky

236

Section V: Managing Expectations in Innovative Businesses 12. Processing the Future: Venture Project Evaluation at American Research and Development Corporation (1946–73) Martin Giraudeau

259

13. Discounting and the Making of the Future: On Uncertainty in Forest Management and Drug Development Liliana Doganova

278

14. The Dilemma between Aligned Expectations and Diversity in Innovation: Evidence from Early Energy Technology Policies Timur Ergen

298

Contributor Biographies Index

xii

319 323

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

1 An Introduction to Uncertain Futures Jens Beckert and Richard Bronk

Human beings make decisions with an eye to the future. They plan for a rainy day, or act now in ways designed to make it easier to feed themselves and their offspring over the next year and beyond. This future orientation takes on a whole new character in modern capitalist economies: the future is no longer bound by tradition; and it can neither be safely assumed that the future will resemble the past, nor that only chance events will disturb the regular cycle of seasons and the predictable needs they imply. Nor is the future any longer generally envisaged in religious or Marxist terms as fitful progress towards a pre-ordained destiny. Instead, actors in capitalist systems face an open and indeterminate future. More precisely, they are able legitimately to imagine and plan for a whole array of possible futures, and choose between a bewilderingly large set of options without fully predictable outcomes. Such uncertain futures are the inevitable result of human creativity and the freedom to imagine new possibilities.1 The open and indeterminate nature of the future is partly a function of three notable features of the capitalist system—its reliance on competition, its tendency to encourage maximizing behaviour, and the partial liberation from inherited constraints it enables. The need for firms and individuals to compete with one another (and pay back interest on loans) if they are to survive in the capitalist system forces them to do more than use existing

1 This introductory chapter draws on the analysis and arguments developed by its authors in two previous books. These are Bronk, Richard (2009), The Romantic Economist: Imagination in Economics, Cambridge University Press, which examines the creative role of imagination in the economy from the standpoint of the history of ideas; and Beckert, Jens (2016), Imagined Futures: Fictional Expectations and Capitalist Dynamics, Harvard University Press, which explores the role played by imaginaries in structuring economic decisions and driving capitalist economies. The ideas analysed in these earlier books are updated in this introduction, not least to reflect new insights developed during the collaborative process of editing this volume.

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

products and resources as efficiently as possible: they must also introduce new products and methods.2 As Joseph Schumpeter (1943 [1976], 84) noted, the type of competition that counts most comes from ‘the new commodity, the new technology, the new source of supply, the new type of organisation’. The consequent ‘process of industrial mutation . . . incessantly revolutionizes the economic structure from within’, and this endogenous ‘process of Creative Destruction is the essential fact about capitalism’ (Schumpeter 1943 [1976], 83). The economic system is subject to constant change, novelty, and ‘unending disruption of the present’ (Beckert 2016, 23). When economists talk about actors ‘maximizing’ opportunities, they normally refer to the technical ‘allocative efficiency’ notion of optimizing among given factors. But, if maximization is considered in social and cultural terms, it is clear that—when present—it often involves the formation of new idealized images of perfection and the urge to enact them in reality. Such aspirations and imaginaries of a better (and more profitable) future are encouraged by market competition and consumer advertising, and are considerably more widespread than in the pre-capitalist era.3 The result is a relentless search for new opportunities, ‘improved’ products, and novel methods. Capitalist actors are constantly striving for more profitable ways of organizing the world and continually reinventing their preferences and even their identities; and this existential freedom to reinvent themselves and the parameters they face—and hence transcend the implicit determinism of rational optimization among given factors within inherited constraints—is one of the main gifts of creativity in markets (Bronk 2009, 219). As George Shackle (1972 [1992], 131) asked provocatively, ‘Is it not by their access to these creative aspects of their choice of conduct, that we can suppose men to have freedom’? It was Shackle who spelt out especially vigorously the crucial link between uncertainty about the future and the innovation and novelty at the heart of all dynamic capitalist economies. Shackle (1979, 52f) wrote of ‘our own original, ungoverned novelties of imagination . . . injecting, in some respect ex nihilo, the unforeknowable arrangement of elements’. Novelty and the imagining of new options inserts disjunctions into previously stable regularities, severs ‘predictable links between the past and the future’, and thereby undercuts the ability of economic agents to make probability forecasts on the basis of historical data (Bronk 2011a, 10). Douglass North (2005, 19) makes the implications of this crystal clear: ‘it is evident that we have been and are creating 2 For discussion of how the ‘expansive dynamism of capitalism’ (Sewell 2008) is driven by competition and the system of credit, see Beckert (2016, 4). 3 For discussion of the disposition to maximize and its ambiguous role in modern capitalism, see Bourdieu (2005), Shwartz (2004, 78–96), and Bronk (2009, 240f); and for analysis of how imaginaries drive the growth of consumer demand, see Campbell (1987) and Beckert (2016, 188–214).

2

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

An Introduction to Uncertain Futures

societies that are unique in comparison to anything in the past’; and it follows that to ‘know the future we would have to know today what we will know tomorrow’ (21). To quote Shackle (1972 [1992], 3) again, ‘What does not yet exist cannot now be known.’ The radical indeterminacy implied by innovation and novelty constitutes a major problem for economic actors: how are they to make decisions, and coordinate their actions with others, if they cannot know what future will follow? How can they form expectations of the future that may legitimately guide them? What is the role for rational analysis when they cannot deduce from past regularities of behaviour and known constraints what the optimal course of action would be? How far does political and economic power determine which expectations actors will develop and project? This volume gives theoretical answers to these questions and develops them through careful empirical and ethnographic studies of how economic actors in modern capitalist economies actually cope with uncertain futures. Suffice to say at the outset that Shackle (1979, 8) suggests a core part of the answer when he writes that ‘the void of time-to-come’ can be filled ‘only by works of the imagination’—that is, by what Beckert (2016) calls ‘imaginaries’ or ‘fictional expectations’. In other words, imagination is not only the root cause of uncertain futures; it is also one of our principal tools for coping with them (Bronk 2011b). Crucially, however, it is only when imaginaries are embodied in narratives and models that they become determinate enough to structure action at the social level and become a suitable object of empirical study. Notwithstanding the efforts of Shackle, as well as several prominent sociologists and political scientists, imagination remains largely ignored by economics and mainstream social science. Indeed, it receives relatively little attention even in the disciplines of philosophy, psychology, and neurophysiology.4 Imagination is usually the name given to a wide array of creative faculties of the mind: these include the ability to visualize counterfactuals, to place oneself in the shoes of another (the basis of sympathy), and to colour perception and analysis by playing with new metaphors. Central to the themes of this volume are two further related functions: first, the conscious or unconscious, willed or accidental, firing of new connections between established pathways in the brain, and the subsequent ability to build and grow the germ of a novel idea into an elaborate vision of the future; and, second, the open-minded receptiveness to new ideas that the poet John Keats (1817 [1998], 1019) called ‘negative capability’—that is, being willing to remain ‘in uncertainties, mysteries, doubts, without any irritable reaching after fact and reason’. This ‘family resemblance’ collection of faculties allows economic 4 The different facets of the human imagination, and their role in economic behaviour, are discussed in Bronk (2009, 196–224).

3

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

actors to see things in new ways; be creative by moving ‘beyond inherited thought-patterns and categories’; invent entirely novel ideas; spot emerging patterns; or choose between visualized but counterfactual options; and, taken together, they have the potential to be deeply ‘subversive of established order’ (Bronk 2009, 201). It is these faculties and related imaginative techniques that are the source of the ideational, preference, product, and process mutations in the economic system that enable societies to evolve and adapt to changing conditions. Equally crucially, these same faculties and instruments enable actors to fashion and refashion the social narratives and shared fictions that structure their expectations and guide their beliefs and actions in conditions of uncertainty. Far from capitalism enslaving us, as Max Weber (1930 [1992]) predicted, in an ‘iron cage’ of instrumental rationality, actors’ beliefs—their ‘fictional expectations’ and imaginaries—may constitute ‘a kind of secular enchantment of the world’ (Beckert 2016, 283). Like reason though, imagination can be a force for evil and folly as well as good. Many thinkers have been disparaging of imaginaries (and the emotions attaching to them) on the grounds that imagination can lead to fantasy—to extravagant and entirely fanciful visions of utopian futures, or dystopian visions of a hell that only the strong ruler can help us avoid. When imaginaries are devoid of any empirical basis in known causal mechanisms or stable constraints and are not anchored in an ethical framework, and when contextual facts and relevant science are rejected as mere ‘elite interpretations’, then those who peddle the imaginaries and associated narratives may be tricksters engaged in a crude struggle for profit or power. If imagination—the ability to conceive and visualize new futures—is to fulfil its potential as a driver of social and economic progress, it must be schooled by reason and empirical analysis and be under the strict governance of ethical deliberation. William Hazlitt (1805 [1998], 21) spoke of the need for ‘a reasoning imagination’; and many commentators since have argued that economic actors, scientists, and others must combine creative inspiration with rational analysis if they want successfully to apply logic and relevant lessons from the past to the elucidation of imagined futures. In business as in the rest of life, imagination should not be seen as antithetical to reason; rather, the two faculties must work hand in hand (Bronk 2009, 205f, 304). Imagination is what allows human beings to expand their horizons beyond what can be deduced from known facts; but imaginaries can be safely used as templates for economic decisions only when schooled by calculative reason. It is this often-fraught co-production of expectations (and economic futures) by imagination and rational analysis that is the subject of much of this book. The remainder of this introductory chapter provides a theoretical framework for considering the role of imaginaries and narrative fictions in 4

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

An Introduction to Uncertain Futures

the economy and how they interact with calculative devices to structure expectations and beliefs. It also analyses the nature of uncertainty in innovative markets and political economies, and examines ways in which economic actors use imaginaries and narratives to coordinate action, determine value, further their own interests, and establish sufficient conviction to act despite the uncertainty they face. The chapter ends by focusing on the challenge faced by all actors in remaining alert to the danger that the future may not accord with their fictional expectations. This opening chapter also serves to place the themes of the volume in the context of broader trends in economics and sociology. For much of the rationale for the book derives from the long shadow cast by the ‘two cultures’ divide (Snow 1959), which drained from standard economics and, to a lesser extent, other social sciences a balanced assessment of the relative roles played by imaginaries and reason. To read most economics texts, you could be forgiven for assuming that imagination plays little or no part in economic decision-making, and that—at system level—the economy is not a creative process at all. The fascination with fully rational expectations and modelling markets as an allocation process tending towards a predictable equilibrium speaks to a determination to keep economics as a form of ‘social physics’ (Mirowski 1989) that can have no place for the vagaries of imagined futures. But the world of entrepreneurs and policy-makers remains a more enchanted, poetic, and political space, and one that is ripe for study by scholars not afraid to bridge the two-cultures divide.

The Nature of Uncertainty in Innovative Markets and Political Economies In an important article, James Buchanan and Viktor Vanberg (1991, 176, 178 and passim) follow Shackle’s lead and claim that the market is a ‘creative process’, where the future is ‘yet to be created’ by choices and innovations as yet unmade. They argue that once ‘the creative-inventive-imaginative element in choice’ is acknowledged, it underscores ‘the tenuousness of the whole notion of equilibrium, defined as the exhaustion of gains from trade’ among given factors and goods (181–2). It also recasts the nature of the problem of knowledge in markets from one of overcoming ‘bounded rationality’ (Simon 1957), information asymmetries (Akerlof 1970), framing biases (Kahneman and Tversky 2000), or the dispersed and tacit nature of information (Hayek 1945 [1948]), to one of coping with uncertain futures—where all actors are in the dark about the world they and others will create. In other words, the challenge is no longer merely how to overcome the shortcomings of individuals as knowing agents and the well-researched structural or institutional 5

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

impediments to their attaining the required knowledge of information already ‘out there’. Instead, in a creative market, the future cannot be known simply because it does not yet exist.5 The relentless tide of product, process, and policy innovations that drive economic growth are partly designed by entrepreneurs and policy-makers; and the constant changes in the preferences and beliefs of investors and consumers that affect market prices are consciously shaped by economic actors ranging from central banks to the advertising departments of firms. But this does not imply that any of the relevant actors can know ex ante what the future they are attempting to shape will involve. This is because the first-order uncertainty implied by any particular novelty or innovation is compounded ‘by uncertainty about the second-order creative reactions of others’ (Bronk 2011, 9) and by the contingent nature of actors’ action-guiding interpretations of the unfolding future. In an ontological sense, there is a high degree of radical indeterminacy in the economy that ‘implies, as its epistemological counterpart, a lack of knowledge’, which David Dequech (2001, 920) calls ‘fundamental uncertainty’. Indeed, it is frequently impossible to know (as opposed to imagine) even the basic categories and entities of those aspects of future reality that have yet to be created (Lane and Maxfield 2005, 10f). Former Governor of the Bank of England Mervyn King (2017, 127) spells out the significance of this: Uncertainty . . . concerns events where it is not possible to define, or even imagine, all possible future outcomes, and to which probabilities cannot therefore be assigned. Such eventualities are uninsurable, and many unpredictable events take this form. A capitalist economy generates previously unimaginable ideas, new products and new technologies.

Uncertainty about the future is not, of course, only the product of the creativity and imaginaries of individual actors interacting with one another; it is also, as Andrew Haldane discusses in Chapter 7, a product of the emergent behaviour of complex economic systems characterized by threshold effects and complex feedback loops. Like Buchanan and Vanberg (1991), Haldane notes that economies often show little or no tendency to equilibrium, and that this renders them unsuitable for predictive modelling according to techniques based on the metaphor of ‘celestial mechanics’ (Shackle 1972 [1992], 4). The importance of emergence and self-reinforcing non-linear dynamics, and the endogenous disequilibrium they imply, leads Haldane to argue in favour of agent-based models that model the complex interaction of heterogeneous 5 For the important distinction between ‘more or less tractable epistemological problems’ related to information asymmetries or the cognitive shortcomings of knowing agents and the deeper problem of ontological indeterminacy resulting from novelties and innovation, see Bronk and Jacoby (2016, 8–11).

6

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

An Introduction to Uncertain Futures

actors; and it has led, more broadly, to considerable interest in complexity economics (Arthur 2015)—where economies are modelled as self-organizing systems, in which small changes in initial conditions may snowball into radically divergent (and path dependent) outcomes. Even these models, though, have a tendency to view the economy in isolation from broader elements of society. As a result, they do not capture the uncertainty that stems from the unpredictable interactions between the economy and other subsystems of society—especially politics and the normative values held in society—which often stand in contradiction to the system demands of the economy. In particular, Karl Polanyi (1944 [1957]) highlighted that the attempt to install a market society inevitably runs up eventually against a countermovement of social protest, as actors find it increasingly hard to bear the instabilities caused by the full commodification of labour, money, and nature. How these counter-movements play out cannot be predicted, but they are associated with profound social, political, and economic crises. Politics can be a major source of indeterminacy and radical uncertainty in the economy, but its impact should not be seen purely in terms of providing external shocks to the smooth functioning of markets. Rather, there is something inherently political about economies built on imaginaries. Since expectations are not anchored in some pre-existing future reality, but rather have an important role in creating the future, they are the legitimate object of political challenge, debate, and choice. The visions and guiding narratives that a society chooses to prioritize usually have important distributional consequences, making them subject to competing interests (Jasanoff and Kim 2009). They are very often also manifestations of the competing values and identities that provide the contours of political debate. Economists may assume that they can model on the basis of revealed preferences and known indifference curves, but the preferences of voters and market participants are frequently the unstable product of a continual process of identitydefining trade-offs between incommensurable and conflicting values. Much of the dynamism and unpredictability of market economies is a function of the frequent social and political ‘redefinition of desired trade-offs between incommensurable and conflicting values’ (Bronk 2009, 195), where there is no single rational and optimal solution. Economists have long been aware of the problem of uncertain futures, even if—with the exception of the authors mentioned here—they have rarely linked it explicitly to the prevalence of innovation and novelty in markets. Nearly a hundred years ago, Frank Knight (1921, 235f) made his famous distinction between measurable ‘risk’ (which allows probabilities to be calculated) and immeasurable or radical ‘uncertainty’ (where probabilities cannot be calculated because each situation is unique). Crucially, Knight viewed 7

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

uncertainty as central to entrepreneurial activity and, indeed, as the main reason that firms can make profits (311), given that any source of profit that is amenable to accurate probability forecasts would quickly be competed away in a competitive system. John Maynard Keynes (1936, 149), too, saw uncertainty as a central feature of economic life: The outstanding fact is the extreme precariousness of the basis of knowledge on which our estimates of prospective yield have to be made. Our knowledge of the factors which will govern the yield of an investment some years hence is usually very slight and often negligible.

The result is that market valuation ‘cannot be uniquely correct, since our existing knowledge does not provide a sufficient basis for a calculated mathematical expectation’ (Keynes 1936, 152). These central insights, however, became increasingly ignored by standard economics, especially from the 1980s onwards (Hodgson 2011); and the reason is clear. As Robert Lucas disarmingly admitted, the theory of rational expectations is not applicable ‘in situations in which one cannot guess which, if any, observable frequencies are relevant: situations which Knight called “uncertainty” ’ (Lucas 1981, 224). Standard economics and finance theory sit atop a set of microfoundations that either simply ignore the problem of radical indeterminacy in innovative markets and political economies or (in Knight’s terms) conflate uncertainty with risk. Rational expectations theory, for example, assumes that actors’ expectations are ‘informed predictions of future events’ and ‘essentially the same as the predictions of the relevant economic theory’ (Muth 1961, 316), and that any errors in these predictions are essentially random. These assumptions are less outlandish than it may at first appear if you also assume, as most economists implicitly do, that the future is ‘ergodic’ or ‘merely the statistical shadow of the past’ (Davidson 2010, 17)—its significant parameters ‘pre-determined’ and ‘immutable’ (Davidson 1996, 479f). For then, expectations can legitimately be based on calculated probabilities—measures of risk based on data on the past—and competitive pressures can be expected to eliminate systematic forecasting errors, so that all relevant information is taken into account. Furthermore, if you make this assumption that the future is ‘implied in the present’ (Buchanan and Vanberg 1991, 170), and that expectations reflect all available information, then it is not far-fetched to assume the efficient market hypothesis, which sees market prices as correctly reflecting the fundamental value of an asset (including its predictable future earnings), subject only to random changes related to chance events. The premise of an ergodic world and the efficient market hypothesis tend to encourage the view that political interference in markets is either useless or outright damaging. If rational actors cannot be ‘fooled’, as rational expectations theory argues, it 8

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

An Introduction to Uncertain Futures

follows that any public policy attempting to ‘manipulate the economy by systematically making the public have false expectations’ (Sargent 2008, 433) is doomed to failure. Economists are fully aware of some of the problems with rational expectations theory and the efficient market hypothesis, not least because the theories are continually called into question by evidence of long periods during which most economic actors hold expectations that later prove to be systematically delusional, while market prices are massively volatile and subject to trends that appear anything but random. The main response within economics, though, has been to focus on information problems (especially asymmetries of information) and biases in the human processing of knowledge (as in behavioural economics) that are predictable and can in theory be corrected. This approach promises to allow bolt-on amendments to rational expectations models to improve their predictive capacity. Although such theoretical responses are pertinent in elucidating certain types of market failure, they fail to address the central problem in innovative economies: namely, that the future is ‘non-ergodic’ and ‘transmutable’; and that, as a result, measurable data on antecedent conditions do not provide a basis for objective probability functions on which rational actors’ expectations can converge (Davidson 1996, 479f; North 2005, 19–22). Economic actors cannot conform to the assumptions of rational expectations theory whenever the future is still to be determined by creative choices they and others will make—when, that is, novel products, processes, imaginaries, and policies are disrupting systematic regularities of behaviour. Nor can they exhibit fully rational expectations tending to a unique equilibrium point— even when entirely free from cognitive bias and information asymmetries—if they are operating in complex economic systems characterized by tipping points, threshold effects, and feedback loops. Instead, all actors operating in these conditions of uncertainty may be subject to ‘symmetric ignorance’ of the future (Skidelsky 2009, 45). This brings us back to the central questions addressed in this volume: if the expectations of those operating in innovative markets cannot be based on the rational calculation of probabilities based on past data, how do they form the expectations and beliefs on which their consequential decisions depend? And if we all live in a world of radical uncertainty and hence are unable to gravitate to a uniquely rational set of expectations, how do we coordinate our actions with one another? Before exploring these questions further, it is important to qualify the basic assumption in this volume that, in innovative capitalist economies, we cannot know what the future holds. In practice, of course, the future is not the totally unknowable ‘void’ that Shackle spoke of, for two important reasons: first, the future is almost always a messy amalgam of persistent regularities of 9

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

behaviour and stable institutional or physical constraints (on the one hand) and novelty and innovation (on the other); and second, we can often discern pointers in the present to patterns that may later unfold, and make pattern predictions about the future in complex and creative systems, based on what we do know about relatively stable constraints and incentives. In other words, actors are rarely faced with a binary situation in which the aspects of the economy that concern them are either ergodic—and hence fully predictable (as standard economics assumes)—or completely non-ergodic—leaving them with no clue about the future (Bronk 2011b). Actors intent on long-term success need to calculate as best they can what is calculable, while making imaginative use of models and narratives to diagnose and interpret newly emerging trends in the uncertain futures they face. Furthermore, they must make judgements—depending on the nature of the problem addressed—about the relative importance of rational calculation and imaginative play with different diagnostic tools. This, as we shall see, has a huge bearing on the role of calculative devices in the formation of fictional expectations.

Fictional Expectations and the Contingency of Market Value Beckert (2016) elaborates a theory of ‘fictional expectations’ to serve as a replacement for rational expectations theory as the basis of suitable microfoundations for economics when seeking to explain and model decisionmaking in conditions of uncertainty. When the future is not already ‘given’, and cannot be assumed to exist as a shadow of the past, economic actors resort to expectations that share many of the attributes of literary fictions. First, the expectations they use delineate visions (of the future) that go beyond observable truths. Second, actors may, prior to decisions, ‘play’ with different possible courses of action by experimenting with different counterfactual images of reality (Beckert 2016, 57; Ricoeur 1991, 177f). Third, expectations often adopt a narrative form that sets out credible causal relationships between the known present and imagined futures, while assigning meaningful roles to the different protagonists. And finally, expectations rely for much of their impact on the rhetorical power of the language and metaphors used (McCloskey 1998). In several respects, of course, fictional expectations in the economy differ from most literary fictions: for one thing, disbelief in them is normally suspended only if the expectations have practical credibility as potentially feasible in the real world; and for another, expectations can frequently inspire action designed to bring about the envisaged future (if the vision is desirable) or foil it (if the vision is distasteful). Fictional expectations have real-world consequences. 10

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

An Introduction to Uncertain Futures

Understanding expectations formed in conditions of uncertainty as ‘fictional’, and as relying on imagination and narrative structure to build up visions of the future from the tenuous pointers available, underscores their essentially contingent nature. In innovative economies there is no fixed anchor in underlying reality for the expectations of even the most rational actors, since the ‘future present’ (Luhmann 1976) does not yet exist and cannot now be known. The result is that the expectations that guide decisions in the economy may be as indeterminate (and unpredictable) as the future to which they relate. This has crucial implications for the capacity of market prices to reflect the true (duly discounted) future value of assets: rather than prices reflecting relevant decentralized knowledge (as Hayek assumed), prices reflect the contingent way we happen at any moment to imagine the future will be (Bronk 2013a, 95), as well as the contingent interpretations we place on incomplete evidence. As Shackle (1972 [1992], 8) epigrammatically puts it, ‘Valuation is expectation and expectation is imagination’. Since market prices and valuations reflect ‘contingent expectations’ (Orléan 2014, 198), it follows that the stability of prices depends primarily on the stability of expectations (and of the imaginaries and narratives that structure them). Moreover, success in forecasting price movements depends above all on forecasting shifts in what Shackle (1972 [1992], 9) calls the ‘expectational kaleidoscope’. This makes traders, as Keynes (1936, 155) famously observed, highly attentive to market psychology, and to the stories or narratives influencing opinion (Beckert 2016, 147); and this in turn provides a rationale for the pleas for a new ‘narrative economics’ (Shiller 2017) discussed later in this chapter. The contingent nature of the imaginaries and stories influencing actors’ beliefs does not imply that the evolution of market expectations and resulting market valuations is random and entirely unpredictable. This holds for three reasons. First, expectations are anchored in inherited social structures, cultural frames, conventions, rules, and institutional settings that reduce the potentially limitless dispersion of expectations.6 As North (2005, 50) observes: ‘The intimate interrelationship of beliefs and institutions, while evident in the formal rules of a society, is most clearly articulated in the informal institutions—norms, conventions, and internally held codes of conduct.’ Economic actors do, of course, from time to time refashion the codes of conduct under which they operate, but for much of the time existing rules and institutions channel the ‘sources of contingency’ (Offe 1998, 682).7 For discussion of the social foundation of fictional expectations, see Beckert (2016, 87–92). The relationship between narratives, imaginaries, and institutions is a complex one. In a sense, widely shared narratives, or ‘public images’ (Boulding 1961) can be seen as the most dynamic, fluid, and contingent of the ‘informal’ institutions that act as ‘scaffolds’ or structures ‘to define the “way the game is played” ’ (North 2005, 48). But it is perhaps more apt to see competing narratives as the 6 7

11

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

The way actors think, and the expectations that guide their decisions, are to some extent path dependent and the product of their ‘habitus’ (Bourdieu 2000). Second, even novel interpretations and entirely contingent expectations are formed in social groups. If imaginaries and interpretations were entirely personal, then economic decisions and the future they shape would indeed be radically indeterminate. In fact, expectations are to a considerable extent the product of social narratives, ‘public images’ (Boulding 1961), and powerful opinion-formers. Moreover, they are shaped by widely shared calculative devices that mould the way actors perceive the future.8 Much of this volume is devoted to documenting this social construction of expectations—whether by policy-makers in central banks, credit rating agencies, business associations, key corporate players, or regulatory standards authorities—and to analysing the power dynamics among these influential expectation-formers. Finally, expectations—particularly those of long-term investors or policymakers—are normally sensitive to established knowledge about persistent physical and institutional constraints, stable causal mechanisms, behavioural regularities, and the relative scarcity of financial resources. Economic actors who wish to succeed in competitive markets are not free to entertain any fantasy they like, since rational analysis can reveal some imagined futures to be almost certainly infeasible in the light of known constraints, power structures, and market trends. In other words, the scope of credible imagined futures is not limitless. There remains a central role for rational analysis in channelling fictional expectations in credibly feasible directions. It is important to remember, though, that—in the short-run at least—the market or political success of a particular narrative or imaginary may depend more on its emotional appeal, the credibility and power of its author, and the rhetorical techniques it employs than on its ability to capture available information about persistent constraints and emerging patterns. This means that much of the rational analysis in which market practitioners engage is either essentially reflexive9—that is, trying to gauge the guiding beliefs or modelling assumptions of other market operators—or designed to assess the emotional salience of competing narratives.

constantly re-imagined, manipulated, and often highly contagious mutations in the DNA of institutional structures. Institutions—such as the US Constitution—may embody, stabilize, and reproduce successful narratives; but periods of dynamic change are those where newly imagined narratives spread ‘like viruses’ (Akerlof and Shiller 2009, 56). 8 In the language of complexity economics (see Arthur 2015), social narratives and widely shared calculative devices can be seen as attractors in a self-organizing system of economic expectations. There is no equilibrium point, but moderately stable patterns of expectations and behaviour can persist for some time until new imaginaries or mutations in social narratives emerge. 9 See, for example, Beunza and Stark (2012, 383–413), who discuss the ‘reflexive modelling’ used by traders to infer the guiding beliefs of other traders from the prices they pay, for use as ‘inputs to their own decision-making’.

12

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

An Introduction to Uncertain Futures

Rational Calculation Devices as Instruments of the Imagination It may at first glance seem odd to focus so much on the role of imagination in the construction of expectations, and on the fictional nature of those expectations, when almost all the practical examples in the chapters that follow show those making decisions in conditions of uncertainty relying on a marked and explicit use of rational calculation devices. The twentiethcentury business plans in innovative industries discussed by Martin Giraudeau in Chapter 12, for example, do not resemble in any obvious way the fanciful, even utopian, prospectuses of the eighteenth-century ‘projectors’ that Giraudeau (2010) and Valerie Hamilton and Martin Parker (2016) have studied elsewhere. Instead, the business plans are apparently sober assessments of feasibility, backed up by careful sensitivity analysis. Likewise, the discounted cash-flow models analysed by Liliana Doganova in Chapter 13 are sophisticated mathematical devices that aim to assess the net present value of long-term investments. And finally, the macroeconomic forecasts examined by Werner Reichmann and Olivier Pilmis in Chapters 5 and 6, respectively, make use of economic models and statistical data, and profess (to some extent at least) to calculate the future. In what sense, then, are such devices ‘instruments of the imagination’ (Beckert 2016) that help in the construction of ‘fictional expectations’? Business plans, discounted cash-flow models, and even country credit ratings are attempts to organize and subject to rigorous analysis visions of the future that have their origins in imagined futures, while at the same time helping to create such futures. It is various outcomes or investment goals that spring ultimately from the visions and counterfactual scenarios imagined by entrepreneurs and investors that form a key input to these calculative devices: in a business plan, the imagined capturing of a new market is carefully assessed for feasibility in the light of known financial constraints and causal mechanisms under various imagined scenarios; in discounted cash-flow analysis, an imagined return on investment is ‘discounted’ in a careful calculation designed to assess whether this return would be worth the expected opportunity costs, given a discount rate based on certain fictional assumptions; and, in the preparation of country credit ratings (discussed by Natalia Besedovsky in Chapter 11), potential (imagined) threats to a country’s stability are assessed in relation to relevant and well-understood causal relationships and known resources. In all these cases, though, imaginaries are more than inputs to models; they are also partially formed by the very models used to explore them. The systematic structure of models helps to flesh out and construct the imaginaries central to fictional expectations. The use of calculating devices in this way is a perfect example of Hazlitt’s ‘reasoning imagination’ in action: it involves the careful stress-testing of 13

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

imagined futures by rational analysis in the light of such knowledge as we have about emerging patterns and stable (knowable) constraints (as a prop to careful judgements about uncertain futures); and it involves the exploration and visualization of different possibilities with the help of systematic tools of analysis. It would, after all, make no sense to argue that because it is impossible to predict the future with precision, economic actors should neither plan for possible eventualities nor examine analytically the threats, opportunities, profits, and losses that, ceteris paribus, could occur on certain specific (imagined) assumptions. At the same time, it is essential to recognize that the output of calculative devices using imagined assumptions and scenarios about the (yet to be created) future is still to a very significant extent ‘fictional’— in the sense of going beyond what can be known as established fact. Calculative devices have two closely related epistemic functions for those seeking to make decisions in conditions of uncertainty. First, while the future is uncertain and will contain genuine novelties, it will also remain heavily influenced by a number of stable and well-understood causal mechanisms or tendencies. As a result, there may be significant aspects of the unknown future that are amenable to the calculation of objective probability functions derived from statistics on the past. Indeed, part of the judgement that a wise decisionmaker must exhibit is to disentangle the incalculable potential impacts of novelty or complex interdependencies from the calculable risks associated with aspects of the economic regime likely to remain immune to novelty.10 Second, even the most dynamic and unstable elements of economies exhibit patterns of behaviour that either have been seen before or (if they are new) can be understood by analogy and with help from (prior) computeraided simulations of complex interaction effects. In other words, in conditions of uncertainty, a key function of models and other calculative devices is to help economic actors diagnose newly emerging patterns, as well as persistent regularities.11 It is often calculative models that enable economic agents to be imaginatively receptive to pointers to new trends. Calculative devices and models carry out this diagnostic function by improving actors’ understanding

10 There is, of course, no way for economic actors to know for certain ex ante whether or not a particular aspect of the economy will remain ergodic and suitable for probabilistic calculation; but it is possible to form qualitative judgements about how likely the future is to be characterized by a significant degree of radical indeterminacy, by considering the prevalence of innovation and complex feedback loops. Some situations are evidently more stable than others, while highly innovative market sectors or periods of rapid political and regulatory change are signals that it is problematic to consider calculative devices as instruments of knowledge rather than generators of more or less useful fictions. Even when dealing with apparently ergodic aspects of the economy, of course, judgement is still required to determine which elements of the past are relevant to probabilistic prediction in a particular case. As Riccardo Rebonato (2007, 146) puts it, ‘the selection of what constitutes the relevant past’ for constructing data sets ‘is not objective’. 11 John Sutton (2000, 16) discusses how economic theory can similarly act as a ‘diagnostic tool’ to tease out such systematic tendencies as may exist in messy reality.

14

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

An Introduction to Uncertain Futures

of existing causal relationships (and known probabilities, where applicable), and by simulating the likely impact of changing certain key variables in complex systems. Consider, for example, climate change. Most scientists would readily admit that they cannot predict with any precision how much the temperature will rise over the next year or the coming century: the complexity of the interactions and the number of variables is such that even if scientists knew how much man-made carbon will be produced, they could not predict exactly what will happen. Moreover, they cannot know what creative humans will come up with to reduce carbon intensity or mitigate the effects of carbon on the climate. But such uncertainties emphatically do not reduce the usefulness of climate-change modelling, which has helped us enormously to understand the mechanisms at play and the sensitivity of outcomes to complex threshold effects. Nor, in most circles, have the uncertainties involved seriously called into question the power (or legitimacy) of such models to shape the contingent expectations of economic and political actors. Climate models are scientific ‘props’12 that have triggered a set of fictional expectations likely to shape economic and political decisions for years to come. In Chapter 7, Haldane explains a new approach to economic modelling that starts from the premise that economic systems are also inherently unpredictable thanks to the complex interaction of highly diverse groups of actors. He shows how the Bank of England is increasingly using agent-based models (ABMs) to simulate the potential impact of different policy options on market dynamics. ABMs assume that economic agents ‘are heterogeneous and interactive’ and that, at system level, ‘multiple, evolutionary equilibria’ are the rule (p. 150). Crucially, system dynamics are ‘endogenously’ driven, thanks to strong feedback effects and emergent behaviours (p. 152). These are models where there is no representative agent, the future is highly uncertain, and economic agents use specified rules of thumb (or heuristic narratives) to guide their behaviour. Nevertheless, Haldane shows that ABMs can give policymakers important pointers to likely patterns in key markets under various policy conditions—especially their sensitivity to changes in certain regulatory conditions and the likely frequency of large economic dislocations or market oscillations. The main practical value of conventional macroeconomic forecasting methodologies may also lie, ironically, in their underappreciated role as diagnostic tools for teasing out emerging patterns rather than as generators of point forecasts. Economic forecasting is in general notoriously unsuccessful at predicting

For discussion of how fictional expectations are structured with the help of certain ‘props’ in the form of models or business plans that trigger contingent imaginaries of the future, see Beckert (2016, 68). 12

15

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

future outcomes with any precision at all. In the case of macroeconomic forecasts, outcomes are often well outside the ranges considered likely by almost all the prestigious forecasters (Beckert 2016, 222f). Haldane explains this as due to the ‘damped harmonic motion’ assumed by the Dynamic Stochastic General Equilibrium (DSGE) models normally used by those modelling the macroeconomy (p. 147–9). One study (Gigerenzer 2013) of euro– dollar exchange rate forecasts showed that tossing a coin would have been as effective as the experts in forecasting even general trends. And yet vast resources are poured into such forecasting exercises. Werner Reichmann explains one rationale for this in Chapter 5: the discursive and interactional process of forecasting he terms ‘foretalking’—in which forecasters engage with a broad epistemic network—‘brings to light possible futures that might not otherwise have been imagined’ (p. 108). Even if these novel imaginaries and stories prove to be wide of the mark as a picture of the future, they may still help the users of the forecasts to understand emerging trends and relevant causal mechanisms better. Olivier Pilmis makes a similar argument in Chapter 6, where he analyses ‘the persistence of forecasting despite recurring “errors” ’ (p. 125). Pilmis provocatively compares the forecasting profession’s ‘will to believe’ in their own abilities (despite evidence to the contrary) to that displayed by magicians, as portrayed in sociological literature. His main argument is that forecasters use the very indeterminacy of the economy (and the presence of unexpected shocks) as proof that errors are due to factors outside their control; and, consequently, what matters to them is whether or not they have followed professional forecasting practices (or rituals). In addition, Pilmis argues that forecasters will often seek to justify their value in the face of apparent errors by pointing to their correct identification of an economic narrative, causal mechanism, or set of scenarios that later proves helpful in understanding the future as it unfolds.13 Calculative models and forecasting methodologies should then be understood as improving the epistemic quality of fictional expectations and decisionmaking in two ways—first, by using formal procedures to stress test pre-existing imaginaries for plausibility in the light of known constraints and calculable probabilities (where systematic regularities and known causal mechanisms make them applicable); and, second, by helping economic agents imagine new possible futures (or scenarios) and identify emerging trends they might 13 The relevance of this argument to the UK Brexit debate is topical. It has been frequently observed that most of the economic forecasts made in early 2016 on the impact of Brexit on the UK economy were misleading at least in respect of the short-term time horizon in which they were couched. But many forecasters would point out that some of the causal mechanisms and scenarios highlighted in the forecasts (such as the impact of sterling devaluation on UK inflation and real wages) have nevertheless proved highly pertinent.

16

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

An Introduction to Uncertain Futures

otherwise have missed. In both cases, the use of a variety of calculative devices and forecasting procedures aids rather than hinders entrepreneurial judgement by allowing for imaginative play with diverse formal interpretations. Whereas premature encasement in a single system of thought may prevent receptiveness to evidence of novel trends, new insights are frequently generated by the flexible use of different analytical tools, the organized juxtaposition of conflicting evidence, and the challenge to received wisdom derived from using a new calculative device. Disciplined analysis is the whetstone of the imagination, and not its antithesis, so long as it is combined with an open mind. Two chapters in this volume shed further light on how closely intertwined the analytical and the exploratory (or diagnostic) functions of calculative devices can be. In Chapter 12, Martin Giraudeau examines the project appraisal procedures in place at American Research and Development Corporation (ARD) in its heyday when managed by the legendary venture capitalist Georges F. Doriot. Giraudeau shows that, despite Doriot’s strong stated commitment to shunning systematic approaches to mapping the future, he in practice required his employees to use a number of formal administrative procedures, decision schedules, and analytical techniques in order to make their judgements. While stressing the need to avoid premature decisions and remain open-minded about the uncertain future, Doriot and colleagues strove constantly to increase their knowledge of relevant factors through formal procedures of investigation. Giraudeau argues that Doriot was attracted by the idea that the good entrepreneur should, like a doctor, develop ‘therapeutic wisdom’ based on experience and pertinent data (p. 267). While Doriot saw a ‘sense of the future’ and ‘feelings’ of conviction as essential to decisionmaking, the required ‘sense’ and feelings nonetheless had to be based on a set of analytical techniques and conceptual grids that could enable objective processing of available information and ‘a shared interpretation of emerging reality’ (p. 271). Giraudeau concludes that, for Doriot and others at ARD, new insights ‘were triggered by the available knowledge’. ‘Imaginative foresight was under strict knowledge oversight’ (p. 274–5). In Chapter 13, Liliana Doganova explores the use of discounted cash-flow analysis (DCF) in its early days as a tool of forestry valuation and management and more recently in the pharmaceutical industry. In both cases, the procedures she discusses organize the assessment of imagined futures with a combination of rational calculation and scenario exploration. Indeed, one of the main functions of DCF, in Doganova’s account, is in ‘making multiple options visible and debatable’ (p. 293). In other words, DCF models serve not only to value different options in terms of their net present value on certain discounting assumptions but also to help agents visualize and explore possible futures they might not otherwise have considered. In this way, DCF may act as a 17

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

source of novel templates for action and open up a multiplicity of possibilities— a function also performed by scenario analyses. More generally, Doganova argues that the uncertain future (as it appears to economic actors) ‘is consubstantial with the instrument of valuation’ (p. 280). In other words, the ‘form it takes is that of the lens through which one looks at it’ (p. 286).

Calculative Devices as Social Justifications for Action and Constitutive of Markets Perhaps the most important role played by calculative devices is the social one of justifying and legitimating action despite uncertainty about the future. The calculation methods used help to justify a decision by providing reasons for action; and they also represent evidence of due diligence consideration of the dangers involved. This crucial legitimizing role of calculative models can be entirely benign even in conditions of uncertainty so long as their epistemic status as generators of ‘informed imaginaries’ is remembered. But these same devices can pose a threat to the stability of markets or societies if their epistemic status is misunderstood, or if they are used by vested interests to hoodwink other economic actors about the degree of uncertainty involved. No amount of calculation and careful analysis of known constraints and causal mechanisms can (in Knight’s terms) turn genuine ‘uncertainty’ into objective probability functions or knowable ‘risk’. But this does not stop some economic actors pretending to themselves or others that such alchemy is possible. In particular, the apparent mathematical precision of devices calculating probabilities or net present values is often harnessed to provide a degree of social or market legitimacy for bold action that is simply unwarranted given the residual uncertainties involved: it enables the actors concerned to pretend (or act as if ) they can know what the future holds sufficiently well to act in a way that would otherwise seem foolhardy. Moreover, calculative devices are used as props in decision-making in part because they meet the requirements of what Michael Power (2007, 197) calls the ‘logic of auditability’, and its associated ‘cultural ideals of precision, proof, and calculability’. Linked with this institutional incentive for unwarranted precision is the broader danger of the sort of ‘scientism’ that F. A. Hayek (1952 [2010], 80) warned about, where economic actors engage in a ‘slavish imitation of the method and language’ of mechanical physical sciences and apply them in ways that are not appropriate to explain the quite different complex socio-economic subject matter of markets. The result is a dangerous illusion of scientific control over the unknowable future. 18

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

An Introduction to Uncertain Futures

The credit rating practices in derivative markets before the post-2007 financial crisis, where agencies presumed they were dealing with the sort of ‘risk’ that is calculable, controllable, and suitable for exploitation in risk markets, enabled the creation of financial products that promoted in their users exactly such an illusion of control. The risk models and credit ratings in the highly innovative (and therefore ‘non-ergodic’) CDO markets prior to 2008 are now known to have fatally confused measurable ‘risk’ with radical ‘uncertainty’ (or indeterminacy). They assumed incorrectly that it was possible to calculate the probability of future default on the basis of data on the past (Bronk 2011, 7; MacKenzie 2011); and, since the market actors using the risk models behaved as if such models could accurately forecast objective probabilities, the results were devastating.14 In Chapter 11, Natalia Besedovsky makes a key distinction between ordinal rankings and cardinal measures of credit risk. Even in areas where objective probabilities are not possible (because—as with country risk—the case is unique), it is still potentially helpful for decision-makers to provide a numerical ranking of the uncertainties faced. This is what credit rating agencies offer with their traditional ‘holistic’ and ‘diagnostic’ judgements (p. 244) of the relative safety of different sovereign bonds, based on their opinion of the current state of each economy. Such ordinal rankings are quite different from the practice of providing statistical probabilities of default (and loss given default) based on past default data and the assumed diversification benefits of pooled investment products. It is the latter sort of ‘precise cardinal measures of credit risk’ (p. 246) that provides a key input to Value at Risk and other risk models used by managers and regulators in the financial sector, and to the pricing of complex derivatives; and, as Besedovsky points out, it is these techniques that allow risk to become something to be ‘managed and exploited as an opportunity for profit’ rather than, preferably, avoided (p. 249). Besedovsky argues that such ‘technical’ risk-measurement practices—that deal with radical uncertainty and indeterminacy either by simply ignoring it or by assigning it a cardinal measure that allows it notionally to be taken into account—constitute the specific ‘epistemic culture’ (Knorr Cetina 2007) that ‘enables the creation of structured finance securities in the first place’ (p. 238). In other words, the enormous structured finance markets in derivatives would not exist in their current form without these cardinal risk-measurement 14 Adair Turner (2016, 102) has articulated the depth of the intellectual errors underlying the whole edifice of Value at Risk models prior to the crisis. The models normally assumed— incorrectly—that relevant probabilities followed a ‘normal distribution’, while the historical data used were based on too short a time period even ‘to capture the full historical experience of price volatility . . . But more fundamentally still, they were based on the flawed assumption that the probability of future developments in financial markets can be inferred from observation of the past.’ Such Value at Risk models ‘thus fail to recognize that the future is governed not by quantifiable probabilistic risk but by inherent uncertainty.’

19

OUP CORRECTED PROOF – FINAL, 9/6/2018, SPi

Uncertain Futures

practices and corresponding assumptions. The practices, assumptions, and models used constitute the grammar and internal logic of these markets, and shape the analysis and behaviour of its participants. To use the terminology of Donald MacKenzie (2006), the cardinal risk-measurement practices should be seen not so much as a ‘camera’ recording current pointers to the future as an ‘engine’ that ‘performs’ the future by creating the possibility of markets. In reality, of course, the two functions (camera and engine) are in tension with one another; and, since cardinal probability measures of risk are useful in illuminating the future only in ‘ergodic’ conditions where the future is genuinely a statistical shadow of the past, it is not surprising that their use in financial markets characterized by widespread innovation was shown in the financial crisis to have been based on an illusion of knowledge. It was an illusion that ensured that models using these cardinal measures of risk did not perform as risk-management assumptions and finance theory implied. The misuse of cardinal risk assessments in highly innovative derivative markets was an example of a more general problem in markets trading in uncertain futures—what might be called the ‘reification’ of calculated fictions about the future as apparently objective features of that future. Since the calculative devices are embedded in the products traded across huge futureoriented markets, they become accepted as genuinely reflecting the future. The same phenomenon is also illustrated in Chapter 10 in Elena Esposito’s analysis of markets trading in derivative products based on volatility calculus. As she explains, financial models calculate—and can only calculate—‘the present future’ (that is, ‘the future as we can expect it today, on the basis of currently available information and statistical models’); but this analytical output often becomes confused with the ‘future present’—that is, unknowable future reality (p. 223). Esposito argues that this confusion was a major cause of market instability in the financial crisis, because the derivative models—built on present calculations of prevailing market opinion about the future and its volatility—themselves had a strong influence on the future present—the (ex ante unknowable) future that then eventuated. Esposito adds: ‘Like all fictions, financial models about the future are extremely controlled constructions’; but since ‘they are not accurate representations of a future reality’, they can end up ‘reproducing’ the very uncertainty they claim to control (p. 228 and p. 233).

Narrative Economics: The Role of Stories, Conversation, and Animal Spirits in the Construction of Meaningful Futures Economic actors who are aware of quite how open the future is in an innovative and complex capitalist system are prone to bouts of paralysing anxiety and disorientation. Gone is the reassuring notion of being able to rely on the 20

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

An Introduction to Uncertain Futures

lessons of history and calculate an optimal course.15 Keynes (1936, 162) argued that investment in uncertain futures will inevitably falter if it depends on ‘nothing but a mathematical expectation’: it ‘will only be adequate’, he added, ‘when reasonable calculation is supplemented and supported by animal spirits’—that is, by the sentiments of confidence and hope. The recessions, and even depressions, into which modern economies not infrequently fall are testament to the negative impact of fear and disorientation in the face of uncertain futures. For capitalism to function normally, economic actors must have fictional expectations that give them the confidence to act and provide ‘interpretive frames to orient decision-making despite the incalculability of outcomes’ (Beckert 2016, 9). Central to this aspect of expectations is the use of ‘narrative structure’, which allows actors to ‘keep ontological uncertainty at bay’ (Lane and Maxfield 2005, 4). Narratives are ‘one of the important devices humans use to give meaning’ to their activities, and ‘create the commitment to act’ (Tuckett 2011, xvii). By integrating existing information and known causal mechanisms, they provide motives for action, and ‘play a big part in decisions taken under conditions of radical uncertainty’ (King 2017, 136). Economic narratives can sometimes be found as stand-alone vectors of general orientation that structure the action-guiding beliefs of those who adopt them—as seen, for example, in stories of new eras, promised fortunes, or dystopias that must be avoided. More often, the stories are embedded in calculative devices whose assumptions they govern, or the stories embed within their own structure an illusion of calculability and an understanding of causal mechanisms derived from the models they draw on. In all these forms, narratives perform a number of related functions. First, they assign roles to actors and objects, and develop a ‘plot’—a storyline of how an imagined future may unfold—that provides a guiding image of anticipated innovations and uncertain futures. This function is particularly evident and necessary in the case of firms operating in radically innovative sectors, where promissory stories provide a script for the future that positions ‘the relevant actors, explicitly or implicitly, exactly as characters in a story are positioned’ (van Lente and Rip 1998, 218). Stories and business plans are a type of ‘forceful fiction’ that reduces the indeterminacy and contingency of the future ‘by providing a blueprint that can be used in action’ (217). As Lane and Maxfield (2005, 11f) argue in their study of innovative start-ups in Silicon Valley, guiding narratives supplement, or even replace, analysis of future

15 As King (2017, 132) puts it: ‘in a world of radical uncertainty even smart people do not find it easy to know what it means to behave in a smart manner.’ Since they cannot know what the future might hold and are therefore unable to optimize, it is perfectly rational for them to rely on ‘coping strategies’—including narratives (135).

21

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

consequences; and they orient actors by providing a logic of action and populating the future with counterfactual artefacts worth investing in. An analogous use of narratives in conditions of macroeconomic uncertainty is discussed in Chapter 8 by Douglas Holmes, where he analyses the ways in which central banks use narratives and forecasts to ‘endow the future with discernible features that we—the public—can reflect and act upon’ (p. 173). By communicating a picture of the future evolution of the economy and the development of inflation, the central bank can, in effect, enlist the public’s help in reaching desired macroeconomic targets with rhetoric alone. Holmes argues that through the mechanism of official statements and ‘forward guidance’, economic agents are prevailed upon to ‘assimilate policy intentions as their own personal expectations’, and so ‘do the work of the central bank’ (p. 177). ‘Communications’ then become ‘the decisive means to achieve the ends of policy’ (p. 178). Perhaps the most famous example of this in recent years was the enormous market effectiveness of the simple statement by Mario Draghi in 2012 that the European Central Bank would do ‘whatever it takes’ to rescue the euro. Holmes’ chapter also points to what is perhaps the most widespread function of narratives—that of making sense of contingent events and uncertain futures, and providing the basis for conversational interaction between different social and economic players. Holmes employs the example of the Bank of England using a network of agencies to access ‘stories continually generated outside the central bank from situated actors who are themselves orchestrating and evaluating economic and financial conditions’ (p. 188); and he argues that such discursive input is critical to the central bank’s ability to decipher emerging trends in the economy. Werner Reichmann makes a similar point in Chapter 5 when he discusses the importance of conversation as part of the ‘interactional process’ of forecasting in German economic institutes: conversation with external contacts helps the forecasters notice emerging developments they would otherwise have missed, and leads to greater appreciation of the intentions of key players in the economy—thereby improving the epistemic quality of their forecasts. This focus on conversations, and on stories accessed from the market, as input to the policy judgements of central bankers and the predictions made by professional economic forecasters might come as a surprise to many lay readers and econometricians alike. Reichmann argues that econometric models are, in fact, ‘increasingly taking a back seat in the process of manufacturing a forecast’ (p. 110) and being supplemented by a generous measure of ‘foretalk’. This partial shift from mathematical modelling using established data series to relying on more discursive inputs makes perfect sense when forecasters are tasked with spotting newly emerging patterns, interpreting ambiguous social data, and making difficult judgements about the relative 22

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

An Introduction to Uncertain Futures

importance of conflicting factors driving the economy. As Gary Saul Morson and Morton Shapiro (2017, 39) argue, the degree of ‘narrativeness’ required for convincing explanations (and predictions) can be calibrated by the degree to which contingency, indeterminacy, and culture play a role in the economy; and the relative importance of qualitative judgement versus calculation techniques ought to be similarly calibrated. A further important function of narratives is to help actors instil confidence and generate sufficient conviction to act. This function has been extensively analysed by David Tuckett (2011) and by Kimberly Chong and David Tuckett (2015). They explain how fund managers use stories to deflect anxiety about the future and develop the conviction required to make investment decisions despite uncertainty. Tuckett draws on this research in Chapter 3 to develop a more general ‘Conviction Narrative Theory’ (CNT), with the help of recent findings in psychology and neurophysiology. Tuckett starts from the twin premises that, in conditions of radical uncertainty, it would not be rational to base decisions on optimizing available information without interpretive and imaginative effort to construct meaning out of the fragmentary indicators to hand; and that actors suffer inevitable anxiety in the face of the unknown and must find ways to conquer doubt about the wisdom of any choice of action. He then builds on extensive research on the role of narratives in constructing meaning and simulating possible outcomes that can be ‘felt as close to real experience’ (p. 71), and integrates recent advances in our understanding of the crucial part played by emotions in decision-making. In a nutshell, Tuckett argues that the conviction that narratives generate is a function of the ratio of the ‘approach’ versus ‘avoidance’ emotions that their sense-making function engenders in those who internalize them. By focusing on the way ‘narratives manage anticipations of gain and loss’ and ‘support action emotionally’, Tuckett argues (p. 74) that CNT can be seen as bringing Keynes’ emphasis on the vital role played by ‘animal spirits’ in line with contemporary research in brain science. Tuckett concludes his chapter by arguing that CNT can be used to develop statistical forms of discourse analysis of news feeds and other types of narrative to measure ‘shifts in the balance of approach and avoidance emotions’ registered in these texts as a tool for understanding and even predicting an economy’s evolution (p. 77). This is an example of the new form of ‘narrative economics’ advocated by Robert Shiller. Shiller (2017, 2, 3, 11) notes that the ‘human brain has always been highly attuned towards narratives . . . to justify ongoing actions’—narratives that ‘are mixtures of fact and emotion and human interest’; and he goes on to argue that economics should include ‘serious quantitative study’ of the widespread narratives associated with economic fluctuations in the hope of better understanding the causal 23

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

relationship between them. Shiller is particularly interested in the role of popular narratives as ‘major vectors of rapid change in culture, in zeitgeist, and ultimately in economic behaviour’ (10), and in how these narratives are subject to forms of emotional contagion amenable to study with models borrowed from epidemiology. The significance of ‘new era’ and other narratives becoming conventional frames involving group emotions16 at a macrolevel was underlined by George Akerlof and Robert Shiller (2009) in their analysis of the post-2007 financial crisis, where they demonstrated how stories—and the emotions of confidence or fear attaching to them—can spread ‘like viruses’ (56). The trigger for a sudden shift in what counts as the conventional narrative—or for a new narrative to become suddenly contagious—is very often an unexpected and economically damaging shock, or an emotionally salient event, that challenges the currently dominant narrative’s sense-making capacity. The consequent ‘narrative revision’ (King 2017, 332) may help cause an abrupt change in economic actors’ perceptions of the future and in their behaviour, leading to profound market (and political) instability. Robert Boyer presents a different version of this new ‘narrative economics’ in Chapter 2, where he provides ‘a history of the narratives associated with a succession of recent socio-economic regimes’ (p. 39). He argues that the modern finance-led economic system, together with increasingly interdependent global markets and high levels of innovation, has fostered unprecedented levels of uncertainty. This makes economic actors even less able to operate with the sort of rational expectations assumed by most economic models and renders them ever more dependent on a series of ‘grand narratives’ that succeed for a time in mobilizing investment and instilling confidence. Examining the temporary sway of such narratives as ‘Japan number one, the new economy, and the omniscience of financial markets’ (p. 41), Boyer develops the thesis that capitalist economies coordinated by such stories are prone to crises, as the failure of each simplistic narrative in turn to capture key dynamics becomes painfully clear and expectations are disappointed.17

Narratives as Coordination Device and Instrument of Power Boyer’s analysis of the ‘Socialization of Expectations by Narratives’ (a section beginning on p. 46) points to another major function of narratives—that of 16 For an analysis by Tuckett of the related notion of ‘groupfeel’ and why it can be destabilizing to markets, see p. 76. 17 For further analysis of the destabilizing nature of grand narratives and ‘analytical monocultures’, see p. 29.

24

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

An Introduction to Uncertain Futures

enabling such coordination of beliefs and action as is necessary or desirable. Of all the coordination problems facing economic actors, those entailed by the indeterminacy of beliefs and expectations in dynamic capitalist markets are the most important. As a number of economists have pointed out (Arthur 2015; Keynes 1936; Orléan 2014), actors operating in a complex and innovative economic system are not only uncertain about the situation they face; they are also uncertain about the contingent expectations and beliefs that other actors are forming in the same situation. When there is no fixed anchor in (future) reality for expectations, the uncertainty faced by any particular economic agent(s) is compounded by uncertainty about the action-guiding beliefs and expectations of other agents. This self-reinforcing uncertainty problem makes it potentially difficult to coordinate action in economic settings. One of the main functions of economic forecasts—and the narratives embedded within them—is to help solve this coordination problem through the creation of shared expectations about the future. As Pilmis puts it, whatever their ‘notoriously poor track record’, actors ‘need forecasts anyway to design strategies’ (p. 126). The future cannot be known and forecasts may turn out ex post to be misleading; but at the time they are produced, they may still—if widely believed—allow for the coordination of expectations and behaviour. The Delphi method of consensus creation in the area of technology forecasts similarly aligns expectations about the range of possible technological futures (Andersson 2013). The aim in this case is not to produce one view of the future but to enable actors to ‘locate and build their own perspective in a structured field of expectations’ (Beckert 2016, 235). The thriving field of the Sociology of Expectations (Brown et al. 2000), which is closely aligned with Science and Technology Studies (STS), focuses on the role of ‘promissory stories’ not only as scripts providing a logic of action for innovators themselves but also as a crucial mechanism for coordinating the allocation of resources across and between whole sectors of the economy. It is, for example, through road maps of expected technological progress that government resources are allocated to research and development. Furthermore, the research trajectories of each industrial sector are shaped by the competition between stories of possible future developments that influence individual firms in their decisions on technological investment (Mützel 2010). For example, as Timur Ergen shows in Chapter 14, the trajectory of renewable energy development in the United States during the 1970s and 1980s was a product of the contingent expectations of relevant political and economic actors concerning future technological progress. Such cognitive frames and narratives determine which development paths are followed; and, equally crucially, they ensure that other possible worlds do not receive the resources required to investigate them, with the result that their innovative potential remains unexplored. 25

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

The coordination properties of narratives—and the forecasts embedded in them—also make them an instrument of government policy and power. The clearest example of this in early twenty-first century capitalist economies is the use by leading central banks of ‘forward guidance’ and official statements to ‘cajole expectations in a particular direction’ (Holmes, p. 177). Indeed, Holmes quotes Ben Bernanke of the US Federal Reserve observing recently that ‘monetary policy is 98 percent talk and only two percent action’ (p. 174). It is important to note, however, that certain preconditions must be met if rhetoric alone is to be an instrument of power—particularly in the hands of the very elites and experts increasingly derided in popular discourse. In the case of central banking, for example, the vision of the future—the economic imaginary—contained in forward guidance will be credible only if the central bank enjoys a high forecasting reputation or demonstrably has the potential fire-power of monetary policy at its disposal to back up its pronouncements. In Chapter 9, Benjamin Braun analyses several recent threats to the credibility of central bank narratives and forecasts that may increasingly weaken their market power. As conventional interest rate adjustments at the short end of the yield curve reached their limits following the 2008 financial crisis, central banks increasingly used a mixture of forward guidance and quantitative easing to force long-term yields (previously set by the market) lower to support investment and asset prices. Braun argues that central banks have incurred several costs by using this combination of rhetorical and ‘hydraulic’ instruments to influence market expectations. First, the rhetorical use by central banks of in-house forecasts in their forward guidance (often in the name of transparency) may have had the perverse effect of weakening their ‘epistemic authority’ (p. 208) when the forecasts were proven wrong, while damaging the credibility of the forecasts as wholly independent of policy concerns. Second, the relative success of the central banks despite this in controlling long-term bond yields has brought about ‘the loss of informational content in financial asset prices’. As in analogous cases of central planning where prices are directly controlled by the state, bond prices may cease, in a regime of quantitative easing, to be an effective ‘barometer of the decentralized beliefs and actions of myriad market actors’ (p. 212). It is increasingly clear that central banks are paying a high price in their bid to maintain their pre-eminence in the discursive game of influencing expectations and shaping the future. The direction of capitalist economies is, in many ways, the outcome of a struggle between different state and market actors to establish their fictional expectations as the most credible; and all credibility is fragile when dealing with uncertain futures. Since the economic forecasts and associated narratives that shape expectations and structure behaviour often have heavy distributional consequences, they also play an important role in the struggles for political power. The EU 26

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

An Introduction to Uncertain Futures

referendum battle in the United Kingdom in 2016, for example, was partly a battle over the credibility of forecasts about the economic consequences of Brexit. In the end, the forecasts of large numbers of experts, forecasting institutes, and government agencies warning of dire consequences were trumped by a simpler narrative of Europe’s (alleged) relative economic decline and vaguely defined export opportunities for the United Kingdom elsewhere, once free of European Union bureaucracy. Above all, the narratives of ‘taking back control’ and kicking elites proved more credible and emotionally appealing than official predictions of economic woe. In Chapter 4, Jenny Andersson examines the use of narratives and ‘scenario crafting’ in a quite different geo-political setting—the Arctic. The competing claims over the Arctic cannot be understood simply as a classic game of bigpower politics and competing economic interests, because the Arctic is a space whose contested post-climate-change future must be imagined and constructed through competing images, narratives, and orders of worth. The struggle to establish rival claims is fought in part, Andersson argues, with the help of predictive technologies that, through a ‘highly selective sorting of available images of the future’, seek to establish the dominance of images of the future that suit particular interests (p. 86). But, crucially, since ‘future opportunities do not by definition yet exist’ and require active definition, the relevant interests in play are also partly constituted by ‘a repertoire of future-making’—ranging from ‘quantitative forecasts and prospecting for natural resources, to highly narrative genres of nation branding’, ‘the mobilization of historical memory’, and normative images of pristine wilderness requiring protection (p. 85). Through this process of building expectations, interested actors attempt to ‘close’ the open future politically by aligning actors behind specific expectations and scenarios. Andersson examines, in particular, the construction of Swedish interests in the Arctic through imaginaries of technology-based market opportunities and the use of historical narratives to project Sweden’s Arctic identity and territorial claims. As in the case of Brexit, political power rests with those able to make their narratives, imaginaries, and expectations count (Beckert 2016, 80). The battle over the relative validity of different economic theories and models can also have significant political significance and power implications. This is because theories and models affect—directly or indirectly—the beliefs, expectations, and therefore the voting patterns of the electorate. For example, public choice theory—a key part of the neoclassical economic paradigm— predicts that government officials and politicians will further the public interest only if it is in their individual interest to do so. But this social-science theory—with only patchy success in predicting actual behaviour and designing governance reforms—has been internalized and promulgated by rightwing politicians in order to help corrode trust in government elites and the 27

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

efficacy of big government (Bronk 2009, 6). Establishing the supremacy of a theoretical frame or model matters politically because the successful shaping of expectations and interpretations is a key aspect of exercising power. In a similar vein, Keynes (1936, 33) famously argued that Ricardian economics became so dominant because it was useful to those in authority, since ‘it could explain much social injustice and apparent cruelty as an inevitable incident in the scheme of progress, and the attempt to change such things as likely on the whole to do more harm than good’. Michel Foucault took an even more extreme view of this relationship between theory and power, and argued, in the words of J. G. Merquior (1985, 85): ‘Theory is not like a pair of glasses; it is rather like a pair of guns; it does not enable one to see better but to fight better’.

The Limits of Performativity and the Need for Dissonance Behind the ‘politics of expectations’ (Beckert 2016), and the power dynamics associated with the theoretical and narrative frames that shape expectations, lies an assumption that theories, narratives, and expectations tend to create futures in their own image—tend, that is, to be self-fulfilling prophecies (Merton 1957), so long as they are internalized by sufficient numbers of actors. The impact of fictional expectations on the future to which they relate is indeed one of their central features. Forecasts, for example, may influence the future by shaping the strategies and actions of economic actors in the present. In this way, expectations can, in a loose sense, have the sort of ‘performative’ effect on market behaviour that Michel Callon (1998) and Donald MacKenzie (2006) articulated in relation to the theories and models of economics and finance theory. Considerable care is needed, however, in applying the concept of performativity to expectations. The impact of expectations, business plans, and forecasts is rarely to create a future that accords ex post with what was expected ex ante. Instead, the so-called ‘performative’ impact of guiding narratives, forecasts, plans, models, and other calculative devices is itself the cause of a new knowledge problem for economic actors (Bronk 2013b, 345): it is impossible to know with any precision whether thinking something is feasible, and predicting or expecting it to happen, will be enough to make it so. This is partly because it is often hard to know whether other relevant actors share your action-guiding beliefs. It is also because the specific theories, models, and perspectives used in the formation of expectations are bound to have limitations in the aspects of reality they encapsulate—particularly in relation to the unknowable future—and may therefore be misleading in important respects. 28

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

An Introduction to Uncertain Futures

There are often other reasons why forecasts and models tend to be ‘counterperformative’ (MacKenzie 2006)—that is, bring about a future opposed to the one that is forecast. In particular, forecasts based on a central scenario are frequently used intentionally to foster political and economic action designed to ensure that what was forecast does not materialize. Examples include climate-change forecasts based on scenarios of continued high carbon intensity providing the impetus for coordinated environmental policies; or budget deficit forecasts on the assumption of no cuts in spending being used by governments to justify austerity. Similarly, bank stress tests are carried out by regulators to calculate capital adequacy under certain extreme scenarios, partly as a means of testing hypothetical resilience, but primarily to convince company boards and politicians of the need to raise new capital. A different problem with the ‘performative’ impact of expectations (and the calculative devices supporting them) is that their very success in shaping the future in the short-term may breed long-term instability. The impact that contingent narratives, calculations, and other forms of structured expectations have on the future depends critically, in many cases, on the number of actors who internalize them in their decision-making processes; hence the emphasis in markets and politics on persuading others of the merits of a particular narrative or forecast. Moreover, actors also have an incentive for mimetic convergence (Orléan 2014) because it often allows them to secure resources and to profit from aligning their strategies with the dominant trend. But if all relevant actors come to internalize the same perspective, and use the same metaphors and models to structure their vision and data, then they will all suffer from similar cognitive biases, given the inevitable limitations and distortions implied by any single method of visualizing the future.18 The result of such ‘analytical monocultures’ is not only shared cognitive blind spots but also high correlations in behaviour, which may then further entrench shared mental models in ‘reflexive feedback loops’ (Bronk 2013b, 347). In other words, homogenization of how actors think about the future (and consequent shared analytical routines) may lead also to high correlations in how they act, and vice versa. If the highly correlated behaviour later proves to be based on analytical errors, the result will almost inevitably be a period of market instability, as markets adjust rapidly to using a new model or narrative in the construction of shared expectations. Chapters 2 and 3, by Boyer and Tuckett, respectively, both examine the role of widely shared narratives and accompanying group emotions in causing cycles of boom and bust. The manifest dangers of analytical monocultures—and of the market, regulatory, and institutional conditions that encourage them (Bronk and Jacoby 18 For discussion of the way in which models (and the metaphors embedded in them) structure perception and analysis as well as action, see Bronk (2010, 103–6).

29

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

2016)—present emotional, practical, and cognitive challenges for economic actors facing uncertain futures. At an emotional level, entrepreneurs and investors need, as Tuckett argues, to find ways to manage their anxiety and disorientation if they are to have sufficient conviction to act; and adopting the current conventional or ‘best practice’ model or narrative is often seen as one way to achieve this. At the same time, although following the trend may pay off in the short-term, it is essential for investors to display a ‘critical attitude’ and constant ‘self-doubt’ (Soros 1998, 25f) if they are not to miss crucial pointers to unexpected developments simply because their preferred investment narrative, model, or conceptual framework has no place for them (Bronk 2010, 105). More broadly, there is in many areas of economic activity and macroeconomic policy a trade-off between the positive coordination effects of stabilizing and homogenizing expectations (on the one hand) and the epistemic value of retaining a diversity of expectation-guiding cognitive, institutional, and calculative frameworks (on the other). Uncertain futures require actors to be imaginatively receptive to newly emerging patterns and able constantly to find innovative responses to novel challenges. David Stark (2009, xvi, 16) argues that successful firms are those that use ‘the ambiguity of multiple evaluative principles to navigate through uncharted territory’, and ‘foster a generative friction that disrupts received categories of business as usual and makes possible an ongoing recombination of resources’. It is precisely this ‘dissonance’ between different frameworks of interpretation that helps economic actors spot, and then make sense of, the unexpected; and it is a clash of outlooks that generates novel interpretations and innovative ideas. Since no single perspective or theoretical framework can make sense of all the incommensurable aspects of socio-economic reality, economic actors stand to benefit from using ‘different cognitive spectacles’— different explanatory models—as diagnostic tools of emerging trends (Bronk 2009, 293). This is one of the principal arguments in favour of modelling pluralism and a multi-disciplinary approach to the study of most problems in applied economics and policy circles.19 All models or narratives are simplifications of reality that focus attention on particular causal mechanisms or aspects of reality to the exclusion of others.20 Being exposed to a new model or forecasting 19 For discussion of how to operationalize ‘disciplined eclecticism’ in applied research and combine the benefits of academic specialization with open-minded assessment of problems, see Bronk (2009, 276–89). 20 Peter Diamond (2011, 1045f)—in an article based on his 2010 Nobel Lecture—links the need for modelling pluralism explicitly to the incompleteness of models. As he puts it: ‘The complexity of the economy calls for the use of multiple models that address different aspects of the determinants of unemployment (and other) outcomes . . . [T]aking a model literally is not taking a model seriously. It is worth remembering that models are incomplete—indeed, that is what it means to be a model.’

30

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

An Introduction to Uncertain Futures

narrative provides, as Haldane puts it, ‘a different—complementary—lens through which we might make sense of our dappled economic and financial world’ (p. 151). In innovative sectors, entrepreneurs often derive benefit from agreeing on a narrative that stabilizes expectations and assigns a common logic of action; but they must also display a willingness to engage in ‘narrative shifting’ (Lane and Maxfield 2005, 16) if they are to react swiftly enough to the evolving challenges presented by rapid technological innovation. In Chapter 14, Ergen explores how difficult it is to navigate this trade-off between the coordination benefits of shared mental models (or homogenous expectations) and the adaptation benefits of cognitive dissonance (or strategic flexibility) in his study of innovative energy sectors in the United States in the 1970s and 1980s. On the one hand, commercialization of these new technologies proved impossible without a significant degree of coordination of expectations and narratives, and a degree of consensus about the future shape of the technologies concerned. Investment was simply not forthcoming without some stabilization of expectations. On the other hand, aligning expectations too thoroughly threatened to undermine the diversity of trial and error activities essential to open-ended search for technological alternatives and induce a degree of cognitive lock-in that would leave parts of the industry seriously wrong-footed when the technology changed due to further innovation. Ergen concludes, however, that this dilemma is less clear-cut than sometimes supposed because—especially in high-tech sectors—even the interpretation of new technological development possibilities in practice requires a heavy commitment of resources to specific paths of research and development, along with feedback from nascent markets. In the area of macroeconomics, policy-makers face a similar dilemma between the benefits of coordination and the cognitive benefits of diversity of expectations. Indeed, the dilemma could be said to underlie the decades-old battle between the followers of Keynes and Hayek. Both Keynes and Hayek took the central importance of uncertainty as their starting point, but they drew startlingly different conclusions. For Keynes and his followers, the prevalence of uncertain futures underscores the paramount need for government intervention to stabilize (and in some cases reset) expectations around a planned target, in order to avoid market instability and depressed investment. By contrast, Hayek maintained that such intervention almost always does more harm than good, since governments never have sufficient knowledge of the long-term implications of their centrally planned interventions. This, he argued, is because the aggregate statistics that government agencies use cannot capture the decentralized and constantly changing knowledge of individuals operating in dynamic situations (Bronk 2013a, 89); and also because the interventions themselves compromise the ability of market prices 31

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

to signal newly emerging trends. In Chapter 9, Braun considers whether central banks now run the risk of making a version of this mistake by undertaking a form of central planning in their programmes of quantitative easing. The answer to these dilemmas and trade-offs may be as simple as it is frustratingly vague: capitalist economies work best at the shifting boundary between shared mental models and cognitive dissonance, between homogenous expectations and conflicting opinions, and between stable strategies (or targets) and flexible blueprints (or goals). Economic actors need to possess enough knowledge of how others will behave, and enough certainty about the regulatory, policy, and institutional context in which they operate, to have the required confidence to act and sufficient time for new strategies to take hold. Without a robust institutional framework, a rapid succession of narratives may shift the economy from one hype to the next, and from one crisis to another. But when the benefits of stability of expectations and coordination of behaviour are bought at the price of inflexible strategies and mental models that ignore the inevitable indeterminacy implicit in the uncertain futures we face, then capitalist economies will also be doomed to have periodic crises as expectations adjust suddenly to novelty and surprise.

Conclusion The focus in this volume on imaginaries, narratives, and fictional expectations lays the foundations for an alternative set of microfoundations for economics that is more suitable than rational expectations theory for analysing decisionmaking in conditions of fundamental uncertainty. Such uncertainty is not the occasional result of random exogenous shocks but is endogenous to capitalist systems characterized by innovation, novelty, and imaginative economic actors reacting to the indeterminate predicament in which they find themselves. Economic actors face the challenge of interpreting dynamic markets and coordinating their actions with at least some others. They must decide how to act despite the uncertainty they face. To do so they combine imaginaries with shared narratives and calculative devices. Calculation plays a key role in developing and stress-testing imaginaries, analysing possible scenarios, and legitimating strategies; but it cannot provide a single determinate set of expectations that could be characterized as uniquely rational. In a world of uncertain futures there is no optimal strategy. The expectations of economic actors are necessarily contingent and frequently contain an element of imagination—the ultimate source of indeterminacy in the economy. The contributing authors of this volume develop this theoretical framework in empirical settings. They show the many different ways in which market 32

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

An Introduction to Uncertain Futures

actors in practice use imaginaries, narratives, and calculative devices to construct expectations, coordinate action, and influence outcomes. The spectrum of settings analysed stretches from policy-makers and central banks, through credit rating agencies and economic forecasters, to financial investors, entrepreneurs, and technology innovators. It is hoped that the book will generate wider interest within the social sciences in analysing the mechanisms actually employed by economic actors to support their decision-making in conditions of uncertainty, and thereby improve our understanding of the dynamics of capitalist economies.

Bibliography Akerlof, George A. 1970. ‘The Market for “Lemons”: Quality Uncertainty and the Market Mechanism’. The Quarterly Journal of Economics 84 (3): pp. 488–500. Akerlof, George A., and Robert J. Shiller. 2009. Animal Spirits: How Human Psychology Drives the Economy and Why It Matters. Princeton, NJ: Princeton University Press. Andersson, Jenny. 2013. Forging the American Future: RAND, the Commission for the Year 2000 and the Rise of Futurology. Paris: Sciences Po. Arthur, W. Brian. 2015. Complexity and the Economy. New York: Oxford University Press. Beckert, Jens. 2016. Imagined Futures: Fictional Expectations and Capitalist Dynamics. Cambridge, MA: Harvard University Press. Beunza, Daniel, and David Stark. 2012. ‘From Dissonance to Resonance: Cognitive Interdependence in Quantitative Finance’. Economy and Society 41 (3): pp. 383–417. Boulding, Kenneth E. 1961. The Image: Knowledge in Life and Society. Ann Arbor, MI: University of Michigan Press. Bourdieu, Pierre. 2000. Pascalian Meditations. Translated by Richard Nice. Stanford, CA: Stanford University Press. Bourdieu, Pierre. 2005. ‘Principles of an Economic Anthropology’. In: The Handbook of Economic Sociology, edited by Neil Smelser and Richard Swedberg, pp. 75–89. Princeton, NJ: Princeton University Press. Bronk, Richard. 2009. The Romantic Economist: Imagination in Economics. Cambridge: Cambridge University Press. Bronk, Richard. 2010. ‘Models and Metaphors’. In The Economic Crisis and the State of Economics, edited by Robert Skidelsky and Christian Westerlind Wigstrom, pp. 101–9. New York: Palgrave Macmillan. Bronk, Richard. 2011a. ‘Uncertainty, Modelling Monocultures and the Financial Crisis’. The Business Economist 42 (2): pp. 5–18. Bronk, Richard. 2011b. ‘Epistemological Difficulties with Neo-Classical Economics’. Southern Economic Association Conference Paper, 19–21 November, Washington D.C. http://eprints.lse.ac.uk/39423/. Bronk, Richard. 2013a. ‘Hayek on the Wisdom of Prices: A Reassessment’. Erasmus Journal for Philosophy and Economics 6 (1): pp. 82–107.

33

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures Bronk, Richard. 2013b. ‘Reflexivity Unpacked: Performativity, Uncertainty and Analytical Monocultures’. Journal of Economic Methodology 20 (4): pp. 343–9. Bronk, Richard, and Wade Jacoby. 2016. ‘Uncertainty and the Dangers of Monocultures in Regulation, Analysis, and Practice’. MPIfG Discussion Paper 16/6, Max Planck Institute for the Study of Societies, Cologne. Brown, Nick, Brian Rappert, and Andrew Webster. 2000. Eds. Contested Futures: A Sociology of Prospective Techno-Science. Aldershot, UK: Ashgate. Buchanan, James M., and Viktor J. Vanberg. 1991. ‘The Market as Creative Process’. Economics and Philosophy 7 (2): pp. 167–86. Callon, Michel. 1998. The Law of the Markets. Oxford: Blackwell. Campbell, Colin. 1987. The Romantic Ethic and the Spirit of Capitalism. Oxford: Blackwell. Chong, Kimberly, and David Tuckett. 2015. ‘Constructing Conviction through Action and Narrative: How Money Managers Manage Uncertainty and the Consequences for Financial Market Functioning’. Socio-Economic Review 13 (2): pp. 309–30. Davidson, Paul. 1996. ‘Reality and Economic Theory’. Journal of Post Keynesian Economics 18 (4): pp. 479–86. Davidson, Paul. 2010. ‘Risk and Uncertainty’. In The Economic Crisis and the State of Economics, edited by Robert Skidelsky and Christian Westerlind Wigstrom, pp. 13–29. New York: Palgrave Macmillan. Dequech, David. 2001. ‘Bounded Rationality, Institutions, and Uncertainty’. Journal of Economic Issues 35 (4): pp. 911–29. Diamond, Peter. 2011. ‘Unemployment, Vacancies, Wages’. American Economic Review 101 (4): pp. 1045–72. Gigerenzer, Gerd. 2013. Risiko: Wie man die richtigen Entscheidungen trifft. Munich: Bertelsmann. Giraudeau, Martin. 2010. ‘Performing Physiocracy: Pierre Samuel Du Pont de Nemours and the Limits of Political Engineering’. Journal of Cultural Economy 3 (2): pp. 225–42. Hamilton, Valerie, and Martin Parker. 2016. Daniel Defoe and the Bank of England: The Dark Arts of Projectors. Winchester: Zero Books. Hayek, F.A. 1945 [1948]. ‘The Use of Knowledge in Society’. In Individualism and Economic Order, pp. 77–91. Chicago, IL: University of Chicago Press. Hayek, F.A. 1952 [2010]. ‘Scientism and the Study of Society’. In The Collected Works of F. A. Hayek. Vol. 13: Studies on the Abuse and Decline of Reason, edited by Bruce Caldwell, pp. 75–166. London: Routledge. Hazlitt, William. 1805 [1998]. ‘An Essay on the Principles of Human Action’. Reprinted in The Selected Writings of William Hazlitt, Vol. 1, edited by Duncan Wu. London: Pickering & Chatto. Hodgson, Geoffrey. 2011. ‘The Eclipse of the Uncertainty Concept in Mainstream Economics’. Journal of Economic Issues 45 (1): pp. 159–75. Jasanoff, Sheila, and Sang-Hyun Kim. 2009. ‘Containing the Atom: Sociotechnical Imaginaries and Nuclear Power in the United States and South Korea’. Minerva 47, 119–46. Kahneman, Daniel, and Amos Tversky. 2000. Choices, Values, and Frames. New York: Cambridge University Press.

34

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

An Introduction to Uncertain Futures Keats, John. 1817 [1998]. ‘Letter to George and Tom Keats’, 21 December. Extract reprinted in Romanticism: An Anthology, edited by Duncan Wu, p. 1019. 2nd edn. Oxford: Blackwell. Keynes, John Maynard. 1936. The General Theory of Employment, Interest and Money. London: Macmillan. King, Mervyn. 2017. The End of Alchemy: Money, Banking and the Future of the Global Economy. London: Abacus. Knight, Frank. 1921. Risk, Uncertainty and Profit. Boston: Houghton Mifflin. Knorr Cetina, Karin. 2007. ‘Culture in Global Knowledge Societies: Knowledge Cultures and Epistemic Cultures’. Interdisciplinary Reviews 32 (4): pp. 361–75. Lane, David A., and Robert R. Maxfield. 2005. ‘Ontological Uncertainty and Innovation’. Journal of Evolutionary Economics 15 (1): pp. 3–50. Lucas, Robert E. 1981. Studies in Business Cycle Theory. Cambridge, MA: MIT Press. Luhmann, Niklas. 1976. ‘The Future Cannot Begin: Temporal Structures in Modern Societies’. Social Research 43 (1): pp. 130–52. MacKenzie, Donald. 2006. An Engine, Not a Camera: How Financial Models Shape Markets. Cambridge, MA: MIT Press. MacKenzie, Donald. 2011. ‘The Credit Crisis as a Problem in the Sociology of Knowledge’. American Journal of Sociology 116 (6): pp. 1778–841. McCloskey, Deirdre N. 1998. The Rhetoric of Economics. 2nd edn. Madison, WI: University of Wisconsin Press. Merquior, J. G. 1985. Foucault. Fontana Press. Merton, Robert K. 1957. Social Theory and Social Structure. Glencoe, IL: Free Press. Mirowski, Philip. 1989. More Heat than Light: Economics as Social Physics, Physics as Nature’s Economics. Cambridge: Cambridge University Press. Morson, Gary Saul, and Morton Shapiro. 2017. Cents and Sensibility: What Economics Can Learn from the Humanities. Princeton, NJ: Princeton University Press. Muth, John F. 1961. ‘Rational Expectations and the Theory of Price Movements’. Econometrica 29 (3): pp. 315–35. Mützel, Sophie. 2010. ‘Koordinierung von Märkten durch narrativen Wettbewerb’. In Wirtschaftssoziologie. Kölner Zeitschrift für Soziologie und Sozialpsychologie, Sonderheft 49/2009, edited by Jens Beckert and Christoph Deutschmann, pp. 87–106. Wiesbaden: VS Verlag. North, Douglass C. 2005. Understanding the Process of Economic Change. Princeton, NJ: Princeton University Press. Offe, Claus. 1998. ‘Political Economy: Sociological Perspectives.’ In A New Handbook of Political Science, edited by Robert E. Goodin and Hans-Dieter Klingemann, pp. 675–90. Oxford: Oxford University Press. Orléan, André. 2014. The Empire of Value: A New Foundation for Economics. Cambridge, MA: MIT Press. Polanyi, Karl. 1944 [1957]. The Great Transformation. Boston: Beacon Press. Power, Michael. 2007. Organized Uncertainty. Oxford: Oxford University Press. Rebonato, Riccardo. 2007. Plight of the Fortune Tellers: Why We Need to Manage Financial Risk Differently. Princeton, NJ: Princeton University Press.

35

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures Ricoeur, Paul. 1991. ‘Imagination in Discourse and in Action’. Translated by Kathleen Blamey. In: From Text to Action, pp. 168–87. London: Athlone. Sargant, Thomas J. 2008. ‘Rational Expectations’. In: Concise Encyclopedia of Economics, edited by David R. Henderson, pp. 432–4. Indianapolis: Liberty Fund. Schumpeter, Joseph A. 1943 [1976]. Capitalism, Socialism and Democracy. London: Routledge. Schwartz, Barry. 2004. The Paradox of Choice: Why More Is Less. New York: Harper Collins. Sewell, William H. 2008. ‘The Temporalities of Capitalism’. Socio-Economic Review 6 (3): pp. 517–37. Shackle, George. 1972 [1992]. Epistemics and Economics: A Critique of Economic Doctrines. New Brunswick: Transaction Publishers. Shackle, George. 1979. Imagination and the Nature of Choice. Edinburgh: Edinburgh University Press. Shiller, Robert J. 2017. ‘Narrative Economics’. Cowles Foundation Discussion Paper, no 2069, Yale University. Simon, Herbert A. 1957. Models of Man. New York: Wiley. Skidelsky, Robert. 2009. Keynes: The Return of the Master. London: Allen Lane. Snow, C.P. 1959. The Two Cultures and the Scientific Revolution. Cambridge: Cambridge University Press. Soros, George. 1998. The Crisis of Global Capitalism: Open Society Endangered. London: Little, Brown. Stark, David. 2009. The Sense of Dissonance: Accounts of Worth in Economic Life. Princeton, NJ: Princeton University Press. Sutton, John. 2000. Marshall’s Tendencies: What Can Economists Know? Cambridge, MA: MIT Press. Tuckett, David. 2011. Minding the Markets: An Emotional Finance View of Financial Instability. London: Palgrave Macmillan. Turner, Adair. 2016. Between Debt and the Devil: Money, Credit, and Fixing Global Finance. Princeton, NJ: Princeton University Press. van Lente, Harro and Arie Rip. 1998. ‘Expectations in Technological Developments: An Example of Prospective Structures to Be Filled in by Agency’. In: Getting New Technologies Together: Studies in Making Sociotechnical Order, edited by Cornelis Disco and Barend van der Meulen, pp. 203–29. Berlin: de Gruyter. Weber, Max. 1930 [1992]. The Protestant Ethic and the Spirit of Capitalism. Translated by Talcott Parsons. London: Routledge.

36

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Section I The Nature of Expectations in Modern Political Economies

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

2 Expectations, Narratives, and Socio-Economic Regimes Robert Boyer

Introduction This chapter provides a history of the narratives associated with a succession of recent socio-economic regimes. It argues that radical uncertainty has reached an unprecedented level thanks to radical innovations and the complexity of domestic and international interdependencies. This uncertainty cannot be overcome by models based on the currently dominant rational expectations hypothesis. Actors are unable to base their estimates on such determinist models because the past is a poor predictor of future socioeconomic regimes. The current agony of the rational expectations hypothesis has opened up a wide space for economic narratives—generally fairly simple in nature—that promise a drastic reduction of radical uncertainty and systemic complexity. The chapter reviews how these narratives are deployed within the business community (storytelling as a method for convincing markets to finance daring and uncertain projects), and among economic policy-makers. The chapter shows that imaginaries (and the narratives that embody them) are crucial in moving capitalist spirits; but it also demonstrates that the related tendency for hegemonic imaginaries or grand narratives to emerge leads to recurring financial and economic crises. In the early 2000s, the economic profession was very proud of the discipline’s achievements. Due to advances in conceptualization, the sophistication of econometric techniques, and the availability of real-time data, macroeconomics had acquired the status of a quasi-natural science. Only details were still to be worked out (Blanchard 2008). Economists were pleased to be one, if not the only, social science able to deliver causal explanations and precise forecasts. This hope was destroyed by the bursting of the subprime bubble and

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertain Futures

its development into a world economic crisis that has yet to be overcome even a decade later. The crisis was a shock for policy-makers: how and why had the scientific discipline of economics been so myopic that it failed to diagnose a speculative real estate bubble (Greenspan 2013) that any American taxi driver might have sensed, but not the head of the Federal Reserve? In order to fight against cumulative depression, central bankers had to give up their previous conservative principles. They had to buy bonds, including toxic ones, to sustain the liquidity and solvency of commercial banks. Unfortunately, nobody knows the way out of this policy of ‘quantitative easing’, a monetary approach that stands in clear violation of the rational expectations hypothesis that remains central to contemporary macroeconomics. This chapter proposes the concept of expectation regimes in order to analyse the succession of distinct periods in which different sets of expectations prevail in the socio-economic order. Expectations can be adaptive or rational within a quasi-stationary world. When key political compromises build complex but coherent institutional architectures, the expectation regime is context-dependent. By contrast, when finance is the hierarchically dominant institution, the radical uncertainty of economic futures has to be reduced by the invention and diffusion of narratives that range from business plans to society-wide utopias. Because self-fulfilling prophecies are exceptional (see the chapter by Esposito in this volume), financial and economic crises typically mark the shifts in narratives required to cure capital imbalances. Expectation regimes that are historically contingent and based on narratives provide an alternative set of microfoundations that—unlike those envisaged by standard economic theories—are able to account for the central importance of uncertain futures in modern economies. The inadequacy of standard economic theories in this regard is well understood. For example, the Walrasian auctioneer can do no more than organize the synchronicity of exchanges, without any concern for uncertainty and different time frames in the various spheres of society. This means that views about the future are excluded from analysis, although they are crucial in any market and even more in a capitalist economy. Walrasian and by extension many standard economic theories have limited applicability because they are unable to deal with time (Sapir 2000). They discard the historical dimension of economic processes (including progressive learning from the past) and fail to recognize the futurity that lies at the heart of markets and capitalist economies. This structural weakness is recognized by some of the modern theoreticians who have tried to generalize the Walrasian theory. To date, however, none of these efforts have been successful because contingent commodity markets are rare, rational expectations internalizing the correct model (Muth 1961) are not valid outside a stable equilibrium, and the 40

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Narratives and Socio-Economic Regimes

creation of futures and option markets entails speculation without ensuring dynamic stability. Luckily, the actors populating real-life economies have been continuously inventing practical devices to reduce uncertainty and complexity. During the post-Second World War Golden Age, indicative planning was used in order to mimic the impact of heterogeneous strategies and discover a minimalist basis for coherence at the macroeconomic level. The neoliberal project that followed aimed at delegating this coordination to financial markets. However, in practice this coordination took place not by means of rational and computational calculation but by the invention of simple and attractive narratives. These practical devices can be grouped under the concept of expectation regimes, defined as the mix of individual and collective coordination mechanisms that facilitate and channel decisions over multiple periods and thus have a long-lasting impact. Since the nineteenth century, quite different regimes have been invented, matured, and decayed. In the modern era when the efficacy of macroeconomic institutions is declining, the heterogeneity of interests is increasing, and intense structural change is obscuring the future, the time of narratives has come. Nevertheless, because such narratives rarely lead to the world they pretend to build, many fail. As this chapter demonstrates, the narratives of Japan number one, the new economy, and the omniscience of financial markets are all good examples of a relentless quest for attractive narrative representations of an unknown future. This analytical framework provides a suggestive interpretation of the higher frequency and severity of economic crises since financial liberalization: the timeframe of financiers’ representations is far too short compared with the time required to change products, technologies, lifestyles, education and training, public spending priorities, and the tax system. As a consequence, the sources (and the nature) of economic crises have changed dramatically. In socio-economic regimes dominated by a financial logic, storytelling is no longer restricted to the road shows of companies’ chief financial officers. Instead, it has permeated the mass media and transformed the very process of democratic political choices and government policies. ‘New era’ narratives and associated emotions and passions (Akerlof and Shiller 2009) sustain attractive but frequently unfounded stories that drive markets and even penetrate the political agenda. This amounts to the end of well-planned political programmes and ushers in an era of vibrant but unstable imaginary visions.

Uncertainty: The Core Ambiguities of Capitalism It might be tempting to restrict the role of narratives and fictions to finance-led contemporary capitalism (Boyer 2018). It is, however, much more insightful 41

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertain Futures

to revisit briefly theories that point out the uncertainty inherent in any monetary economy, before examining the particular prevalence of uncertainty (requiring resolution with the help of narratives) in globalized and finance-led capitalist systems.

The Three Uncertainties of Any Capitalist System Many theoretical and historical approaches have shown that the institution of money is a necessary condition for a market economy to exist, contrary to the neoclassical myth whereby it is the extension of barter that calls for the introduction of money. This raises the question of how a market economy reaches equilibrium via the price mechanism. Karl Marx seems to be the first to have pointed out that the issue is not only the determination of relative prices; rather, the existence of money allows actors to deal with the uncertainty stemming from the frequent temporal separation of selling one good and buying another (Marx 1867 [2017]). A further source of uncertainty relates to the time lag between investment in a productive process and the delivery of the expected returns: because competition between capitalists leads to overaccumulation, the succession of booms and busts is the consequence of an inability to anticipate the timing of the relevant stages of an economy during the course of an investment. American institutionalists provided particular insights regarding the close links between capitalism and the forging of images about the future. The concept of ‘futurity’ stresses this core property of any monetary economy (Commons 1934 [1989]). A third level of uncertainty is introduced once innovation becomes a key driver of competition: by definition, innovation undermines repetition and thus cannot be assessed by rational calculus. The challenge is the more daunting, the longer the time lag between the investment in research and development and the delivery of new goods or services to the market. This is the core message of Joseph Schumpeter (1926). Logically, the standard microeconomic approach is unable to deliver any reasonable solution, because no probability distribution can be derived from previous innovations. This central importance of innovation and novelty in modern economies (see also Shackle 1972) dictates a crucial role for contingent imaginaries and social representations of uncertain futures in contemporary capitalism.

Uncertainties Compounded in Finance-Led Capitalism While these general causes of uncertainty hold for capitalism in all its historical formations, there are reasons to believe that uncertainty is exacerbated in 42

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Narratives and Socio-Economic Regimes

contemporary capitalism and is leading to a growing role for storytelling (Salmon 2007; West and Micht 2000). The first reason is that the deepening of the division of labour creates strong structural interdependencies: any localized event might have surprising fallouts in distant sectors and territories. Generally, economic agents have neither the power nor the resources to diagnose and control this uncertainty. They have therefore to forge ad hoc representations of these interdependencies, taking into account only a few of the actually existing causal relations. The second reason is that globalization has developed global value chains that are difficult to master even for the leading firms that organize the segmentation of production. Rather than uncertainty being limited to the local environment of each production site, it is related to worldwide shocks that affect both demand in final markets and the supply of resources. Clearly, even the most talented experts with huge data sets have been unable to anticipate the successive crises of the new economy. Finally, the explosion of derivatives linked to each basic transaction has created an impressive complexity in financial instruments. Because the position of each trader is based in large part on private information in a context in which bilateral ad hoc contracts are the norm, outside observers are unable to detect emerging structural imbalances: they are revealed only when a node of the international financial system goes bankrupt. The collapse of Lehmann Brothers is a good example of this obscurity at the heart of the financial markets that standard economic theory had generally assumed were transparent. Financiers and traders are left to invent fictions in order to pretend they are masters of a process that ultimately they cannot comprehend.

Beyond Standard Economic Theory: How Institutions, Planning, Market Prices, and Narratives Guide Expectations Has contemporary economic theory found practical solutions to overcome or channel this uncertainty and make investment decisions easier and safer? Despite all the efforts in economics, only unsatisfactory theoretical solutions with little practical relevance have been found. This holds for general equilibrium theory as much as for rational expectations theory. Both these strands of standard economics are built on a series of thought experiments designed to deal with the future, but few of their insights can be applicable to the real world because that would require the complete redesign of economic institutions in accordance with the requirements of a normative model assuming total price flexibility, a complete set of contingent markets, and the ability of actors to solve complex analytical problems through an exclusive reliance on a full measure of substantial rationality. Since—notwithstanding the partial 43

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertain Futures

performativity of economic models in some areas (see MacKenzie 2006)—it is not possible to build the social world according to such idealized economic models, firms, individuals, and public authorities have had to invent partial and imperfect solutions to overcome the inhibiting role of uncertainty.

Economic Institutions: Devices for Guiding Expectations and Behaviours Despite the beliefs of standard economic theory, markets are not the only social constructions invented for coordinating heterogeneous behaviours. Conventions might emerge from the repetition of successful interactions (Lewis 2002) and organizations are built to socialize information and individual strategies (Aoki 1988). Both embed collective responses to recurring as well as unexpected events by partially redefining the social identity of members of society (Douglas 1986). At a higher level, some society-wide institutions help in structuring both the incentives and constraints that individuals face (North 1990), thereby reducing the scope of relevant behaviours. Similarly, regulation theory states that institutional forms monitor some key interdependencies among different domains—industrial relations, the competition regime, as well as the monetary and exchange rate regime (Boyer 2015; Boyer and Saillard 2001). In this way, institutions help construct the strategies necessary to cope with the deep complexity of societies (Delorme 2010). De facto, institutions address the various sources of uncertainty diagnosed by economic sociology (Beckert 2016, 43). Institutions frame the context for individual decisions; they constrain the range of alternatives considered; and they provide a simplification of the interactions with other domains. Of course, the rules of the game that institutions set do not fully determine the outcome but they delineate strategies and restrict radical uncertainty. They give structure to a problem with otherwise unlimited dimensions (Aoki 2010).

Indicative Planning: A Method for Socializing Individual Strategies While these general institutional forms may reduce ‘local’ uncertainty, their interaction can still display a lack of compatibility that generates chaotic dynamics that appear as uncertainty at the global or system level. Indicative planning may provide a method for diagnosing such incoherent regimes, promoting deliberation about possible solutions, and simulating likely consequences. If all the stakeholders are involved in the patient and recurring process of planning, then a shared vision of the future is diffused in everyday decisions about private and public investment, education and research, the geographical distribution of economic activity, and so on. In this way, the uncertainty of the future is partially but not totally reduced. This was the explicit philosophy of French public authorities in the heyday of the French Golden Age: the 44

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Narratives and Socio-Economic Regimes

statistical system, macroeconomic modelling, cost/benefit analysis of sectoral projects, and economic policy in terms of taxation or public spending were jointly mobilized to sustain the central trajectory decided after many iterations between technical expertise, the conciliation of dominant interests, and strategic choices made by the government. From a theoretical point of view, planning was presented as a tool to fight uncertainty (Massé 1965). It is not so easy to assess its role in the rapid and fairly steady growth in France in the thirty years after 1945 (Kindleberger 1967), but a similar configuration could be observed in Japan and had remarkable outcomes. In both countries the catching up process could be organized under the monitoring of the public authorities. But the very success of this socio-economic regime promoted a diversification of production away from basic goods, and the national economy was progressively opened up to international trade. These two structural transformations gradually eroded the efficiency of indicative planning, because firms tend to become the lead players in the stiffening of international competition, while public authorities lose some key economic policy instruments, such as control over the exchange rate or the taxation of mobile capital. When the trans-nationalization of value chains and financial asset portfolio management became the dominant mechanisms governing macroeconomic evolutions, French indicative planning experienced a severe crisis and the diffusion of ‘laissez-faire’ discourses convinced later governments that the state is the problem and markets are the solution.

The Creation of a Multiplicity of Futures Markets: The Neoliberal Solution When modern economies became too complex to be monitored by interventionist tools, the institutional arrangements of the post-Second World War regimes were replaced by competition mechanisms that operate through liberalized markets. This holds for the goods markets, labour contracts, and, more fundamentally, financial markets. If the best experts, both public and private, can no longer make accurate forecasts, market price signals are tasked with revealing the viability of public and private strategies. Under these conditions markets and states switch places. In the past, governments set medium-term expectations through programmes, while markets provided short-run adjustment processes. Since the end of 1990, international finance has been tasked with setting the stock market value of leading firms, assessing the relevance of their investments and innovation decisions, and checking the viability of national public finance. The future has become the consequence of private strategic choices, and governments are in charge of reacting to unexpected perturbations alongside the long-term trajectories set by leading private actors. 45

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertain Futures

Many futures markets are thus created and become the compass guiding economic actors. A large fraction of the economic profession followed the hypothesis that these markets are efficient in the sense that they deliver the best synthesis of scattered individual information. Nevertheless, the literature has shown that the volatility of stock market valuations is far higher than the variability of the underlying process of profit generation (Shiller 2000). It might be troubling for traders to observe that the ups and downs of financial markets follow a pure random walk. As a result, the related erratic movements tend to be attributed to the ‘mood of the market’, according to an astonishingly anthropomorphic vision of financial markets. Last but not least, instead of a smooth adjustment of stock markets to new information, brutal collapses of financial valuation continually surprise the best experts and financial gurus. The creation of futures markets, which was supposed to reduce uncertainty, appears in the end to have increased it, rendering decision-making by firms still more difficult than in the era of administered economies.

The Socialization of Expectations by Narratives This has opened up a new epoch for devices that actors invent to deal with futurity and uncertainty—in particular, for the spectacular narratives invented by financiers and entrepreneurs. In the words of the French poet Jean Cocteau (1927): ‘Let us pretend to be the authors of mysteries we are totally ignorant of ’. Since interdependencies between the domestic and the international, finance and the real economy, polity and economy display multiple channels that cannot be captured by a rational business plan and balance sheet, firms, investors, and traders need to simplify dramatically the mysterious world they must live in. It would be foolhardy to summarize an investment plan by a series of quantitative indices that are likely to be falsified by subsequent events. So instead the focus is on a suggestive discourse that offers a vision of the intentions of the large firm’s CEO, the startup manager, or the business angel of Silicon Valley. Faced with the multiplicity of transformations occurring in technological paradigms, lifestyles, geography of production, and social stratification, large multinational firms and investors must address the deep uncertainty affecting their decisions: how to invest in a changing world in which the old principles are invalid but new ones have not yet emerged? Business plans describing expected cash flows along a series of trajectories become mere fictions, because firms are unable to know the distribution of probability over the successive events that will affect the outcome of today’s decisions in terms of products, investment, and R&D expenditures. Inventing a simple and attractive narrative cuts through the inhibiting complexity and uncertainty. This solution suggests that entrepreneurs are mastering their future by describing it in a 46

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Narratives and Socio-Economic Regimes

narrative and allocating resources to make it happen. The initiative shifts from government and public agencies to charismatic entrepreneurs in the Schumpeterian sense of the term: they escape from the management of purely repetitive decisions sustaining the exhausted productive paradigm from the past and bet on the imagined success of new products, organizations, methods, and territories. This should normally deliver fairly chaotic macroeconomic developments, given the intrinsic heterogeneity generated by a market economy. Two features of contemporary societies, however, polarize the a priori unlimited pool of possible narratives. Firstly, the stock market and, more generally, the different financial instruments coordinate heterogeneous expectations via price formation: the rapid appreciation of a company’s stock is interpreted as a mark of the success of its visions and strategies, the more so when theoreticians and practitioners believe that the market valuation reveals all available information about the future. Frequently a dominant narrative emerges and becomes adopted by actors whose alternative bets have been disappointed. A typical fiction-led boom moves new industries and by extension the national economy. Secondly, the mass media, both old and new, widely diffuse not only financial valuations but also the stories of leading entrepreneurs at the forefront of the ‘new economy’. Thus a powerful mechanism reinforcing the role of dominant narratives is embedded in both the modern financial system and the media system. Such statements can be self-fulfilling if sufficient people invest money in the story. Instead of trying to decipher an obscure future, actors take decisions designed to make an imagined future real. ‘The new economy is the future’, ‘Capitalism has moved to Asia’, ‘Quantitative finance allows us to master risk and deliver an unprecedented rate of return’, or ‘The Euro is irreversible’ are all good examples of the contemporary narrative approach to uncertain futures; and these stories have to a considerable extent structured recent economic behaviour. The narratives told to the financial markets by government agencies also differ from the government policies prevailing at the time of indicative planning. The key public actor is no longer the ministry of finance, which is too slow and cumbersome to react to unexpected events, but rather the central banker: she or he has the task of interacting with financiers’ narratives by issuing statements designed to maintain trust in the positive evolution of financial markets (Greenspan 2013; see also the chapters by Braun and Holmes in this volume). This is a novelty in the long history of central banking (Blinder 1997). Governments are even happy to delegate difficult distribution choices to the so-called ‘markets’ in the context of a zero-sum game brought about by the quasi-stagnation of productivity (Krippner 2011). Furthermore, political authorities tend to take into account the interests of 47

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertain Futures

international financiers more than citizens’ demands (Streeck 2017). It is clear that communication has become one of the main policy instruments for governments and central bankers (Lordon 1997), while deciphering the message of markets has become one of the main policy inputs.

A Succession of Grand Narratives in Recent Economic History Because most imaginaries are unable to deliver the expected transformations in the economy, they generally end up with a brutal readjustment of expectations. In some cases, there is a long period of contestation between competing grand narratives and one eventually gains temporary dominance. In some other configurations, after a period of doubt and uncertainty, an alternative vision of the future emerges, becomes hegemonic, and leads to the recovery of investors’ ‘animal spirits’.

Back to Malthus: The Limits to Growth When the brutal spike in oil prices in the 1970s revealed the neglected dependency of the industrialist post-war model on cheap natural resources, economists expressed conflicting views. For most of them, the oil shocks of 1973 and 1979 represented temporary periods of turbulence for prosperous economies and were expected to be overcome quickly, provided that wise economic policies were followed: energy-saving innovations would respond to higher energy prices and allow production-based growth to resume. However, a dissenting group at the margins of the economic profession challenged this optimism: economic growth and demography had reached the limits set by the volume of natural resources available in the world (Meadows et al. 1972). An intriguing image was popularized by a simulation model, namely that of the water lily whose exponential growth will completely cover a pond in 30 days. Whatever the theoretical and technical controversies, the oil crises succeeded in reintroducing Malthus’ generic fears and his narrative of natural limits to growth. Public opinion supported, and governments initially adopted, some drastic measures in order to save energy. But as soon as a modest recovery had taken place, most economists resumed their defence of the guiding vision of Prometheus unbound: price and income variations, substitution, and technical change could overcome any natural resource scarcity.

Japan Invents the Machine that was Supposed to Change the World The debate shifted in the 1980s to analysis of the juxtaposed trajectories followed by the United States and Japan: industrial decline of typically Fordist 48

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Narratives and Socio-Economic Regimes

industries, on one hand, and the invention of a new production paradigm, on the other. The Asian ‘dragons’ and Japan in particular seemed to have discovered a way of avoiding the stagnation of productivity observed in North America. In practice, of course, the Asian emerging industrial economies displayed many differences with American and European capitalisms in terms of the interplay between state and market, social stratification, and integration in the world economy. Institutional theorists have pointed out the complementarity between the various institutional forms and different coherent development modes (Hollingsworth and Boyer 1997), which implies that successful institutional architectures cannot easily be imported (Amable 2003). Despite this, American and European managers harnessed the power of narrative to propose shortcuts that promised to resolve a complex web of causalities. The post-war ‘golden age’ was reinterpreted as a direct consequence of the breakthrough of mass production and the diffusion of the assembly line, inherited from the sophistication of ‘model T Ford’ production (Nye 2013). According to an imaginary of technological determinism that transcends different epochs, the future was now seen as being dependent on the adoption of so-called lean production. This inspiration was captured by the titles of two bestsellers: Japan Number One (Vogel 1999) and The Machine that Changed the World (Womack et al. 1990). This vision became quite effective in mobilizing managers and shaping macrodynamics. All over the world, business schools were teaching how to implement the so-called ‘Japanese model’, and many authors anticipated its domination at the global level—especially in terms of firms’ organization of production—as evidenced by the opening of Japanese plants in the United States and Europe. The belief in the superiority of the Japanese brand of capitalism created an unprecedented speculative stock market and real estate bubble, in an overreaction to what was perceived to be Japan’s bright future, with its possible replacement of the United States as the leading industrial power. This speculation had to be reassessed, however, because the simple narrative of lean production turned out to be an unjustified generalization, given the diversity of production models within the same sector and country (Boyer and Freyssenet 2002). Furthermore, Japanese macroeconomic performance had to be attributed to many other distinctive features, such as the wage–labour nexus, the nature of competition, and the financial system. Moreover, foreign admirers of Japanese capitalism totally missed the growing tensions and disequilibria generated by the very success of past strategies of firms and successive governments (Boyer and Yamada 2000). The lean production fiction has long been discarded, but it nevertheless achieved a partial transformation of the industrial world while it lasted. 49

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertain Futures

From Information to Knowledge: The New Economy After a period of uncertainty in search of promising new sectors, in the mid1990s American business became convinced that the convergence of various advances in information and communication technologies (ICT) would open up a new epoch in the history of production systems: information would become more important than energy and natural resources. This conception emerged in the United States, where public opinion tends to consider any innovation to be potentially beneficial—an attitude that stands at odds with the more cautious approach characteristic of Europe (World Value Survey 2016). Experts praised the unique innovative potential in North America, anchored in the excellence of top universities and research organizations. Policy-makers were eager to regain technical and economic hegemony over Japan. While the old industrial basis was shrinking, totally new enterprises quickly established dominant and sometimes quasi-monopolistic positions in the production of information products and equipment goods. A new and powerful narrative emerged: information and communication technologies were believed to abolish the barriers of time and space. As communication and information technologies were diffused, it became clear that profits can only be earned in the longer term if the bulk of information is converted into knowledge. This led to the rise of the second generation of internet companies, specializing in software (Boyer 2004). The narrative of the new economy permeated society as a whole: intangible capital frequently eclipsed typical equipment goods; start-ups dictated the speed of the economy; and when they were converted into public firms quoted on the stock market, their capitalization exploded and often superseded those of the old economy. Companies that never earned any profit enjoyed high stock prices because they explored the future and were destined (so the narrative went) to encounter success ‘eventually’. Silicon Valley and the newly created Nasdaq were allies that jointly bet on the bright future of the New Economy: they channelled a flow of investments into new companies out of typical goods production. The enthusiasm was so frantic that in the 1990s some experts even speculated that ‘economic laws’ had now been invalidated. Very early on, dissenters challenged the anticipated impact of ICT on productivity, but they were not listened to (Gordon 2000). In due course, they were proved right, however, and the new economy did not succeed in building a stable regime. Instead, the dot-com bubble burst in the early 2000s. From this a general lesson can be drawn: the grand narratives that guide behaviour may change, but they all end up in a financial crisis (Garber 2000; Kindleberger 1978; Reinhart and Rogoff 2009). In the end, the guiding fictions have to cope with the real components of accumulation: although they may succeed in transforming the economy, they tend not to do so quickly or 50

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Narratives and Socio-Economic Regimes

deeply enough to triumph as self-fulfilling prophecies. The conflict of time horizons between financial imaginaries and the actual economic activity they impel is the deeply rooted source of major financial crises (Boyer 2013). Furthermore, every narrative has its blind spots; and, if the same narrative is widely shared, so too is the cognitive myopia implied (Bronk 2013; and Beckert and Bronk in this volume).

The Hegemony of Finance: The Heyday of Imaginaries and Narratives The dot-com crisis did not trigger a depression because the central banks reacted quickly, lowering interest rates. This created a permissive condition for a recovery of the economy. But where to invest, given the reappraisal of the prospects opened by ICT and the New Economy? The solution was found within the financial system: the advances in mathematical finance opened up a new territory promising the reconciliation of high rates of return on capital with a reduction of risk. As Haldane (2009) puts it: ‘A new era had dawned, one with simultaneously higher returns and lower risk. . . . Or so ran the rhetoric.’ The breakthroughs of academic research delivered methods for pricing new financial products; first options, then derivatives, and finally derivatives of derivatives (Black and Scholes 1973; Merton 1973; see also Esposito in this volume). The volume of contracts and transactions exploded and they came to be a source of profits for banks and financial intermediaries alike. Academic research had set in motion a transformation of contemporary financial markets (MacKenzie 2006). Paradoxically, some conceptual continuity prevailed with the previous periods: scientific progress was invoked to justify the hegemony of finance that developed. The specialists in statistical physics or engineering shifted from industry to finance and became so-called ‘quants’: old-style financiers could not assess the nature and risks of these new businesses and were unable to challenge the legitimacy of quants because they did not share the prestige of such high calibre scientists (see also Besedovsky in this volume). Dissenters who stressed that the whole financial industry was built upon erroneous calculations of probability distributions that exclude the ‘fat tails’ associated with crises were not listened to (Mandelbrot and Hudson 2005; Taleb 2010). A new grand narrative was born. It stated that any economic or social issue can be overcome by an ad hoc financial instrument, whose introduction is a better strategy than implementing painful reforms and unpopular policies (Shiller 2003, 2008). Are developing countries suffering from recurring exchange rate adjustments and financial crises? Let us design a sophisticated insurance contract that obviates the need to change either the currency regime or economic policy! Is a rock star suffering a liquidity constraint but rich in terms of future royalties? Let us convert the flow of income into a 51

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertain Futures

security that can be sold to an investor and cashed-in immediately. Do minority and low-income families face persistent difficulties purchasing housing? Why not reduce personal income and asset requirements for mortgages, and securitize these credits to ‘spread’ the risk? If this opportunity is widely exploited by banks, a housing boom generates price increases that sustain the illusion that gains from speculation can replace income. Unfortunately from a macroeconomic point of view, this is impossible in the long run, and the strategy was bound to end in a dramatic collapse of a financial system that had fully embraced the narrative that systemic risk was now under control (Boyer 2008, 2011a). The subsequent crisis was far more severe than most previous ones. This was because the accumulation regime was totally finance-led, with few breakthroughs in the real economy that might lead to productivity increases (Gordon 2016). Moreover, the deep integration of various financial systems reinforced the synchronicity generated by the worldwide diffusion of the same basic narrative about the efficiency of financial markets (see also Bronk and Jacoby 2016). But this belief was not in line with rigorous research that had proven that financial markets could not be efficient in terms of information (Grossman and Stiglitz 1980). The typical pattern of financial crises confirms this assessment: the reversal from speculative boom to brutal and deep collapse is always spectacular. After the collapse of Lehman Brothers in the autumn of 2008, the panic was interrupted only by the central banks— which played their role of lender and buyer of last resort to the hilt (Eichengreen 2015)—and by the resilience of the rare accumulation regimes still based on productive capital, especially that of China (Boyer 2011b). In the end, the economic policies introduced via a succession of imagined futures derived from finance theory have led to collapse or have tended to become less and less effective, leaving most modern economies that were built around these imaginaries facing the prospect of long-term stagnation (Summers 2016).

The Green Economy: The Most Recent Narrative to Move Financial Markets Once the risk of a cumulative depression and deflation was removed in the 2010s, governments and financiers faced a daunting question: what will be the next engine of growth? The very active Chinese economic policy achieved a return to rapid growth fuelled by a credit boom that reverberated all over the world. The strong dependency of Chinese manufacturing on imported natural resources triggered a commodity boom and a spectacular reversal of the terms of trade between manufactured goods and natural resources. This was, for a time, interpreted as a confirmation of the possible blocking of growth by the exhaustion of oil, metals, and even food in some regions of the world. 52

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Narratives and Socio-Economic Regimes

But this was not a mere repetition of the oil shocks of the 1970s (Meadows et al. 1972, 2004). On one hand, the debate on climate change reinserts world economic activity into the biosphere and the physical barriers to continuing the past modes of development: economists and policy-makers have to take this structural change into account (Stern 2006). On the other hand, the market for oil futures fed intense speculation because the expectation of future scarcity drives the spot market: primary commodity prices were quoted as financial assets (Cournot Center 2016). This invasion of the domain of natural resources by finance was the explicit strategy of some investors and it was designed immediately after Lehman Brothers’ collapse (MacCall 2009). A new narrative emerged around the green economy and it permeated the redeployment of investment, innovation, environment regulations, taxation, and public infrastructure. One variant of this project proposed that sustainability and prosperity should replace growth as governments’ key concern (Jackson 2009). As in previous cases, this new narrative triggered a modest economic boom, but it failed to allow the effective implementation of the new growth regime based on long-term ecological sustainability, as anticipated by governments and reiterated by the Paris Conference on Climate Change. The transformation of the real economy was far slower than the unfolding of the speculative bubble in oil and natural resources. Furthermore, the narrative of the greening of national economies is a drastic simplification, given the complexity of ongoing and interrelated structural transformations: the maturing of the e-economy, the development of new services, the constant rise of health-care demand, the delocalization of polluting industries, an uncertain exit policy from quantitative easing, and recurring protectionist temptations. The conflict of time horizons between finance and the economy and the extreme simplification associated with the concept of the Green Economy explain the brutal reversal of oil prices seen since 2014, its repercussion on stock markets, and the re-emergence of major uncertainties concerning the direction of investment and the future regime for oil prices (Garnier and Sølna 2017).

The Destiny of Successful Narratives: A Source of Major Financial Crises This brief historical perspective on developments since the 1970s— summarized in Table 2.1—suggests the existence of some crucial and invariant mechanisms that explain both the reasons for narratives and their power to determine the dynamics of capitalist economies. It seems to confirm the importance of representations of the future as stressed by recent theoretical breakthroughs: widely shared imaginaries and associated grand narratives are 53

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertain Futures Table 2.1. A synoptic view of the chronology of imaginaries and expectation regimes in the contemporary period Socioeconomic regime

Imaginary

Expectation regime

Sources of crisis

Limits to growth (1970s) Japan no. 1 (1980s)

A neo-Malthusian conception

Oversimplification of links between economy and natural resources Misreading of the source of Japanese growth

The New Economy (1990s)

A Schumpeterian technological revolution

Transition towards a stationary economy Transition towards another socioeconomic regime Technological impatience

Finance-led accumulation (2000s) Environmental limits to prosperity (2010s)

Successful control of risk by scientific advances in mathematical finance A green economy

Decline of American hegemony, Japanization of the world

Driven by storytelling Transition to a new socio-economic regime

Conflict of time horizons between technical imaginaries and the real economy Conflict of time horizons between financial models and the real economy Global challenge versus national interests

one of the (imperfect) responses to the uncertainty of modern economies (Beckert 2013, 2016; Bronk and Jacoby 2016).

The Higher the Uncertainty and Complexity, the More Urgent the Need for Simple Narratives All individuals or firms have to face the uncertainty associated with a monetary economy and develop relevant strategies, knowing that economic calculus deals only with measurable (or ‘Knightean’) risk and becomes unfeasible in complex economies. In conditions of uncertainty, rules of thumb (Heiner 1983), reliance on bounded rationality (Simon 1997), learning from past episodes (Nelson and Winter 1982), or the influence of ‘animal spirits’ (Keynes 1936) are all mobilized in order to enlighten decisions with intertemporal consequences. These devices help to overcome specific and localized uncertainties, and they imply fairly diverse behaviours (including the bankruptcy of some actors) that are made compatible only ex post by the price mechanism. More difficult to master is a second level of uncertainty created by the complexity of the long-distance interactions between sectors, and by the macroeconomic uncertainties generated by innovations within the financial system and government policy. As argued earlier in this chapter, economic institutions aim precisely at reducing this second-degree uncertainty. But the exhaustion and open crisis of the 54

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Narratives and Socio-Economic Regimes

post-war socio-economic regime eroded belief in institutional constructivism. Given the lack of convincing alternatives, most governments have, as a result, delegated the exploration of the future to financial markets. Such markets do allow for the pooling of scattered visions and the coordination of decentralized information through the price mechanism—one definition of financial market efficiency—but this does not warrant the extreme faith in market knowledge evident before the crisis. Individuals and firms recognize that they must rely upon the price signals of financial markets because they have only limited knowledge. Since these markets appear to be highly stochastic, experts have frequently gone on to attribute a clear intentionality and rationality to the ups and downs of financial markets. A curious personification of the markets occurs, with statements such as ‘the markets think that . . . ’, whereas in reality the ‘message of markets’ is merely the unintended consequence of the confrontation of initially heterogeneous expectations, objectives, and financial resources. This introduces two entry points for narratives. As this chapter has shown, each era displays a grand vision (or narrative) about the direction of change and the source of profit. But in order to interpret the everyday market process, secondary narratives have to explain the observed discrepancies with respect to the imagined ideal future. To be convincing, both primary and secondary narratives have to be fairly simple.

Mimetism Is Rational and Leads to the Hegemony of a Single Narrative In the post-war period, social and political deliberations—via collective bargaining, economic policy debates, and indicative planning—partially succeeded in synchronizing, ex ante, behaviours across different social groups with different, if not opposing economic interests. This has not been the case since the 1980s. What, then, are the processes that convert the heterogeneity of representations of the future into a common macroeconomic dynamic? This chapter argues that the logic of financial markets entails a definite pattern—alternating bull and bear assessments of a succession of grand narratives. Conventional economic theory sticks to the hypothesis that in equilibrium rational expectations will prevail because agents gain economic rewards by anticipating the future correctly, while competition eliminates all those who fail to build correct forecasts. But this neglects the fact that the learning process converges only if the final rational equilibrium is stable (Grandmont and Laroque 1991). Since socio-economic regimes mature and decay repeatedly, agents are unable to deploy rational expectations during periods of dynamic change between alternative emerging regimes characterized by radical uncertainty. 55

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertain Futures

A second standard economic solution considers the sequence of price signals on product and factor markets as sufficient for the monitoring of inter-temporal and long-term decisions about production, investment, and innovation. Again, this assumes an ergodic and ultimately predictable system tending towards a stable and long-term equilibrium—a contradiction in terms because technical, organizational, and institutional innovations progressively transform the existing economic regime (Shackle 1972; and Beckert and Bronk in the introduction to this volume). The overwhelming role of finance in contemporary capitalist regimes opens up a more relevant solution. The futures markets guide decentralized individual strategies by proposing not only prices but also shared representations of the future. When uncertainty increases, agents tend to rely more on the expectations of others than on their own, and this rational mimetic behaviour may move the price of assets away from so-called fundamental value, computed as the discounted value of future incomes (Orléan 1990). Indeed, two extreme narratives may then alternate: one a bet on the complete success, the other on the total failure, of a given financial investment. In a sense, finance has replaced collective deliberation. This comes at a cost: the reliance on widely shared and frequently misleading narratives implies an intrinsic financial and economic instability. This calls for a new theory of value that is different from past analyses that used to look for objective foundations for prices and values (Orléan 2014).

Conclusion The analysis presented in this chapter is synthesized in Figure 2.1. Three basic features of monetary and capitalist economies render powerless any economic calculus assuming rational expectations. First, at the individual level, investment and innovation face radical uncertainty because not all possible states of the world can be known. A second category of uncertainty relates to the reaction of other agents that belong to the same sector and sphere of competition. Finally, the relevant web of interactions is difficult to decipher, given the deepening of divisions of labour within and across national borders. This central problem of uncertainty for modern economies has been dealt with since the immediate post-war period by applying two different strategic approaches. During the epoch of constructivism that followed the Second World War, core socio-political compromises and indicative planning led to the emergence of fairly coherent socio-economic regimes, and a remarkable reduction of the three main sources of uncertainty was achieved. With the subsequent deregulation of product and labour markets, the dynamism of financial innovation, and the move towards globalization, financial markets 56

The need for alternatives

A misallocation of capital

Story-telling as tentative self-fulfilling prophecies

2. Radical uncertainty of outcomes of individual decisions

Institutions and organizations as uncertainty reduction devices

An impossible rational calculus

An open major crisis

Decisions on production, investment, innovation

An economic boom

Emerging disequilibrium

A cyclical downturn 3. Dependency with respect to other agents’ decisions

Futures markets allow agents to follow price signals guided by shared narratives

No early detection of disequilibria

From a set of anti-uncertainty devices to economic dynamics The feedback from economic crisis to the invention/implementation of anti-uncertainty devices

Figure 2.1 An interpretation of economic evolution: the emerging, maturing, and crisis of successive narratives

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

1. Complexity of interactions

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertain Futures

were charged with designing an alternative regime. Ever since, the price signals on financial markets have been progressively complemented by grand narratives that are supposed to synchronize the strategies of heterogeneous actors. Narratives are now the key instruments available for top managers and governments. But the downside of this strategy has been made painfully clear by the financial crisis of 2007 and the instability that followed. Behind the major differences between different eras, a common dynamic pattern is operating. The invention of narratives as an ‘anti-uncertainty device’ makes possible a wave of investments that is initially successful but finally hits the barrier associated with an unbalanced accumulation regime, a disjunction between the time frames of imaginaries and actual economic outcomes, and a failure of narratives to capture key dynamics. In particular, a ‘structural crisis’ tends to take place when a common belief about the direction of investment and innovation breaks down. Radical uncertainty then tends to return amid calls for the invention of alternative narratives and economic institutions. This chapter has described the existence of a two-way causality between a succession of key narratives and various configurations of capitalism. At the most basic level, narratives and socio-economic regimes co-evolve alongside major social transformations. Across the unfolding of different historical eras, there are significant regularities in the pattern of emergence, maturing, and crisis of socio-economic regimes, in which narratives and expectation regimes play a crucial role.

Bibliography Akerlof, George A., and Robert J. Shiller. 2009. Animal Spirits: How Human Psychology Drives the Economy and Why It Matters. Princeton NJ: Princeton University Press. Amable, Bruno. 2003. The Diversity of Modern Capitalism. Oxford: Oxford University Press. Aoki, Masahiko. 1988. Information, Incentives and Bargaining in the Japanese Economy. Cambridge: Cambridge University Press. Aoki, Masahiko. 2010. Corporations in Evolving Diversity: Cognition, Governance, and Institutions. Oxford: Oxford University Press. Beckert, Jens. 2013. ‘Capitalism as a System of Expectations: Toward a Sociological Micro-Foundation of Political Economy’. Politics and Society 41 (3): pp. 323–50. Beckert, Jens. 2016. Imagined Futures: Fictional Expectations and Capitalist Dynamics. Cambridge, MA: Harvard University Press. Black, Fischer, and Myron Scholes. 1973. ‘The Pricing of Options and Corporate Liabilities’. Journal of Political Economy 81 (3): pp. 637–54. Blanchard, Olivier J. 2008. ‘The State of Macro’. NBER Working Paper no 14259, August. http://www.nber.org/papers/w14259.

58

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Narratives and Socio-Economic Regimes Blinder, Alan S. 1997. Central Banking in Theory and Practice, Cambridge, MA: MIT Press. Boyer, Robert. 2004. The Future of Economic Growth: As New Becomes Old. Cheltenham, UK: Edward Elgar. Boyer, Robert. 2008. ‘History Repeating for Economists: An Anticipated Financial Crisis’. Prisme, no 13, November. Paris: Centre Cournot pour la Recherche en Economie. http://robertboyer.org/fr/category/publi/livre/page/2/. Boyer, Robert. 2011a. Les financiers détruiront-ils le capitalisme? July. Paris: Economica. Boyer, Robert. 2011b. ‘A New Epoch but still Diversity within and between Capitalism: China in Comparative Perspective’. In Capitalist Diversity and Diversity within Capitalism, edited by Christel Lane and Geoffrey Wood, pp. 32–68. Abingdon: Routledge. Boyer, Robert. 2013. ‘Les crises financières comme conflit de temporalités’. Vingtième Siècle. Revue d’Histoire, no 117, pp. 69–88. https://www.cairn.info/revue-vingtiemesiecle-revue-d-histoire-2013-1-page-69.htm. Boyer, Robert. 2015. Economie politique des capitalismes: Théorie de la régulation et des crises. Paris: La Découverte. Boyer, Robert. 2018. ‘From Economics as Fiction to Fiction-Led Capitalism’. In Representations of Capitalism in American Fiction, edited by Jacques-Henri Coste and Vincent Dussol. London: Palgrave Macmillan. Boyer, Robert, and Michel Freyssenet. 2002. The Productive Models: The Condition of Profitability. London: Palgrave Macmillan. Boyer, Robert, and Yves Saillard. 2001. Eds. Regulation Theory: The State of the Art. London: Routledge. Boyer, Robert, and Toshio Yamada. 2000. Japanese Capitalism in Crisis: A Regulationist Interpretation. London: Routledge. Bronk, Richard. 2013. ‘Reflexivity Unpacked: Performativity, Uncertainty, and Analytical Monocultures’. Journal of Economic Methodology 20 (4): 343–9. Bronk, Richard, and Wade Jacoby. 2016. ‘Uncertainty and the Dangers of Monocultures in Regulation, Analysis, and Practice’. MPIfG Discussion Paper 16/6. http://www. mpifg.de/pu/mpifg_dp/dp16-6.pdf. Cocteau, Jean. 1927. Les mariés de la tour Eiffel, collection folio, réédition Paris: Gallimard. Commons, John R. 1934 [1989]. Institutional Economics: Its Place in Political Economy. London: Routledge. Cournot Center. 2016. ‘Régime de formation des prix et volatilité: Le pétrole face au charbon et au gaz en longue période’. April. Paris: Centre Cournot. Mimeograph. Delorme, Robert. 2010. Deep Complexity and the Social Sciences: Experience, Modelling and Operationality. Cheltenham: Edward Elgar. Douglas, Mary. 1986. How Institutions Think. New York: Syracuse University Press. Eichengreen, Barry. 2015. Hall of Mirrors: The Great Depression, The Great Recession, and the Uses—and Misuses—of History. New York: Oxford University Press. Garber, Peter M. 2000. Famous First Bubbles: The Fundamentals of Early Mania. Cambridge, MA: MIT Press. Garnier, Josselin, and Knut Sølna. 2017. ‘Time-Frequency Analysis of Oil Price Data’. Prisme, no 33, Paris: Centre Cournot. Gordon, Robert J. 2000. ‘Does the “New Economy” Measure up to the Great Inventions of the Past?’. Journal of Economic Perspectives 14 (4): pp. 49–74.

59

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertain Futures Gordon, Robert J. 2016. The Rise and Fall of American Growth: The U.S. Standard of Living since the Civil War. Princeton, NJ: Princeton University Press. Grandmont, Jean-Michel, and Guy Laroque. 1991. ‘Economic Dynamics with Learning: Some Instability Examples’. In Equilibrium Theory and Applications. Proceedings of the Sixth International Symposium in Economic Theory and Econometrics, edited by William A. Barnett, Bernard Cornet, Claude D’Aspremont, Jean Gabszewicz, and Andreu Mas-Colell, pp. 247–73. Cambridge: Cambridge University Press. Greenspan, Alan. 2013. The Map and the Territory: Risk, Human Nature, and the Future of Forecasting. London: Penguin Press. Grossman, Sanford J., and Joseph E. Stiglitz. 1980. ‘On the Impossibility of Informationally Efficient Markets’. American Economic Review 70: pp. 393–408. Haldane, Andrew. 2009. ‘Why Banks Failed the Stress Test’. Speech given at the MarcusEvans Conference on Stress-Testing, 13 February 2009. http://www.bankofengland. co.uk/archive/documents/historicpubs/speeches/2009/speech374.pdf. Heiner, Ronald A. 1983. ‘The Origin of Predictable Behavior’. American Economic Review 73 (4): pp. 560–95. Hollingsworth, J. Rogers, and Robert Boyer. 1997. Eds. Contemporary Capitalism: The Embeddedness of Institutions. Cambridge: Cambridge University Press. Jackson, Tim. 2009. Prosperity Without Growth: Economics for a Finite Planet. London: Earthscan. Keynes, John Maynard. 1936. The General Theory of Employment, Interest and Money. London: Macmillan. Kindleberger, Charles P. 1967. ‘French Planning’. In National Economic Planning, edited by Max F. Millikan, pp. 279–303. Cambridge, MA: NBER. http://www.nber.org/chap ters/c1426. Kindleberger, Charles P. 1978. Manias, Panics and Crashes: A History of Financial Crises. New York: Basic Books. Krippner, Greta R. 2011. Capitalizing on Crisis: The Political Origins of the Rise of Finance. Cambridge, MA: Harvard University Press. Lewis, David K. 2002. Convention: A Philosophical Study. Oxford: Wiley Blackwell. Lordon, Frédéric. 1997. Les quadratures de la politique économique. Paris: Albin Michel. McCall, Matthew. 2009. The Next Great Bull Market. Hoboken, NJ: Wiley. MacKenzie, Donald. 2006. An Engine, Not a Camera: How Financial Models Shape Markets. Cambridge, MA: MIT Press. Mandelbrot, Benoît, and Richard L. Hudson. 2005. Une approche fractale des marchés: Risquer, perdre et gagner. Paris: Odile Jacob. Marx, Karl. 1867 [2017]. Capital: A Critique of Political Economy, Vol. 1. London: Forgotten Books. Massé, Pierre. 1965. Le Plan ou l’anti-hasard. Paris: Gallimard. Meadows, Donella H., Dennis L. Meadows, Jørgen Randers, and William W. Behrens. 1972. The Limits to Growth. New York: Universe Books. Meadows, Donella H., Jorgen Randers, and Dennis L. Meadows. 2004. The Limits to Growth: The 30-Year Update. London: Earthscan. Merton, Robert C. 1973. ‘Theory of Rational Option Pricing’. Bell Journal of Economics and Management Science 4 (1): pp. 141–83.

60

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Narratives and Socio-Economic Regimes Muth, John F. 1961. ‘Rational Expectations and the Theory of Price Movements’. Econometrica 29 (3): pp. 315–35. Nelson, Richard R., and Sidney G. Winter. 1982. An Evolutionary Theory of Economic Change. Cambridge, MA: Harvard University Press. Nye, David E. 2013. America’s Assembly Line. Cambridge, MA: MIT Press. North, Douglass C. 1990. Institutions, Institutional Change and Economic Performance. Cambridge: Cambridge University Press. Orléan, André. 1990. ‘Le rôle des influences interpersonnelles dans la détermination des cours boursiers’. Revue économique 41, pp. 839–68. Orléan, André. 2014. The Empire of Value: A New Foundation for Economics. Cambridge, MA: MIT Press. Reinhart, Carmen M., and Kenneth S. Rogoff. 2009. This Time Is Different: Eight Centuries of Financial Folly. Princeton, NJ: Princeton University Press. Salmon, Christian. 2007. Storytelling: La machine à fabriquer des histoires et à formater des esprits. Paris: La Découverte. Sapir, Jacques. 2000. Les Trous noirs de la science économique: Essai sur l’impossibilité de penser le temps et l’argent. Paris: Albin Michel. Schumpeter, Joseph Alois. 1926. Théorie de l’évolution économique (Theorie der wirtschaftlichen Entwicklung). Translated by Jean Jacques Anstett. 2nd edn. Dalloz: Paris. Shackle, George L. S. 1972. Epistemics and Economics: A Critique of Economic Doctrines. Cambridge: Cambridge University Press. Shiller, Robert J. 2000. Irrational Exuberance. Princeton, NJ: Princeton University Press. Shiller, Robert J. 2003. The New Financial Order: Risk in the 21st Century. Princeton, NJ: Princeton University Press. Shiller, Robert J. 2008. The Subprime Solution: How Today’s Global Financial Crisis Happened and What to Do about It. Princeton, NJ: Princeton University Press. Simon, H. 1997. Models of Bounded Rationality: Empirically Grounded Economic Reason. Cambridge, MA: MIT Press. Streek, Wolfgang. 2017. Buying Time: The Delayed Crisis of Democratic Capitalism. London: Verso. Summers, Lawrence H. 2016. ‘The Age of Secular Stagnation: What It Is and What to Do About It’. Foreign Affairs 95 (2): March/April. https://www.foreignaffairs.com/ articles/united-states/2016-02-15/age-secular-stagnation. Stern, Nicholas. 2006. Stern Review on the Economics of Climate Change. London: HM Treasury. Taleb, Nassim Nicholas. 2010. Le cygne noir: La puissance de l’imprévisible (The Black Swan: The Impact of the Highly Improbable). Paris: Les Belles Lettres. Vogel, Ezra F. 1999. Japan Number One: Lessons for America. Cambridge, MA: Harvard University Press. West, Scott, and Mitch Anthony. 2000. Storytelling for Financial Advisors: How Top Producers Sell. Chicago, IL: Dearborn Financial Publishing. Womack, James P., Daniel T. Jones, and Daniel Roos. 1990. The Machine that Changed the World. New York: Macmillan. World Value Survey. 2016. ‘Survey Wave 2010–2014’. http://www.worldvaluessurvey. org/WVSDocumentationWV6.jsp.

61

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

3 Conviction Narrative Theory and Understanding Decision-Making in Economics and Finance David Tuckett

‘Standard macroeconomics’ recognizes the crucial importance of expectations in its models but at the same time has nothing to say about their content or the source of their formation or alteration. It also does not consider how they function in conditions of uncertainty. This chapter looks at the nature of expectations through the lens of a new socio-psychological theory of decisionmaking called ‘Conviction Narrative Theory’.1 Conviction Narrative Theory is a multidisciplinary approach developed to understand how human social actors make decisions to act in conditions of radical uncertainty. Founded on psychoanalytic theories of thinking and unconscious phantasy, it combines established approaches in psychoanalysis and sociology with more recent ideas in psychology and cognitive and affective neuroscience. Radical uncertainty is here taken to refer to situations in which the outcomes of planned actions cannot be known, to any measurable extent, at the time the decision is taken and for a long time afterwards, if ever. The reason is that such decision-making contexts are both equivocal and indeterminate— meaning, in formal terms, that to define and sample states and events relevant to the decision is difficult, and also that it is not possible to calculate either the probabilities of events or the relative probability of each state occurring (Lehner 2002). 1 Parts of this chapter—especially the section headed ‘Conviction Narrative Theory’ and Figures 3.1 and 3.2—are derived from Tuckett and Nikolic (2017). These elements are used as building blocks for a new application of the theory to operationalize Keynes’ concept of ‘animal spirits’ and to explain market instability.

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Conviction Narrative Theory

Although notably absent from most current work in decision science, understanding decision-making in radical uncertainty is important.2 Many decisions to commit to action in the complex, dynamic, and interconnected world in which we live have outcomes that are radically uncertain, whether in economics, finance, politics, government, or commercial organizations. Innovation is one obvious driver of such radical uncertainty, but even in an apparently stable environment, unforeseen consequences are regularly observed from new complex constellations of interdependent events. Arguably, a failure to incorporate radical uncertainty and how actors cope with it into economic and finance models (as, indeed, into most approaches in decision science) is one factor responsible for the unthinking reliance on formal modelling prior to the recent economic and financial crisis (Gigerenzer 2014; Kay 2015; King 2016). It remains a major limitation of current economic and finance theory (Tuckett et al. 2015). Drawing on arguments in Chong and Tuckett (2015), this chapter begins by describing the radically uncertain context faced by money managers and how they cope by developing conviction narratives. It then generalizes that idea to introduce a wider theory of decision-making under radical uncertainty, termed ‘Conviction Narrative Theory’ (henceforth, CNT). CNT (Tuckett and Nikolic 2017) differs from standard approaches to decision-making, which limit themselves to theories of information processing in contexts where data is available to calculate future probabilities. In contrast, CNT draws on the human capacity to organize experience through narrative, which is able to combine cognitive and affective responses generated by the subjective experience of action in conditions of radical uncertainty. Although novel in psychology, economics, and sociology, this chapter describes how CNT is supported by a significant body of current theory and research. Finally, it argues that CNT is a way to operationalize Keynes’ (1936) formulation of ‘animal spirits’ as a human solution to radical uncertainty. As such, it provides alternative, more plausible, and empirically substantiated microfoundations on which to build a macroeconomic understanding of the development of monocultures and financial market instability, as well as tools to understand and forecast an economy’s evolution more generally.

2 Related terms for radical uncertainty are ‘ontological’ (Lane and Maxwell 2005); ‘deep’ (Petersen 2006), ‘Knightian’ (Knight 1921), or ‘model’ (Chatfield 1995) uncertainty. These are all forms of uncertainty in which the system model generating outcomes and the input parameters to the system model are not known or widely agreed on by the stakeholders in a decision (Lempert 2002). In psychology, there has been some work on contexts characterized by ‘ambiguity’ or ‘ignorance’, but almost all work on judgement and decision-making retains the well-defined problem structure of gambles (for example, Fox and Tversky 1995) and so cannot address the decision context that is the concern of this chapter.

63

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

Money Management as Case Study Money management necessarily creates an emotionally subjective experience. First, trading financial assets, like undertaking fixed investment in capital goods or infrastructure, creates an experience of what can be considered dependent object relations (Tuckett 2011; Tuckett and Taffler 2008): the activity makes the creditor who holds the asset experientially dependent on the uncertain future of the debtor over time. This is true whether the debtor is a sovereign state, corporation, or individual and whether the asset is a cash deposit, bond, equity, or derivative. The outcome may be gain or loss. Such dependency experience necessarily creates an emotional state, as in human relationships more generally. Second, all financial assets are abstract entities— debt contracts whose worth depends not on their intrinsic properties alone but also on other market actors’ reflexive evaluations (Soros 1987) of the future income-generating capacity of the organization issuing the contracts. To take on dependency that might involve loss, as in taking on a partner, financial agents must find ways to be convinced about the profitability of the uncertain opportunities for future gain they hypothesize to exist. Money ‘managers’ are finance professionals who exercise highly remunerated legal mandates to buy and sell financial assets. They build portfolios by holding a basket of assets with the intention that the overall worth of the portfolios they manage will grow. Because holding assets of uncertain worth over time is a different experience from very rapid buying and selling they are usefully differentiated from financial ‘traders’, who usually hold positions for no more than a few hours or indeed minutes. To fill their portfolios, money managers must search for profitable opportunities. They look for assets whose relative value they suppose will grow in the medium to long term; and, to do that, they make two linked judgements: they try (a) to find entities whose relative value they think is underestimated by their market price today, and (b) to anticipate what value others will place on these entities’ worth in future (the Keynesian ‘beauty context’; Keynes 1936). Calculation, imagination, and analysis underlie their search, but because the future of the entities (including their relations to other competing entities) and the future judgements of other agents are unknowable, ex ante, their decisions also require coping with the emotional experience of radical uncertainty. In 2007 and again in 2011, Tuckett (2011, 2012) and Chong and Tuckett (2015) interviewed fifty-two highly experienced and high-reputation money managers in London, New York, Boston, Edinburgh, Paris, and other financial centres and asked them to describe in detail some decisions with which they had been pleased in the previous twelve months and some with which they had not. Tristan Cooper (not his real name) is a typical example of a money manager from those interviews. He made portfolio decisions by 64

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Conviction Narrative Theory

selecting individual stocks. At the time of interview, he was deciding whether to include in his portfolio a large construction company valued at several billion. He said it was doing ‘interesting things’ that would remain highly profitable. The ‘opportunity’ had been flagged to him by a ‘quant’ screening system because the share price had fallen after one of the company’s subsidiaries had disclosed accounting irregularities, causing distrust and some millions in losses. Cooper had to decide: was the asset cheap due to a past difficulty or because there will be ongoing problems? If the former, it is cheap and will recover, providing a gain. After a lot of data analysis, Cooper went to visit the company’s Chief Financial Officer (CFO) ‘to try and get a sense of what kind of guy he is’. At the meeting, he concluded that the CFO came ‘up with a very decent explanation as to why they had screwed up’, providing a number of reasons. Based on that explanation, he argued to himself: From a valuation standpoint . . . if you are a construction company worth 2 billion euros then, if you have to write off 60 million, once, it shouldn’t matter a lot . . . Earnings for the year are going to be destroyed but once the stock has fallen you can forget about it.

Cooper bought the stock, but six months later the company revealed a second accounting irregularity, making Cooper upset. Investment portfolio decisions must generally be maintained over time or transaction costs will extinguish gains. ‘Nothing had changed really. I should have said, fine, that is another just 50 million . . . from a strict mathematical standpoint . . . it doesn’t matter.’ In fact, at first, he had ‘hung on . . . probably because I trusted the guy and I thought I was smarter than everyone else’. But then he saw others he knew selling: ‘I just couldn’t hang on anymore . . . The stock was down like 14%; I just sold it.’ Decisions are made in a social environment. As it turned out, he reported: ‘The stock has long ago made back what it lost and has been a super star since then.’ His valuation case was ‘right’ but he let himself be ‘distracted’ and ‘that’s happened to me before’. With hindsight, his original thesis had been correct but he was not able to maintain it under the emotional experience of uncertainty. The example illustrates how, without hindsight, there can be no ‘right’ answers at any stage. Note that the context was arguably unique. Buying was one of two plausible options. Calculation could explore options but not resolve them, and he needed to anticipate what others would do. All this made outcomes inherently uncertain. Cooper’s account involved the use of a decision rule—find a company temporarily prejudiced—which is the basis of many business models in asset management (see Dreman 2012). It can also be noted that large gains or losses were potentially involved, evoking emotions such as excitement and anxiety, and that new data could be expected to 65

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

emerge and potentially challenge a thesis (in the role of counterfactuals and conflicting causal models) before outcomes could be known.3 Elaborating on the irreducible cognitive and emotional conflicts facing actors like Cooper and threatening their daily operations, Chong and Tuckett (2015) argue that the crucial question is not what it is optimal to do but how Cooper and others like him manage to act at all. The state of confidence in financial markets is necessarily constructed through repeated psychological and social actions that create, renew, and legitimate the existence of expertise and conviction on an ongoing but ultimately fragile basis. Action requires both that expectations about outcomes excite and that hypothetical doubts about actions’ future potential to create loss are overcome. Indeed, from an outside analytic viewpoint, money managers like Cooper face two ongoing action problems: first, they have to convince themselves and their clients that what they are doing in general is worthwhile and will bring success more often than failure; and secondly, each time they make a portfolio decision, they need to be convinced that the particular action is worthwhile. Conviction narratives—that is, stories which support activities and action and so, in effect, turn uncertainty into a degree of subjective certainty—are required in both cases. Narratives must be developed to support conviction that particular investment strategies and processes are effective and that profitable opportunities can be identified. They are based, as illustrated with Cooper’s decision-making, on locally prevailing beliefs, rules of thumb, underlying ideas of causation and, above all, arguments and exemplification dispelling potential doubts. Detailed analysis by Chong and Tuckett (2015) demonstrated how the money managers interviewed used ‘conviction narratives’ to ‘think’: the narratives provided grounds to approach a decision and, at the same time, dispel the potential doubts that uncertainty necessarily created. Examining the narratives recorded in the interviews, the paper’s authors and a third colleague, making judgements independently, found that within every decision made by the respondents, one or more of six core characteristics, each containing combinations of attractor and/or doubt-repelling elements, could be identified. There were two ‘pure attractor’ situations—either those in which there was opportunity for gain discovered via exceptional effort or ability on the part of the manager (35 Decisions) or by finding an exceptional or unique quality of the underlying entity (68 Decisions). There were three ways of dispelling doubt: by including in the narrative an explanation as to how there was a 3 Subjectively, actors themselves may be so successful at avoiding the existence of alternative but possible narratives that they are not aware of the degree of equivocality unless they are questioned about it. This unreflective type of conviction is analysed later in the chapter as a Divided State (DS) mental state. Examples might include largely unquestioned risk assessment practices in financial markets prior to 2007 (see Kay 2015; Taleb 2004).

66

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Conviction Narrative Theory

limit to possible bad surprises (57 Decisions); or an availability of proven, solid individuals who would not ‘mess up’ (18 Decisions); or by including reasons to suppose the interviewee, in this instance, had not got emotionally carried away with enthusiasm (9 Decisions). Twenty-two decisions contained features that simultaneously acted as attractors and doubt-repellers—for instance, a story element that an entity enjoyed a monopoly or market power and so could both extract high returns and be able to go on doing so.4 Constituted in these forms, their narrative hypotheses allow financial actors to feel committed to their beliefs and to manage dependency on the uncertain future. They could then feel confident in their particular decisions and more generally promote themselves as skilful, and hence survive in an industry in which (according to academic analysis) no individual player is in fact consistently more successful than would be expected by chance (for example, Barras et al. 2010; Busse et al. 2010; Fama and French 2010; Wermers 2011; and see also Kahneman 2011).

Conviction Narrative Theory Looked at from the outside, the problem for fund managers is how do they convince themselves—and any others that they need to influence—that a proposed action will bring about gain rather than loss. Decomposing the deliberations Cooper revealed in his interview, a narrative is revealed within a causal chain. He starts with a search for undervalued securities. He identifies that the value of a particular security [VA-] is depressed by a particular factor [market prejudice C1]. Because he thinks himself capable of making rational rather than emotional arguments [C2], he thinks this is the potential opportunity he is searching for. By talking to the CFO, he carefully investigates the company’s plan [C3] to deal with its problems. He judges it will work out [C4], so that the prejudice influencing value will eventually be dispelled among other market actors [C5], who will revalue the security [VA+]. Cooper’s deliberations comprise a valuation narrative linking the action of investing to a desired outcome. Through it, he became convinced that investing in A would give him gain (so long as the elements of his causal understanding underpinning his narrative [C1, 2, 3, 4, 5] would prove correct). Note, however, that there is more than one narrative within the overall plot and they are all necessary to support different elements of his argument. Indeed, there are sub-narrative plots underpinning the variously identified causal factors: for example, in support of his judgement [C4], Cooper drew on his visit to the company and his conversation with the CFO [C3]. We can see 4 Further details and detailed examples of ‘conviction narratives’ supplied by respondents can be found in Chong and Tuckett (2015, Box 1, p. 17.)

67

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

this argument potentially resting on still deeper level narratives: for instance, to support his judgement about the CFO, Cooper could tell a story, if pressed, about why he is convinced of his personal capacity to know whom to trust and whom not [CC3] and so on. In CNT, this analysis is generalized. To act, when outcomes are (objectively) uncertain, actors faced with radical uncertainty draw on (subjectively) preferred narrative plots of how a planned action will lead to a particular outcome. Such narratives depend on other part narratives (narrative chunks) and, in this way, actors develop a conviction that their intended action will bring about the desired outcome and allow them to make a planned gain rather than a loss. In short, the subjective confidence to act (and often to carry collaborators with you) is enabled by creating or adhering to a conviction narrative linking action and planned outcome through a plot, itself composed of what one might think of as sub-narrative chunks at different levels of the underlying argument.

Narratives Analysis of Cooper’s narrative has revealed that narratives or narrative chunks regularly exist inside each other. A consequence is that defining the term ‘narrative’ precisely becomes somewhat intractable—like deciding which is the ‘real’ doll in a Russian doll. Despite such inherent inexactness, the term is useful. Previous work across a range of disciplines has already deployed it to understand many processes relevant to decision-making. As the arguments rehearsed in this section show, narratives allow actors to give meaning to everyday events and happenings (along with their causal implications), simulate how actions play out, and communicate what they plan to do or have done to others. Importantly, because cognition is embodied, as narrative plots are rehearsed, they also stimulate emotions of approach and avoidance. Bruner (1990), like other narrative theorists (for example, Sarbin 1986; Schank and Abelson 1977; Spence 1984), argues that narratives allow human actors to construct the everyday meaning of events and happenings they experience, including implicit causal mechanisms. His work draws explicitly on related ideas in anthropology and sociology (e.g. Evans Pritchard 1974; Garfinkel 1967; Schutz 1973; Weber 1921) to note that action is always situated. From this standpoint, he questioned the usefulness of the dominant trend in cognitive science, which takes meaning for granted. Computational theories treat cognition as the processing of pre-coded information units; essentially as equivalent to mechanical computation but with limits on memory and processing power. Such an approach is difficult to apply to decisionmaking in radical uncertainty because it ignores the necessary construction of meaning. In radical uncertainty contexts, optimizing available information without considering its meaning and relevance is unlikely to be a ‘rational’ 68

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Conviction Narrative Theory

way to proceed. In that context, we simply cannot know which bits of information to hand—or even which causal models—will actually be useful to model the future. We have to use our imagination. Bruner’s argument (1990, 55–6) is that it is precisely the capacity for narrative framing or schematizing, along with affect regulation, which provides human actors with the predisposition to order experience. For him, narrative provides a typical means of constructing the world without which we would be left ‘lost in a murk of chaotic experience’ (1990, 56). Also for him, cognition and affect are not in conflict but closely connected—something he illustrates by reference to Bartlett’s (1932) classic account of memories as narratively constructed accounts of events that organize experience on the basis of cultural schemata and the pleasant or unpleasant emotions they evoke (Bruner 1990, 57–8). Narratives, therefore, are a crucial element in organizing experience into what one can think of as meaningful and manageable ‘chunks’ (Miller 1956). Such formulations of the role of narrative as a general form of mental organization central to consciousness lie at the heart of CNT. Similar ideas are also at the heart of the psychoanalytic notions of unconscious phantasy and imaginative internal object relations (Freud 1908; Tuckett 2011) and have been proposed, albeit in somewhat different ways, in various branches of psychology and social science. For instance, Tomasello et al. (2005) argue that humans have a highly developed capacity to understand others’ inner states and possess the drive to express their own states. The way they do this is through creating accounts of their experience in inner speech—constructing a running verbal commentary on their activities (Gazzaniga 2000). This paves the way for narrative representation of themselves and their relations to others and for the communication to others of these state-of-the-world narrations and associated plans (for example, Baumeister and Masiacampo 2010).

Four Functions of a Conviction Narrative A conviction narrative combines these general characteristics of narratives with other functions that together support decision-making in radical uncertainty.5 In particular, narratives permit individuals: (i) To make meaningful sense of situations in which they find themselves, or more precisely to identify opportunities for action in any context based on the implicit causal explanations they attach to their observations. 5 Although it is useful to separate the four functions to show how conviction narratives work for expository purposes, they are intrinsically interwoven. For instance, pattern recognition, simulation, and feeling a narrative is accurate tend to occur simultaneously rather than in a convenient analytic causal order. As an example of how this may occur, several current models suggest that most of our brain functions seem to have a predictive nature, so that perception of the present seems to entail a modelled prediction of the future (Clark 2013; Friston 2003; Pezzulo et al. 2015; Wolpert and Miall 1996).

69

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

In the example, Cooper had to interpret data and understand its causal relevance to his subjective plans. He started out with a general conviction narrative as to how to perform his job defined by himself and the institutional context in which he worked: he was looking for pairs of companies—one believed to be overvalued and another, in the same sector, undervalued. One can then think of his environment as full of action cues (revealed through screening systems he has set up) waiting for further investigation and ready to be turned into planned action through the active construction, interpretation, and causal modelling of current realities in terms of available rules of thumb that had worked before. Did the construction company fit the label ‘temporarily “prejudiced” stock’—in which case, he could apply the underlying causal model that stocks whose price is depressed by rumour will eventually rebound? Actors are faced with virtually limitless data. They are able to act because they find cues to implement rules of thumb, based on models they have of how their relevant world works (Tuckett 2011). Procedural rules such as ‘look for prejudiced stock’ organize an actor’s world of opportunity narratives and function like adaptive heuristics (Gigerenzer 2014; Gigerenzer and Gaissmaier 2011). They classify situations into potential action opportunities with predicted outcomes—each rule implying underlying causal mechanisms and the sequential consequences to be expected from the action. In this way, they both indicate what further search might be required and make an uncertain situation intelligible and actionable (see Weick et al. 2005). (ii) To articulate to themselves and simulate alternative representations of the future outcomes of their actions, so as to predict their impact on their plans. Cooper thought about what to do by sketching out scenarios that might follow depending on whether he held on to or sold his investment, consequent upon various imagined events. Whether narratives are formed through ‘telling’ them to oneself, reading, or listening to them or by telling them or writing them for others, they draw on and express the human capacity for foresight and simulation. They permit the future to be imagined, deliberated upon, expressed, and communicated. Although some of the details remain debated topics in contemporary cognitive neuroscience, it seems that the brain processes underlying the ability to travel mentally backwards (memory) or forwards in time (planning) are fairly similar (Suddendorf and Corballis 1997). In one approach, the primary function of what Schacter et al. (2008) call the ‘prospective brain’ is essentially to store (past) experiences for the purpose of anticipating future events. In this way, one can suppose that as different potential actions are considered, feelings are triggered associated with various chunks of experience stored together in the brain (i.e. bits of pattern recognition, categorization, 70

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Conviction Narrative Theory

available adaptive heuristics, sense making, and narrative typification). The different outcomes simulated as the result of different actions are, therefore, simultaneously felt and thought by human actors to be more or less plausible. (iii) To communicate about their intended actions to others in order to gain support in social contexts where action is collaborative. Conviction narratives provide what in sociology might be called ‘logics of action’ (DiMaggio 1997) in a given domain. In psychology, Mar and Oatley (2008) have shown some of the detailed ways narratives serve to communicate complex messages in a comprehensible way. Additionally, because emotion is easily shared, conviction narratives seem likely to be particularly useful to coordinate social support for strategies within organizations or societies. Their framing and development in social contexts (in the development of a corporate plan, for example) will also anticipate the need to use them for those ends. (iv) To articulate and feel convinced about their preferred action, making it possible to sustain a commitment to it, even at the risk of loss. In radical uncertainty, plans have the intrinsic property that they may succeed or fail. Ex ante we cannot know. Consequently, contemplating action to exploit a perceived opportunity must potentially stimulate approach emotions associated with ‘let’s do it’ and/or trigger loss aversion and inhibition (stimulating avoidance emotions). Ambivalence in feeling is intrinsic and inescapable as a potential experience. The crucial function of a conviction narrative, if it is to facilitate action in a radical uncertainty context, is, then, to manage ambivalence. Since action involves embracing a project with the subsequent potential for loss or gain, it requires a narrative that resolves the possibilities in one direction. In the Cooper example, his ability to develop a narrative to repel feelings of doubt was crucial. Initially, his narrative managed his doubt but, as events unfolded, it could no longer do so. There have been significant advances in understanding the role of emotions in action and decision-making in recent decades. It is increasingly clear that emotion and cognition are not separate processes but intertwined at all stages from perception to action (Phelps 2006). Unlike machine learning, human cognition is embodied. In other words, it is ‘the outcome of interaction between perception, action, the body, the environment and other agents, typically during goal achievement’ (Barsalou 2008, 619). For example, narratives have the characteristic that they can create conviction at the cognitive and affective levels. Indeed, a property of narratives is that they allow simulations of the outcome of action to be felt as close to real experience. In particular, conviction narratives play the crucial role of managing approach-avoidance emotional conflicts and so motivating action under 71

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

radical uncertainty. In short, when narratives—that make sense, make predictions, and simulate the outcomes of action—are run internally, told, read, or heard by human actors, they facilitate the development of a particular quality of subjective ‘knowledge’ about the outcome of a plan which combines cognitive and affective experiences to create a sense of what Bruner (1985) called ‘verisimilitude’. ‘Knowledge’ in this formulation is cognitive and affective, based on felt experience: as a narrative is rehearsed, the component elements produce ‘happenings’ in neural architecture. What one might think of as ‘yes, that makes sense’, or ‘no, that doesn’t seem right’ emotional experiences occur. It is a dynamic process giving actors the potential to notice positive or negative feelings and to be motivated to respond to them. Asset managers, of course, know things might go wrong, and their narratives manage that possibility.

CNT and Standard Decision-Making Models in Psychology: The Role of Emotions In CNT, emotional processes play a different role from that in most other models of judgement and decision-making. Figure 3.1 represents a simplified model of ‘rational’ decision-making in which decisions are mediated by the separate operation of emotional and deliberative processes on action. Kahneman (2011), for example, conceives of emotions as heuristics (an affect heuristic [Zajonc 1980]) evoked automatically and belonging to what he calls ‘System 1’, but not as an essential and useful component of deliberative thought in System 2. In such models, emotions are treated as simply a bias or hindrance to reflective thought. Figure 3.2 depicts the role of emotion in the CNT model. Here, cognitive and emotional processes, activated in an individual’s local social context, interact in a mutually reinforcing fashion to produce a feeling of conviction about narrative prediction (the condition for action involving potential loss or gain under radical uncertainty). Approach emotions evoked in the particular narrative dominate so that the planned action feels convincing. In this way, conviction narratives support action and go on doing so, unless updated. The role given to emotion in the CNT formulation draws on the literature just mentioned and on further literature suggesting that feelings play an organizing or metacognitive role in human thinking and have an evolutionary purpose linked to maintaining homeostatic control (Damasio and Carvalho 2013). An important element of this literature has been the recognition of the importance of the fundamental relationship between emotion (including approach/avoidance motivation) and behaviour at multiple levels from the primitive primary emotional system of the brain through 72

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Conviction Narrative Theory

Deliberative Processes (S2)

Action

Emotional Processes (S1)

Figure 3.1 The role of emotion in deliberation in standard models (for example, Kahneman 2011) Cognitive (Deliberative) Processes Conviction Narrative: Approach > Avoid

Action

Emotional Processes

Figure 3.2 The role of emotion in decisions to act in CNT

learning processes to higher order cognitions (Panksepp 2013; Rolls 2013; Solms 2013). Thus, not only base affective orientations, but also higher order emotions (with complex cognitive appraisal elements) may be understood as mechanisms for mediating approach behaviours in relation to rewarding opportunities and avoidance behaviours in relation to adverse threats.6 From these foundations, one can then think of narrative simulation as literally creating states of the body—deep subjectively experienced states of well-being or discomfort evoking approach or avoidance (Damasio and Carvalho 2013). While simulating outcomes, actors imaginatively project their bodies into the future to anticipate the experience of their future self as well as that of the others represented. To feel there are good grounds to act, a decision-maker must, overall, be able to repel doubts that may come up. Here, the ability of human actors to draw on feelings of conviction provides an

6

It cannot be developed here but this definition of the emotions that matter is not the same as traditional approaches to emotional valence (which contrast what are called negative versus positive emotions).

73

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

advantage unavailable to a computer generating only scenarios. While a computer model operating as a non-human observer may be unable to identify secure grounds to support a particular narrative of the future in radical uncertainty and so commit to a particular decision, a human decision-maker can generate a feeling of conviction sufficient to act. Narratives create experienced rather than just abstract ‘knowledge’: they provide support for action founded on an emotionally coloured and subjective feeling of ‘knowing’ what will happen.

CNT and Animal Spirits Keynes (1936) critiqued his contemporaries’ economic models by arguing that—given radical uncertainty—investment decisions cannot depend on strict mathematical expectation, ‘since the basis for making such calculations does not exist’ (163). Rather, he argued, sufficient actors are usually able to ‘supplement’ and support reasonable calculation with ‘animal spirits’, to put aside thoughts ‘of ultimate loss . . . as a healthy man puts aside the expectation of death’ (1936, 162). Based on such microfoundations, Keynes considered that macroeconomic developments depend on the overall state of confidence and its influence on individual decision-makers: [I]f the animal spirits are dimmed and the spontaneous optimism falters, leaving us to depend on nothing but a mathematical expectation, enterprise will fade and die;—though fears of loss may have a basis no more reasonable than hopes of profit had before. (Keynes 1936, 162)

CNT focuses precisely on the way that narratives manage anticipations of gain and loss and so support action emotionally. In this sense, it can be seen as a theory of Keynesian animal spirits, but one based on contemporary research in brain and social science. To summarize the theory, human actors use socially constructed narratives to make sense of their world and the opportunities it presents them, and to feel sufficiently convinced (via simulation) about the outcomes of their planned actions to take those opportunities. Narratives evoke reactions, ultimately within the human interoceptive system, which stimulate felt emotions of approach and avoidance towards action. Conviction narratives, in which approach predominates, support actors in selecting and committing to preferred actions, although loss (and so loss aversion) is possible. They allow actors to draw on their cognitive and emotional resources both to motivate action and repel doubt. Such narratives draw on biologically and socially evolved capacities that allow individuals subjectively to prepare to execute particular actions even though, because of radical uncertainty, they cannot objectively know what the outcomes will be. Such narratives are also 74

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Conviction Narrative Theory

facilitating because they provide an easy means for actors to communicate and gain support from others for their selected actions, as well as justify themselves when this is required. In short, conviction narratives perform coordination and audit-trail functions, as well as motivating action.

Divided States and New Microfoundations for Macroeconomics Calculative processes based on ‘rational expectations’, as defined in economics, depend on the availability of a decision-making context in which the probabilities of all possible actions and their outcomes are available to the decision-maker, ex ante. It is a context that excludes radical uncertainty and most decisions made in the worlds of finance, business, and government. These broader contexts are what Savage (1954) termed ‘large worlds’ to which he considered it would be ‘ridiculous’ to apply Bayesian probabilistic ideas from the ‘small worlds’ of gambles and risky choice (Volz and Gigerenzer 2012). Rational choice in an economics sense can apply only in those limited instances where the choice context is stable or predetermined. Theories based on rational expectations, therefore, necessarily recede before a great deal of empirical reality (Berezin 2005). Because current economics relies on a restricted equilibrium version of rational choice theory, it has significant difficulties (as recent events have proved) when faced with the task of understanding a macroeconomy composed of numerous decisions made in radical uncertainty, even if modified by the limited importation of psychological thinking found in behavioural economics. CNT, on the other hand, focuses on the fact that radical uncertainty stimulates irresolvable conflicts of opportunity and loss; and it takes advantage of current science, suggesting that such conflicts must activate the twin pathways evoking approach and avoidance in brain architecture. It extends decision and economic theory to allow them to be less unreal and to create greater explanatory power to understand how and why economies are dynamic. The sociologist and psychoanalyst Smelser (1998), in a Presidential Address to the American Sociological Association, also emphasized the potential value of theories that incorporate conflicting aims. He argued that ambivalence rather than rational expectations might be a promising basis on which to build aspects of economic and social theory. Ambivalence—formally the state of holding ‘opposing affective orientations towards the same person, object or symbol’ (1998, 5)—is, he suggested, characteristic of how individuals experience social life. He provided numerous examples of states that generate it. For example, bonding with others creates dependence that can be positive but can also signal a loss of freedom; parents can be both happy and sad when their 75

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

children grow up and leave home; consumers can be both attracted to and feel guilty about the pursuit of status goods; and members of groups and organizations can be both enriched and depleted by belonging. Drawing on the concept of ambivalence to understand financial instability, Tuckett (2011; Tuckett and Taffler 2008) extended Smelser’s analysis to propose the psychoanalytic-based concepts of Integrated (IS) and Divided State (DS) mental states. They are dispositional properties (Stinchcombe 2005) that refer to relations between thoughts and the feelings they evoke. One state, DS, is conceived as an orientation towards a particular narrative characterized by the apparent absence of felt ambivalence. It is recognizable in contexts when, although different outcomes are conceivable, conflicting narratives are absent in discourse. In DS, feelings such as doubt, frustration, humiliation, defeat, or disappointment, for example, which might evoke avoidance and create a shift towards abandonment of the current exciting, promising, fulfilling narrative, are absent. In DS, only partial non-ambivalent narratives of self and other relationships are allowed. The other state, IS, discussed among psychoanalysts with reference to the poet Keats’ ‘negative capability’,7 is an alternative mental disposition characterized by the emotional ability to tolerate feelings of doubt or ambivalence when they are aroused by thoughts and to retain curiosity about both their source and potential evolution. In such states, actors can reflect on alternative and contradictory narratives of the future and act even if some thoughts create unpleasant feelings because they threaten convictions about the outcome of plans. IS and DS are conceived as omnipresent and shifting states, each with very different implications for appraising the outcome of action. They influence perception of elements in an actor’s environment and are influenced by shifts within it—other people’s behaviour, news, innovation, and so on. They exist as dispositions simultaneously and overlap one another, but with one always dominating mental proceedings at any one moment. What is termed groupthink ( Janis 1982), groupfeel (Tuckett 2011), or an analytical monoculture (Bronk and Jacoby 2016) are perhaps best seen as a dominant DS state at a group level— functioning in what Bion (1952) labels a ‘Basic Assumption’ group way. Group members no longer assess risk or work in the heterogeneous mode assumed by equilibrium economic theories, which is closer to IS. Both the cognitive diversity and emotional pluralism considered essential for the healthy operation of markets are likely to be absent when DS are not suitably balanced by IS.

7 See Britton (2013) and the editorial introduction to this volume. Beckert and Bronk make the point (p. 3) that ‘negative capability’ is a key facet of imaginative thinking—being receptive to new pointers that challenge settled views: being willing to remain ‘in uncertainties, mysteries, doubts, without any irritable reaching after fact and reason’.

76

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Conviction Narrative Theory

CNT and the concepts of Ds, IS, and groupfeel together provide analytic tools to test Keynes’ intuition that shifts in animal spirits lie behind the market instability seen in the 1930s and more recently in the events leading up to 2008 and afterwards. In markets characterized by radical uncertainty, such as finance, individual decisions are enabled by conviction narratives influenced both by the state of mind of decision-makers and the shared narratives they find convincing in their surrounding reference network. Exciting phantastic object8 narratives circulating through relevant networks can become irresistible in Ds because they evoke great excitement not mediated by anxiety and enquiry into doubts. A clear emotional basis for groupfeel (Tuckett 2011) then exists. When all or most actors share a non-ambivalent conviction in the same narrative, the conditions are ripe for the sort of ‘deceptive reflexive feedback loops’ (Bronk 2013) that Soros (1987) argues lie behind boom and bust. Emotions of conviction and confidence are as contagious and potentially destabilizing as those of fear (Akerlof and Shiller 2009), unless they are moderated by IS. Congruent with the ideas set out in this chapter, Shiller (2017) suggested in an address to the American Economic Association that the human brain has always been highly attuned to narratives, whether factual or not, to justify ongoing actions. Because narratives can ‘go viral’, he elaborated, it is plausible that they might drive economic and financial fluctuations. He went on to suggest that the quantitative study of narratives using new methods might, therefore, help to gain a better economic understanding. One way to study narratives and their influence on economic and financial fluctuations not previously attempted is to make use of the emphasis in CNT on the precise role in economic action played by approach and avoidance emotions, as indicators of animal spirits. Some preliminary work that does this is proving empirically useful. First, evidence has been obtained that changes in the balance of approach and avoidance emotions in news documents (shifts in animal spirits, one might say) appears to contain significant information for forecasting the US and UK economies in the period 1996–2014 (Tuckett and Nyman 2017) and also around the time of the Great Depression ( James et al. forthcoming). Second, statistically sustained shifts in the balance of approach and avoidance emotions in selected texts, taken as an indication of the development of Ds narratives, appear to offer the potential to provide early warning of developing financial groupthink (Nyman et al. 2017; Tuckett et al. 2014).

8 Phantastic Objects (Tuckett 2011) are subjectively very attractive or idealized ‘objects’ (people, ideas, or things) that we find highly exciting and idealize, imagining (feeling rather than thinking) that they can satisfy our deepest desires (biological priors), the meaning of which we are only partially aware. Examples: dot-com stocks or CDOs.

77

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

Summary and Conclusion This chapter has described the radically uncertain context faced by money managers and how they cope by developing conviction narratives. It then generalized these findings to introduce a wider theory of decision-making under radical uncertainty, termed Conviction Narrative Theory (CNT). CNT differs from standard approaches to decision-making in economics and behavioural psychology, which are limited to theories of efficient and inefficient information processing in contexts where data is available to calculate future probabilities. In contrast, CNT draws on the human capacity to organize experience through narrative, and demonstrates how cognitive and affective responses are combined to facilitate action opportunities. Although novel in psychology, economics, and sociology, CNT is supported by a significant body of current theory and research into narratives and cognitive functioning. Building on this research, the chapter has argued that CNT can operationalize Keynes’ (1936) formulation of animal spirits as a human solution to radical uncertainty, and provide alternative, more plausible, and empirically substantiated microfoundations on which to build a macroeconomic understanding of the development of monocultures and financial market instability. The chapter concluded by arguing that discourse analysis of news feeds and other narrative forms can provide new tools for understanding and forecasting an economy’s evolution.

Bibliography Akerlof, George A., and Robert J. Shiller. 2009. Animal Spirits: How Psychology Drives the Economy and Why It Matters. Princeton, NJ: Princeton University Press. Barras, Laurent, Olivier Scaillet, and Russ Wermers. 2010. ‘False Discoveries in Mutual Fund Performance: Measuring Luck in Estimated Alphas’. Journal of Finance 65 (1): pp. 179–216. Barsalou, Lawrence W. 2008. ‘Grounded Cognition’. Annual Review of Psychology 59: pp. 617–45. Bartlett, Frederic Charles. 1932. Remembering: A Study in Experimental and Social Psychology. Cambridge: Cambridge University Press. Baumeister, Roy F., and E.J. Masicampo. 2010. ‘Conscious Thought Is for Facilitating Social and Cultural interactions: How Mental Simulations Serve the Animal–Culture Interface’. Psychological Review 117 (3): pp. 945–71. Berezin, Mabel. 2005. ‘Emotions and the Economy’. In Handbook of Economic Sociology, edited by Neil J. Smelser and Richard Swedberg, pp. 109–27. 2nd edn. New York and Princeton, NJ: Russell Sage Foundation and Princeton University Press. Bion, Wilfred R. 1952. ‘Group Dynamics: A Review’. International Journal of Psychoanalysis 33 (2): pp. 235–47.

78

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Conviction Narrative Theory Britton, Ronald. 2013. ‘Commentary on Three Papers by Wilfred R. Bion’. Psychoanalytic Quarterly 82 (2): pp. 311–21. Bronk, Richard. 2013. ‘Reflexivity Unpacked: Performativity, Uncertainty and Analytical Monocultures’. Journal of Economic Methodology 20 (4): pp. 343–9. Bronk, Richard, and Wade Jacoby. 2016. ‘Uncertainty and the Dangers of Monocultures in Regulation, Analysis, and Practice’. MPIfG Discussion Paper 16/6, Max Planck Institute for the Study of Societies, Cologne. Bruner, Jerome. 1985. ‘Narrative and Pragmatic Modes of Thought’. In Learning and Teaching the Ways of Knowing. Eighty-Fourth Yearbook of the National Society for the Study of Education. Part II, edited by Elliot Eisner, pp. 97–115. Chicago, IL: University of Chicago Press. Bruner, Jerome. 1990. Acts of Meaning. Cambridge, MA: Harvard University Press. Busse, Jeffrey A., Amit Goyal, and Sunil Wahal. 2010. ‘Performance and Persistence in Institutional Investment Management’. Journal of Finance 65 (2): pp. 765–90. Chatfield, Chris. 1995. Problem Solving: A Statistician’s Guide. 2nd edn. London: Chapman and Hall. Chong, Kimberly, and David Tuckett. 2015. ‘Constructing Conviction through Action and Narrative: How Money Managers Manage Uncertainty and the Consequences for Financial Market Functioning’. Socio-Economic Review 13 (2): pp. 309–30. Clark, Andy. 2013. ‘Whatever Next? Predictive Brains, Situated Agents, and the Future of Cognitive Science’. Behavioral and Brain Sciences 36 (3): pp. 181–204. Damasio, Antonio, and Gil B. Carvalho. 2013. ‘The Nature of Feelings: Evolutionary and Neurobiological Origins’. Nature Reviews Neuroscience 14 (2): pp. 143–52. DiMaggio, Paul. 1997. ‘Culture and Cognition’. Annual Review of Sociology 23: pp. 263–87. Dreman, David. 2012. Contrarian Investment Strategies: The Psychological Edge. New York: Free Press. Evans Pritchard, E.E. 1974. Nuer Religion. New York: Oxford University Press. Fama, Eugene F., and Kenneth R. French. 2010 ‘Luck versus Skill in the Cross-Section of Mutual Fund Returns’. Journal of Finance 65 (5): pp. 1915–47. Fox, Craig R., and Amos Tversky. 1995. ‘Ambiguity Aversion and Comparative Ignorance’. The Quarterly Journal of Economics 110 (3): pp. 585–603. Freud, Sigmund. 1908. Creative Writers and Day-Dreaming. Standard Edition, 9: pp. 141–53. London: Hogarth. Friston, Karl. 2003. ‘Learning and Inference in the Brain’. Neural Networks 16 (9): pp. 1325–52. Garfinkel, Harold. 1967. Studies in Ethnomethodology. Englewood Cliffs, NJ: Prentice Hall. Gazzaniga, Michael S. 2000. ‘Cerebral Specialization and Interhemispheric Communication: Does the Corpus Callosum Enable the Human Condition?’ Brain 123: pp. 1293–326. Gigerenzer, Gerd. 2014. Risk Savvy: How to Make Good Decisions. New York: Viking/ London: Allen Lane. Gigerenzer, Gerd. and Wolfgang Gaissmaier. 2011. ‘Heuristic Decision Making’. Annual Review of Psychology 62: pp. 451–82.

79

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures James, Harold, Ali Kabiri, Rickard Nyman, and David Tuckett. Sentiment Shifts and the Great Depression. Forthcoming. Janis, Irving L. 1982. Groupthink. 2nd edn. Boston, MA: Houghton Mifflin. Kahneman, Daniel. 2011. Thinking, Fast and Slow. New York: Farrar, Straus and Giroux. Kay, John. 2015. Other People’s Money: Masters of the Universe or Servants of the People. London: Profile Books. Keynes, John Maynard. 1936. The General Theory of Employment, Interest and Money. London: Macmillan and Co. King, Mervyn. 2016. The End of Alchemy: Money, Banking and the Future of the Global Economy. London: Little, Brown. Knight, Frank H. 1921. Risk, Uncertainty, and Profit. Boston, MA: Houghton Mifflin Co. Lane, David A., and Robert R. Maxfield. 2005. ‘Ontological Uncertainty and Innovation’. Journal of Evolutionary Economics 15 (1): pp. 3–50. Lehner, Johannes M. 2002. ‘Metaphors, Stories, Models: A Unified Account of Decisions’. Philosophy of Management 2 (1): pp. 35–46. Lempert, Robert J. 2002. ‘A New Decision Sciences for Complex Systems’. Proceedings of the National Academy of Sciences 99 (3): pp. 7309–13. Mar, Raymond A., and Keith Oatley. 2008. ‘The Function of Fiction is the Abstraction and Simulation of Social Experience’. Perspectives on Psychological Science 3 (3): pp. 173–92. Miller, George A. 1956. ‘The Magical Number Seven, Plus or Minus Two: Some Limits on our Capacity for Processing Information’. Psychological Review 63 (2): pp. 81–97. Nyman, Rickard, David Gregory, Sujit Kapadia, Paul Ormerod, David Tuckett, and Robert Smith. 2017. ‘News and Narratives in Financial Systems: Exploiting Big Data for Systemic Risk Assessment’. For presentation at the NBER Summer School on Forecasting, 11 July 2017. Panksepp, Jaak. 2013. ‘Cross-Species Neuroaffective Parsing of Primal Emotional Desires and Aversions in Mammals’. Emotion Review 5 (3): pp. 235–40. Petersen, Arthur C. 2006. Simulating Nature: A Philosophical Study of Computer-Simulation Uncertainties and Their Role in Climate Science and Policy Advice. Apeldoorn, the Netherlands and Antwerp, Belgium: Het Spinhuis. Pezzulo, Giovanni, Francesco Rigoli, and Karl Friston. 2015. ‘Active Inference, Homeostatic Regulation and Adaptive Behavioural Control’. Progress in Neurobiology 134 (November): pp. 17–35. Phelps, Elizabeth A. 2006. ‘Emotion and Cognition: Insights from Studies of the Human Amygdala’. Annual Review of Psychology 57: pp. 27–53. Rolls, Edmund T. 2013. ‘What Are Emotional States, and Why Do We Have Them?’ Emotion Review 5 (3): pp. 241–7. Sarbin, Theodore R. 1986. Ed. Narrative Psychology: The Storied Nature of Human Conduct. New York: Praeger. Savage, Leonard J. 1954. The Foundations of Statistics. New York: Wiley. Schacter, Daniel. L., Donna Rose Addis, and Randy L. Buckner. 2008. ‘Episodic Simulation of Future Events: Concepts, Data, and Applications’. Annals of the New York Academy of Sciences 1124: pp. 39–60.

80

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Conviction Narrative Theory Schank, Roger C., and Robert P. Abelson 1977. Scripts, Plans, Goals and Understanding: An Inquiry into Human Knowledge Structures. Mahwah, NJ: Lawrence Erlbaum Associates. Schutz, Alfred. 1973. The Problem of Social Reality. Vol. 1. Collected Papers. The Hague: Martinus Nijhoff. Shiller, Robert J. 2017. ‘Narrative economics’. Cowles Foundation Discussion Paper no 2069. Cowles Foundation for Research in Economics, Yale University. Smelser, Neil J. 1998. ‘The Rational and the Ambivalent in the Social Sciences: 1997 Presidential Address’. American Sociological Review 63 (1): pp. 1–16. Solms, Mark. 2013. ‘The Conscious Id’. Neuropsychoanalysis 15 (1): pp. 5–18. Soros, George. 1987. The Alchemy of Finance. New York: John Wiley and Sons. Spence, Donald P. 1984. Narrative Truth and Historical Truth: Meaning and Interpretation in Psychoanalysis. New York: Norton. Stinchcombe, Arthur L. 2005. The Logic of Social Research. Reprint from 1970. Chicago, IL: Chicago University Press. Suddendorf, Thomas, and M. C. Corballis. 1997. ‘Mental Time Travel and the Evolution of the Human Mind’. Genetic, Social, and General Psychology Monographs 123 (2): pp. 133–67. Taleb, Nassim N. 2004. Fooled by Randomness. London: Penguin Press. Tomasello, Michael, Malinda Carpenter, Josep Call, Tanya Behne, and Henrike Moll. 2005. ‘Understanding and Sharing Intentions: The Origins of Cultural Cognition’. Behavioral and Brain Sciences 28: pp. 675–735. Tuckett, David. 2011. Minding the Markets: An Emotional Finance View of Financial Instability. London: Palgrave Macmillan. Tuckett, David. 2012. ‘Financial Markets are Markets in Stories: Some Possible Advantages of Using Interviews to Supplement Existing Economic Data Sources’. Journal of Economic Dynamics and Control 36 (8): pp. 1077–87. Tuckett, David, Antoine Mandel, Diana Mangalagiu, Allen Abramson, Jochen Hinkel, Konstantinos Katsikopoulos, Thierry Malleret, Igor Mozetic, Paul Ormerod, Robert Elliot Smith, Tommaso Venturini, and Angela Wilkinson. 2015. ‘Uncertainty, Decision Science, and Policy Making: A Manifesto for a Research Agenda’. Critical Review 27 (2): pp. 213–42. doi:10.1080/08913811.2015.1037078. Tuckett, David, and Milena Nikolic. 2017. ‘The Role of Conviction and Narrative in Decision Making under Radical Uncertainty’. Theory and Psychology 27 (4): pp. 501–23. Tuckett, David, and Richard Taffler. 2008. ‘Phantastic Objects and the Financial Market’s Sense of Reality: A Psychoanalytic Contribution to the Understanding of Stock Market Instability. The International Journal of Psychoanalysis 89 (2): pp. 389–412. Tuckett, David and Rickard Nyman. 2017. ‘The Relative Sentiment Shift Series for Tracking the Economy’. Under review. https://www.researchgate.net/publication/ 316923347_THE_RELATIVE_SENTIMENT_SHIFT_SERIES_FOR_TRACKING_THE_ ECONOMY. Tuckett, David, Robert Elliot Smith, and Rickard Nyman. 2014. ‘Tracking Phantastic Objects: A Computer Algorithmic Investigation of Narrative Evolution in Unstructured Data Sources’. Social Networks 38: pp. 121–33.

81

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures Volz, Kirsten G., and Gerd Gigerenzer. 2012. ‘Cognitive Processes in Decisions Under Risk are not the Same as in Decisions Under Uncertainty’. Frontiers in Neuroscience 6, Article 105, 1–5. DOI=10.3389/fnins.2012.0010 https://www.frontiersin.org/article/ 10.3389/fnins.2012.00105. Weber, Max. 1921. The Theory of Economic and Social Organization. Translated by A. M. Henderson and Talcott Parsons. New York: Oxford University Press. Weick, Karl E., Kathleen M. Sutcliffe, and David Obstfeld. 2005. ‘Organizing and the Process of Sensemaking’. Organization Science 16 (4): pp. 409–21. Wermers, Russ. 2011. ‘Performance Measurement of Mutual Funds, Hedge Funds and Institutional Accounts’. Annual Review of Financial Economics 3: pp. 537–74. Wolpert, Daniel M., and R. C. Miall. 1996. ‘Forward Models for Physiological Motor Control’. Neural Networks 9 (8): pp. 1265–79. Zajonc, Robert B. 1980. ‘Feeling and Thinking: Preferences Need No Inferences’. American Psychologist 35 (2): pp. 151–75.

82

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

4 Arctic Futures Expectations, Interests, Claims, and the Making of Arctic Territory Jenny Andersson

Introduction AI: The Chinese are there, they’re doing some kind of weird research. LT: The Chinese even have an icebreaker. AI: And the Norwegians are moving forwards. (Interview 1, Board members, Mistra, Swedish Strategic Environmental Research Agency)

The Arctic is a territory made up of expectations. The North Pole is a mythical place of emptiness, absence, and virginity in historical as well as contemporary representations. It is not, of course, an empty space at all. The North Pole is and has been inhabited both by Inuit populations and colonial settlement (Bravo and Sörlin 2002; Jorgensen and Sörlin 2013). Since the end of the Cold War, the Arctic region has gone through a process of profound redefinition, from historical and colonial notions of the polar region as periphery, to notions of the Arctic as an emergent global space (Anderson 2010). This global space is shaped by key processes of world ordering in which both geopolitical and environmental issues are at stake and in which natural phenomena interact with socio-economic processes of representation and regulation. In this ordering process, expectations of the future play a key role. The chapter suggests that these expectations involve a set of claims on the future. Through these claims, a number of actors on different levels set out their future stakes and interests in the Arctic continent, thereby charting the contours of future Arctic territory.

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

The chapter begins by using the case of the Arctic to offer a number of critical reflections on the literature concerning expectations and futurity, focusing in particular on the mutual constitution of interests and imaginaries in the making of future claims. The chapter then explains how the complex actor positions that exist in the Arctic are shaped by the crafting of interests through predicting and imagining the future or reinterpreting the past, and by the drafting of future claims; and how expectations (and the narratives embodying them) are influenced by economic interests and geopolitics. Finally, the chapter pays particular attention to the case of Sweden, which in 2011 became an ‘Arctic nation’.

Claims on the Future: The Mutual Constitution of Expectations and Interests The Arctic is a hub of the global future, as the future of world markets and of the relationship between humanity and nature is played out there. Climate change is intensified in the Arctic, where the temperature is changing twice as quickly as the global average. Key markers of environmental degradation—for instance, persistent organic pollutants (POPs)—are concentrating in Arctic marine and human life. Since the 1970s, the Arctic has also been the site for a set of competing expectations and images related to the region’s future. These competing expectations can be understood as a struggle for dominant images of the Arctic future: they reflect competing narratives and conflicting orders of worth and future interests. This chapter seeks to show that expectations are often directed by interests. Actors pursue dominant images of the future because they have interests in, or normative preferences for, that future; but these interests and normative preferences are themselves partly constituted by expectations or imaginaries. Through the negotiation of such conflicting interests and orders of worth, the relative value of biodiversity and environmental preservation versus the value of continued exploitation of natural resources in the name of growth and competition is settled. In the case of the Arctic, this process of settling future use involves important notions of historical human heritage and ‘stewardship’ of a continent commonly defined as of interest to all humankind. But Arctic futures are also the imagined territorial playground of core national and corporate interests. Because of the magnitude of economic interests in the Arctic continent in an expected de-iced future, the competition over future expectations in the Arctic is also a geopolitical scramble (Young 1998, 2009). As actors on different levels struggle to claim an influence on the region, their capacity to ‘make a claim’—to demonstrate that they have expectations and a future interest in the region—is of vital importance. The many national action plans and 84

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Arctic Futures

strategies discussed in the coming pages serve the purpose of making these claims, and of orienting actor positions around a set of dominant future visions. Arctic actors agree that the future of the Arctic in a world marked by climate change will bring fundamental economic opportunities. As these future opportunities do not by definition yet exist, they need to be actively defined. This explains the very active role played in the definition of relevant interests by a repertoire of future-making and predictive techniques in the Arctic: these techniques vary from quantitative forecasts and prospecting for natural resources, to highly narrative genres of nation branding and the mobilization of historical memory (see Bravo and Sörlin 2002). Understanding how claims-making works in the Arctic offers an opportunity to make some critical observations on the literature on expectations in a range of fields, from innovation studies through Science and Technology Studies (STS) to sociology and financial sociology. As argued most recently by Beckert (2016), expectation, anticipation, and prediction are attempts to manage uncertainty. Scenarios and forecasts play a specific role as ways of stabilizing and making apparently ‘knowable’ an inherently unknown future. Forecasts turn inherent uncertainty about coming developments into a set of seemingly manageable risks by creating fictions about the future, which confer coherence on forms of action in unknown territory (Beckert 2016). Beckert argues that such expectations are not rational, but rather fictional, and that actions do not reflect a set of pre-existing preferences, but rather a belief system of fictions. Capitalism as a system functions because of the stabilizing and coordinating role of expectations, and through a play between this stabilization effect and the opening of new future horizons through innovation, creativity, and novelty (see also Beckert and Bronk’s introduction to this volume). From this perspective, predictive technologies play the role of social and economic coordination by giving actors grounds for more or less coherent anticipations of the future. The high-level scenario processes conducted in relation to the Arctic can be understood as playing precisely this role of structuring action by giving actors forms of information about the anticipated future actions of the other stakeholders involved. The management of expectations should not, however, be thought of as directly related to inherent and naturalized uncertainty. Actors can—a number of studies have shown—produce uncertainty for instance by obscuring the possibly deleterious future effects of their actions in the predictive technologies they use (see Oreskes and Conway 2010 and see also Doganova in this volume). Moreover, narratives and expectations of the future can create coordination effects that have a profoundly destabilizing effect over the long term—for instance, by entrenching unsustainable expectations or postponing solutions to collective action problems. While expectations may be shaped by a range of 85

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

ingenious and creative devices, their real-world effects need not necessarily be thought of as stabilization. Financial forecasts, for example, contributed to profound inbuilt instabilities in the financial system (Holmes 2013; MacKenzie 2008). Similarly, strategies of sustainability, which are designed to bring together incommensurable goals of resource extraction and nature preservation, are not necessarily constructive forms of action for the purpose of avoiding a potentially catastrophic future. They may serve rather to legitimize fundamentally unstable future expectations as part of an essentially capitalist and possibly apocalyptic dynamic (Andersson and Westholm 2018; Swyngedouw 2010). From this perspective, it cannot be taken for granted that forms of prediction help solve problems of coordination in the face of an unforeseeable future. Rather, a critical assessment is needed of the contexts in which predictive technologies might help resolve long-term problems and of contexts in which they might instead be involved in a form of active postponement or depoliticization of solutions to long-term challenges. Indeed, in the event, prediction may neither stabilize nor open up a creative future, but rather close off many potentially valuable options. This criticism can be extended to the literature in, for example, the STS field on prediction as a kind of performative social imaginary (Mallard and Lakoff 2012). Against the idea that predictive technologies somehow put forward a collectively shaped and shared image of the future, predictive technologies can be suspected of creating biases by not giving equal weight to all available future images. They are charged with a highly selective sorting of available images of the future, and give credibility and legitimacy to dominant images of the future. It is therefore important to try to understand this sorting process, and how and through what kinds of social process particular images of the future become dominant. Beckert suggests that fictional expectations can be contested and are subject to a ‘politics of expectations’, but also provide a certain coherence and stability to the future through the structuring effects of narrative. Fictions are dependent on credibility—on providing a coherent belief system (Beckert 2016, 65). This chapter proposes to stress much more strongly that images are often dominant not because of their coherence, but rather because of the geopolitical or economic power of the actors producing them and because of the images’ role in projecting political and market power. Actors pursue certain images of the future because they have an interest in their realization. It seems crucial to put this notion of interest front and centre in the understanding of expectations and imagined futures. It can be argued that forms of prediction provide the basis of claims on the future that are expressions of the ambition to extend key interests over time. In this process, expectations of the future and interests are mutually constituted: expectations clearly form

86

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Arctic Futures

the basis for the articulation of interests; but interests also lead to a production of expectations (scenarios, images, and stories concerning the future) in a highly symbolic and strategic way. In the example of the Arctic, dominant geopolitical actors in the region reiterate long-standing claims to the future of its natural resources by mobilizing new expectations that support existing interests; while less dominant geopolitical actors (for example, Sweden) use imaginaries, narratives, and forecasts actively to craft new and historically virtually non-existent interests in the continent. These interests are dependent on the emergence of a large set of stakeholder expectations on the Arctic future. In this way, expectations play a key role in the shaping of interests that cannot be understood as mere expressions of pre-existing preferences, but rather as actively constituted by notions of what the future might offer. The play between dominant and less dominant expectations described in this chapter establishes a hierarchy of sorts, through which competing notions of the Arctic continent’s future worth are settled. In the process, Arctic territory is made; borders are defined and redefined; powerful and less powerful stakeholders are identified; and strategies of action are drawn up (Dittmer et al. 2011). It can be questioned from this perspective whether the future really remains ‘open’. Rather, the Arctic future is the object of a set of complicated and competing foreclosures (compare Adam and Groves 2007; Anderson 2010). Paradoxically, it is the potentially catastrophic future of the region, namely de-icing, that make these foreclosures possible and leads to a new era of colonization of the Arctic (Dittmer et al. 2011; Dodds and Nuttall 2016; Sörlin and Lajus 2013; Steinberg et al. 2015; Stuhl 2016). To sum up, the chapter makes three contributions to the literature on expectations. It suggests, first, that expectations are often projections of interests over time and are key to the structuring of the Arctic as a future economic space; secondly, that predictive techniques are used by actors in order to make claims in and on the future in a quintessentially selective process in which certain futures, and not others, become dominant; and thirdly, that this competitive process is creative and makes use of a range of repertoires and calculative and narrative future-making techniques, but also a number of social activities and actors that become drawn into the co-production of a future ‘imaginary’ (Jasanoff and Kim 2015). In particular, environmental research is a way of constructing core images of the future within what is known as the eco-modernization paradigm (Andersson and Westholm 2018). Equally important is the social interplay between environmental research, geopolitics, and culturally constructed images of territory.

87

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

Crafting Interests after Ice As historian of the environment and technology Sverker Sörlin points out, the Arctic became human territory through technology, and this is still the case ( Jorgensen and Sörlin 2013; Josephson 2014). As global players meet over the contested borders of a melting continent, natural facts meet with socioeconomic projections in a process that is quintessentially dependent on a multitude of techniques of representation and anticipation (Dittmer et al. 2011). These predictive techniques include such varied items as charts of ice melting; oceanographic maps of sea beds and continental shelves; prospecting for minerals; devices for calculating future gas prices or the effects of gas emissions on the atmosphere and cryosphere; estimations of the reproductive rates of sea mammals and birdlife; economic and social forecasts of the wellbeing of the region’s inhabitants; and geopolitical games and scenarios. Other, less obviously predictive techniques are involved in the making of the Arctic future, too, including narrative devices such as the writing of national action plans and Arctic strategies, storytelling and images about Arctic belonging, and the mobilization of historical memory and heritage (Cooke 2013; Dittmer et al. 2011). All these narrative techniques are crucial. Accessing the future opportunities offered by the Arctic depends on an actor’s capacity to be Arctic, to prove in various ways the existence of a legitimate claim to the continent. ‘Being Arctic’ is a question of a complicated hierarchy of actors, and of judicial and geopolitical positions in an interplay between notions of sovereignty, actual or imagined borders, and national or transnational identities (Gerhardt et al. 2010; Hough 2013; Kraska 2011; Lasserre 2010; Ruel 2011; Young 2009). Iceland, for instance, is an island surrounded by Arctic waters. But Iceland is not a recognized Arctic coastal state. Sweden has had northern territory since the making of the Swedish nation state in the sixteenth century but put significant policy effort into becoming an ‘Arctic state’ only in 2010 and 2011. The Arctic Council represents all Arctic states, but applications to the Arctic Council have been made by both China and the European Union, which are not Arctic states (Jakobson 2010). The return of the Arctic to the tables of border drawing is shaped by a fundamental biophysical process, de-icing. Calculations of the speed at which the sea ice is melting vary, but recent estimates indicate that it is now occurring at an unexpected rate (Comiso et al. 2008; Oechel et al. 1993; Stroeve et al. 2007). Underneath the ice is a seabed, partly a global common governed by UNCLOS, but partly also the territorial property of a number of sovereign nation states, the so-called Arctic Five. UNCLOS, the UN Law of the Sea, was signed in 2008 but stipulates a ten-year period during which coastal nations can produce estimates of the continental shelf. Such estimates became 88

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Arctic Futures

central when, in the same year, a US geological survey showed that the Arctic continental shelf hides approximately one-third of the world’s unexplored oil reserves and up to two-thirds of global gas reserves (US Energy Information 2009; Gautier et al. 2009). The survey, produced at a time of rising oil prices and volatility in commodity markets, reignited geopolitical tensions in the region and conflicts between actors, such as an unsettled historical territorial conflict over the continental shelf between Iceland and Russia. The expectation of carbohydrate resources also caused an important fissure in international transpolar collaboration between those with seabed claims and those without. But access to core natural resources is only one of the crucial expectations relating to Arctic futures. Even more important is the fact that, with no ice, the mythical North West and North East Passages open up, freeing direct trade routes from Europe to Asia. The opening up of these passages is a longstanding colonial dream that is now expected to alter world shipping routes and control over world trade. While Russia is the main protagonist with the United States over gas reserves, China expects core benefits from Arctic shipping (Baker 2007; Chen 2012). In addition, the Nordic countries and the Inuit nations have formulated expectations concerning the future Arctic as a significant tourist destination, which would provide a source of economic recovery for the largely postindustrial Arctic regions. Sweden and Norway foresee a future for their Northern regions as new playgrounds for the creative industries. Sweden has an ice hotel in Jokkmokk, and Facebook has server halls in Kiruna (ironically, because it’s cold). At the same time, China and the EU are competing over the use of Arctic waters as a central ground for commercial fishery. As world oceans warm, fish stocks move north, making Arctic waters crucial sources of global food supply. The Arctic is also a central site for research, including groundbreaking climate and environmental research, high-level international science collaborations, and forms of planetary data production enabled by multifunction platforms and Satellite Arctic Observation Networks or SAONS (see Edwards 2014). These involve very high-level public-private partnerships and funding schemes involving both national public actors and private and financial institutions. While several important planetary regulations and transnational treaties— UNCLOS, the UN Commission on the Limits of the Continental Shelf (CLCS), the UN Framework Convention on Climate Change (UNFCCC), the Convention on Biological Diversity (CBD), and the UN Declaration on the Rights of Indigenous Peoples—constitute attempts to govern the Arctic as a global commons, the prospect of de-icing has led to the reaffirmation of key national interests in the region. Most national actors have thus set out action plans and 89

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

strategy documents explaining their interest in the future Arctic, thereby implicitly making a claim. This activity intensified noticeably in the period after 2008, with the Obama administration developing an Arctic policy based both on environmental sustainability and possible resource extraction in Alaskan drilling and oil sands. Russia pursues an Arctic strategy in which the Arctic is defined as central to Russian strategic interests, and has declared that it intends to use seabed research to support its extended claim on the continental shelf. Russia also claims a privileged right to control the North West and North East Passages (Josephson 2014, 2016). Norway is in competition with Russian mining interests in Svalbard and the Barents Sea. The 2008 estimation of Arctic oil and gas reserves also led to the breakaway from the so-called Arctic Eight—the transnational body of the Arctic Council— of the Arctic coastal nations, the Arctic Five: Canada, the United States, Russia, Denmark (Greenland), and Norway. The Arctic Five’s Ilulissat Declaration broke with the idea of the Arctic as a common interest, which was the starting point of international declarations on the Arctic from the Cold War era. Instead, it defines the Arctic as made up of agreed-upon national territorial claims. It also states—in direct contrast to the declarations of the Arctic Council of Eight (Arctic Five plus Sweden, Finland, and Iceland)—that the Arctic Five are against a global Arctic regulatory framework and that territorial disputes and future resource conflicts will be solved by agreement between themselves by virtue of their sovereignty (Ilulissat Declaration). Not only Iceland but also indigenous peoples have been excluded from the Arctic Five. So were the countries and organizations that are observers at the Arctic Council of Eight, including the World Wildlife Fund. These observers have rapidly increased in number and variety since the constitution of the Arctic Council in 1996. Today they include not only nation states but a plurality of non-governmental organizations and financial organizations (Arctic Council [AC] Observers). The EU’s Arctic strategy emphasizes its massive economic interests in the region, in particular in fish and rare minerals (see EC Communication 2008; EC Conclusion on Arctic Issues 2009). The EU hopes that Greenland, with an Arctic coastline, will eventually be an independent member of the European Union. Iceland, which considered EU membership before the financial crisis but lost interest in the wake of the euro crisis, is dependent on American and Nordic support in its territorial conflict with Russia. With the opening up of the Arctic sea passages Reykjavik would be a central world port. In 2011, the Nordic countries signed—for the second time—a Solidarity Declaration, intended to support Iceland, emphasizing regional forms of cooperation, as well as EU interests, and a common Arctic foreign policy based on sustainability and the Sami people’s right to self-determination (Nordic Council 2008). 90

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Arctic Futures

Re-Imagining Climate Change Ilulissat is an ice fjord, into which icebergs break off from one of the fastest moving glaciers in the world. UNESCO has classed several Arctic sites as World Heritage sites, including the Chikchi Sea, in which Royal Dutch Shell began drilling for oil in 2008, until deciding that this posed too high a risk for marine life (see UNESCO 2013; IUCN 2017). As a World Heritage Site, Ilulissat is marketed by Greenland’s tourist agency1 and UNESCO as ‘climate change Ground Zero’ and as a place where climate change, and the human capacity to adapt to it, can be experienced first hand. Adaptation, the tourist agency website claims, is a human capacity championed by Arctic populations and Greenlanders in particular. There have indeed been colonial settlements on Greenland since the fifteenth century, and strategies for coping with the climate have been key to the human presence on the continent (see Jorgensen and Sörlin 2013). Today, this history of human adaptation is a resource being mobilized in a branding exercise in which climate change is construed essentially as an opportunity for the continent despite its obviously detrimental effects on human and animal life. Climate science is not certain of the exact impact of the de-icing of the Arctic continent, although it is likely to have a dramatic impact on world climate through a complex range of factors. These include melting sea- and land-ice, the thawing of permafrost, and the transformation of Arctic ice from a carbon sink to a carbon source. Scenarios of environmental degradation tend, in contemporary Arctic strategies, to be addressed as a set of uncertainties that potentially present fundamental opportunities for the Arctic region. The Arctic Council actively promotes the key expectation that climate change will allow the Arctic region to become a central economic region, due to de-icing, and a leader on climate change issues (AC Declaration 1998, 25; Keskitalo 2002). Mitigation and resilience surfaced in the aftermath of the UN Copenhagen Climate Change Conference in 2009 as ways of providing strategies for dealing with what are now considered the unstoppable effects of global warming (Rockström et al. 2009). Resilience requires a package of strategies in which contemporary markets and financial actors are centrally involved (Mirowski 2014). Following the Copenhagen summit, the Arctic Council has showcased the Arctic as a cutting-edge world example of adaptation strategies, through the development of new surveillance technologies and early warning systems, the intelligent exploitation of Arctic resources, and the mobilization of so-called ‘indigenous knowledge’. It might be suggested that this

1

https://visitgreenland.com/things-to-do/ilulissat-icefjord/.

91

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

sustainability agenda functions as a prerequisite for continuing oil and gas prospecting after 2008. In other words, it serves ‘environmentally sound oil and gas activities that may contribute to the development of the Arctic region’ and ‘technologies that will reduce the risks of oil and gas activities’, including the reduction of spills, as well as CO2 emissions (AC Declaration 2009, 55). This is part of the production of legitimacy around the expectation of extraction as a dominant future. The emphasis on adaptation as a specific promise on the future was a product of intensified transnational collaboration, and also of the inclusion of the indigenous Arctic nations in setting out a future vision for the Arctic. Within this framing of the future, indigenous peoples become ‘leaders’ of adaptation strategies. The Nuuk Declaration in 2011 recognized (for the first time) the Arctic as an inhabited region (AC Declaration 2011). The so-called Kiruna vision for the Arctic, produced by the Swedish presidency of the Arctic Council, stressed: ‘the economic potential of the Arctic in sustainable development is enormous’, in particular through the development of ‘Arctic knowledge’, in the sense of traditional, indigenous knowledge and ways of life that could help humanity prepare for climate change (AC Kiruna Vision 2013). Since 2015, the Arctic Council has actively promoted not only traditional knowledge and new technologies, but also various forms of financial vehicle, such as emissions trading and Methane to Markets Partnerships as part of mitigation strategies. Through these strategies, climate change is indeed being reimagined from an ecological disaster to a set of economic promises. What is often referred to as ‘eco-modernization’ (Baker 2007) can be understood as an economization of the future, through which market processes are used to reconcile conflicting future values. Importantly, this reconciliation requires co-production (Jasanoff and Kim 2015)—that is, the participation of a variety of stakeholders, some of whom can provide important forms of symbolic legitimacy. Dominant images of the future seem to rely on such forms of co-production. In the Arctic case, environmental research and the production of climate data play a particular role in the production of dominant expectations. Environmental research was a key component of the demilitarization of the Arctic at the end of the Cold War and of emergent forms of transnational cooperation from the mid-1990s onwards. Sustainability was a key objective in the construction of the Arctic as a transpolar and global region. Such constructions played on the idea of the Arctic as having a particular connection to the past and future of humanity and being a heritage of humankind. The notion of stewardship—a core element of the idea of resilience—was put forward by the Inuit Circumpolar Council (ICC), consisting of Inuit nations. Following conflicts with petroleum companies in Canada and Alaska in the 1960s, Inuit peoples formed the ICC 92

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Arctic Futures

in order to demand recognition as ‘stewards of the Arctic’, including the right to determine the extraction and use of the continent’s natural resources. They saw this demand as part of a claim to autonomy and nationhood (Cameron 2012; Shadian 2006, 2014). Influenced by the ICC initiative and the appointment of an Inuit minister for Arctic issues in Canada, the Arctic Council of Eight gave indigenous peoples permanent participant status, even if the term ‘peoples’ was defined as having ‘no implications as to the rights that might apply to the term according to international law’ (standing US footnote in Arctic Council Declarations 1996–2008). The symbolic notion of stewardship was taken over by the Arctic Council in the Ottawa Declaration, which emphasized the sustainable development of the region and its resources, and transnational governance as the way to avoid military security issues in favour of a ‘common’ Arctic future (AC Declaration 1996). As we have seen, since 2008 such notions of the Arctic as a common interest to all humankind have been increasingly challenged by national claims to the continental shelf and by the split between the Arctic Eight and the Arctic Five. Environmental research is a core activity of the Arctic Council through the Arctic Science Committee. A key event in the shaping of expectations concerning the Arctic was the Second International Polar Year in 2008, established under the US Arctic Council presidency (AC Declaration of the International Polar Year 2007). As an institution, the Polar Year was a massive exercise in Arctic identity building, particularly by drawing together humanities and social-science research programmes focused on the history of human presence in the Arctic continent. The Polar Year in 2007 had harked back to the first Polar Year, held in the late nineteenth century as the big powers first scrambled for Arctic raw materials (Josephson 2014). Celebrating the memory of the many historic polar excursions that were the basis for colonial ventures in the Arctic was one of the features of that Polar Year. For example, celebrating polar research was one of the core elements in Sweden’s Arctic policy, in which narratives of historic polar expeditions were used strategically as a way of demonstrating a historical presence in the Arctic region.

Arctification: Branding, Narrative, and Historification in Making Arctic Territory Sweden’s approach to the Arctic region is based on the notion that Sweden has core future interests in it. This is paradoxical for two reasons. First of all, while the Arctic plays a role in Swedish history as a hinterland, its Arctic status is ambiguous: Sweden has held territory above the polar circle in Norrland since the colonization of Lapland in the sixteenth and seventeenth centuries, but does not have an Arctic coastline; nor, at least until 2010, did it have an Arctic 93

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

identity (as opposed to, say, Canada or Iceland in which the idea of being northern or Arctic has a strong cultural history legacy; see Avango et al. 2013; Sörlin 1989). Secondly, while there is certainly an historical Swedish presence in the Arctic, Sweden does not as a matter of present fact hold vital current interests there. The writing of the Swedish Arctic Strategy (SAS) in 2011 was therefore a most interesting process, as it was based on the idea that opening up Arctic territory after de-icing presented a set of ‘opportunities that Sweden needs to be able to influence’ (Regeringskansliet 2011). Influencing these opportunities depended on the ability to articulate an Arctic interest and to set out an active future claim. In the SAS, this was formulated as a concrete policy goal, but as one that depended on a highly heterodox set of strategies, which included nation branding, the mobilization and creation of forms of historical memory, and projecting the existence on Swedish territory of Arctic indigenous peoples. High-level Swedish diplomats refer in interviews to a telephone call from US Ambassador John Farrell in 2010, stressing the need for neutral actors in the Arctic region and urging the Nordic countries to actively develop interests that might stave off a direct conflict between the United States, Russia, and China in Arctic waters. The Americans pointed out that Sweden is an Arctic nation: ‘A third of your territory is north of the polar circle. But everyone we talk to talks about Norrland or the far North or whatever. You need to realize that you are an Arctic nation’ (Interview 2, Senior Diplomat, Swedish Foreign Ministry).2 During Sweden’s presidency of the European Union in 2010, the Nordic countries took an active role in drafting the EU’s Arctic policy. The drafting of SAS in 2011 followed on the articulation of this European policy, and involved diplomats who had worked with the EU presidency and were now involved with Sweden’s presidency of the Arctic Council. The strategy paper states at the beginning: ‘Sweden is an Arctic nation with interests in the region’ (Regeringskansliet 2011, 7). It continues by setting out the components of what is referred to as Sweden’s Arctic identity, and by drawing up a list of potential future Swedish economic interests in the Arctic. Developing an Arctic identity required the mobilization of history in order to demonstrate an Arctic presence. In the Strategy, this demonstration of historic presence makes up for a lack of territory and coastline. Interestingly, research—and in this case polar research—is again a source of the creation of expectations and legitimation of a future stake. As the SAS argues: Sweden has been a research nation in the Arctic since Carl Linnaeus’s journey to Lapland in 1732. Linnaeus’s journey to Lapland was a colonial

2 As this is an ongoing research project, the identities of interviewees cannot be disclosed. Interview 1 was conducted on 30 November 2015. Interview 2 was conducted on 17 October 2015.

94

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Arctic Futures

undertaking, part of the Swedification of northern territory. It led to the development of mining in the north, as well as to the imposition of Swedish language and culture among the Sami people. This journey has now been remobilized, as part of Arctification: ‘Swedish research efforts and the Swedish state’s commitment form an independent claim to a Swedish presence in the part of Arctic which is outside of Swedish territory’ (Regeringskansliet 2011, 12). This is not the end of the historical examples cited. Linnaeus’s Laponia trip was the first of many expeditions. In 1758, there was a Swedish meteorological expedition, and in 1875 Nordenskiöld ‘prophetically’ sailed through the North East passage and exclaimed that he had found a new route for trade with China. From the mid-nineteenth century, Sweden sponsored research on Svalbard, the result of a ‘well-functioning regional collaboration between skilled seafarers and scientists’, leading to the mining of coal in the Svea-Mine. These historical polar expeditions are defined in the SAS as precursory examples of Swedish interests in Arctic territory, and as setting out a line of continuity in Sweden’s presence in the region. In this manner, modern forms of polar research, which have served both environmental purposes and prospecting for resources, are viewed in light of a benign tradition of Arctic knowledge creation. In actual fact, modern Swedish polar research began as the Arctic became a theatre of the Cold War. It gained in importance in the 1980s and 1990s. In 1991, the research platform Oden began prospecting sea beds, as well as collecting climate data by satellite technology. These expeditions were forerunners in the setting up of much larger platforms for strategic environmental research in research stations in Abisko and Kiruna that are today part of satellite-empowered networks and key hubs in the Arctic Council research strategy. The SAS cites Sweden as having ‘one hundred years of experience’ in developing environmental monitoring techniques, which produce not only climate data, but also research technologies that will be essential, it is hoped, for more sustainable mining, oil, and gas industries. Sweden thus hopes to be a major producer of green technologies for resource extraction as Russian and Norwegian mining interests take off in coming decades. The Sami people also play a central role in SAS, as a ‘Swedish minority present in the region for 8000 years’, preceding the presence of the Swedish state. The so-called Lappkodicillen in 1751, which regulated Sami migratory rights, is celebrated as a world-leading example of transnational Arctic governance, while SAS is silent about the fact that Sweden has not ratified ILO Convention No. 169 on the rights of indigenous peoples, and that the migratory rights of Sami and reindeer remain an unsettled source of conflict over both territory and industrial interests (Sörlin and Wormbs 2010). Again, co-production and

95

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

the mobilization of research have played a key role in producing narratives of the future. The historical dimension of SAS was based on the historical research produced by a group of Swedish historians within the larger research platform Arctic Futures, piloted by the Swedish Agency for Strategic Environmental Research, Mistra, and set up as one of a number of large research platforms on issues of particular strategic relevance for the future. Arctic Futures was not only a research programme, but also a geopolitical intervention. The research platform was an outcome of the Polar Year, which among other things introduced cross-disciplinary research between environmental scholars and historians, and the idea that such forms of research collaboration could play a key role in shaping the future of the continent. The idea of a major Swedish research platform was thought up during Sweden’s chairmanship of the Arctic Council and partly drafted by the Swedish Ambassador to the Arctic Council (Interview 1). The use of future-creating methods, such as games, scenarios, and forecasts, was an integral part of the platform, and included a number of subcomponents. These included a research project devoted to Arctic games; another called Managing Competition and Promoting Cooperation; a project on economic futures entitled From Resource Hinterland to Global Pleasure Periphery (stressing tourism as a central Swedish interest in the region); and a project entitled, more critically, Assessing Arctic Futures: Voices, Resources, Governance, which aimed to understand the ‘powered production of Arctic futures’ and the role of dissenting voices and national narratives in this process. Arctic Futures was defined in the policy process as an integral part of the strategy for building Sweden’s credibility as an Arctic country and hence as part of building expectations about the future. ‘Of course Sweden’s ambition was to point out that Sweden is an Arctic country’ (Interview 2). According to Mistra, its aim was to contribute to an ‘identification process in Arctic questions’, spread knowledge about Arctic issues, and help Swedes identify with the Arctic (Interview 1). The board of Arctic Futures included both the Swedish Foreign Ministry and a representative of the US Arctic Commission. The ties between the programme and the policy process were manifest. During Sweden’s Arctic Council presidency, Arctic Futures’ researchers were asked by the Ambassador to write policy briefs on Arctic identity and on Sweden’s interest in the region. Some refused. The group of historians in charge of assessing ‘critical future voices of the Arctic’ produced a series of high-level academic works on the history of human settlements in the Arctic and the continent’s complicated relationship between native and colonial identities. These works provided input for SAS. SAS was a key document narrating a story of Sweden as an Arctic nation with Arctic interests, and a strategic element in the active organization of Swedish 96

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Arctic Futures

industrial interests in a de-iced region. Sweden’s interests in the future Arctic are not oil and gas reserves, nor timber. Its interests are in fact dependent on the realization of the expectations of other and more important actors in the core areas of resource extraction and shipping. The Swedish shipping industry, however, is aiming for a strategic role in the construction of ice breakers and in search-and-rescue missions when the North East and North West Passages open up and world navigation travels through Arctic waters. As one of our interviewees points out, Swedes already in the early twentieth century rescued an Italian polar expedition that would otherwise have perished. Russians, Chinese, and Americans will need search-and-rescue teams once traffic increases (Interview 2). Swedish mining companies are also hoping that the development of Arctic mining for rare minerals will be an opportunity to sell extremeconditions equipment. Ten-page ads by the mining industry in the Swedish daily papers indeed show mining as no longer being about people descending into the now doomed domestic Kiruna mine, but increasingly involving the manufacturing and global marketing of high-tech equipment for mining that takes place in other parts of the world, and by other countries (Dagens Nyheter 2016).

Concluding Remarks Futures are, the literature proposes, imagined, calculated, and narrated. Futures are also colonized, and repertoires of future making are directly involved in a process of claims-making that constitutes symbolic socio-economic spaces and territories. As an emerging future economic region, the Arctic is the site of a complicated interplay of interests, in which the expectation of a future after ice is the trigger for a struggle for the right to partake in the opportunities that this may bring. This chapter has shown how this takes place through a complicated and hierarchical game of actor positions—a game in which the shaping of future expectations is key to the projection and protection of future interests, and in which interests also shape the images and expectations of the future that are put forward. In contemporary societies, expectations are objects of management and rationalization processes, through which—so this chapter argues— futures are in fact created. Images of the future, in this sense, are not all equally important, because claims and expectations of the future reflect a social hierarchy and, in this case, a geopolitical hierarchy. Imaginaries can only hope to be performative if those who wield them have power. As Koselleck (1981) argues, in modern societies the future is a dangerously open terrain. States and other powerful actors therefore try to foreclose this future through claims of the kind this chapter has revealed. These claims are 97

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

contested, but the diverging projections are not symmetrical or equal in the influence they can exercise. Against dominant visions of the future stand native, marginal, and alternative images that often lack both scientific and economic backing and thus have little influence in politically contested space. Environmental research is directly caught up in this struggle over future images and is part of the production of expectations and their embedding in a larger narrative of sustainability and resilience of great socio-economic importance. There is an urgent need to understand the socio-economic processes by which claims on the future are made, and by which expectations serve dominant or less dominant interests. Far from having a mere stabilizing function of enabling coordination around one possible future, expectations produce path-dependent effects by closing off possible options. Indeed, they are often part of highly problematic and partisan social imaginaries on future problems (see Adam and Groves 2007). As Anderson (2010) and others have suggested, predictions may well serve to constrain and foreclose the openness of the future by perpetuating structures of the present that make change hard or impossible to imagine. Expectations should thus be understood as powerladen socio-economic constructs, and their link to the interests of the actors propagating them should be examined carefully.

Bibliography Adam, Barbara, and Chris Groves. 2007. Future Matters: Action, Knowledge, Ethics. Leiden: Brill. Anderson, Ben. 2010. ‘Preemption, Precaution, Preparedness: Anticipatory Policy Action and Future Geographies’. Progress in Human Geography 34 (6): pp. 777–98. Andersson, Jenny, and Erik Westholm. 2018. ‘Closing the Future: Environmental Research and the Construction of Selective Consensus’. Under review. Arctic Council. 1996–2016. Declarations. Arctic Council. 2007. Declaration of the International Polar Year, Washington, D.C. Arctic Council. 2013. Kiruna Vision for the Arctic. Avango, Dag, Annika E. Nilsson, and Peder Roberts. 2013. ‘Assessing Arctic Futures: Voices, Resources and Governance’. The Polar Journal 3 (2): pp. 431–46. Baker, Susan. 2007. ‘Sustainable Development as Symbolic Commitment: Declaratory Politics and the Seductive Appeal of Ecological Modernisation in the European Union’. Environmental Politics 16 (2): pp. 297–317. Beckert, Jens. 2016. Imagined Futures: Fictional Expectations and Capitalist Dynamics. Cambridge, MA: Harvard University Press. Bravo, Michael, and Sverker Sörlin. 2002. Eds. Narrating the Arctic: A Cultural History of Nordic Scientific Practices. Canton, MA: Science History Publications.

98

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Arctic Futures Cameron, Emilie. 2012. ‘Securing Indigenous Politics: A Critique of the Vulnerability and Adaptation Approach to the Human Dimensions of Climate Change in the Canadian Arctic’. Global Environmental Change 22 (1): pp. 103–14. Chen, Gang. 2012. ‘China’s Emerging Arctic Strategy’. The Polar Journal 2 (2): pp. 358–71. Comiso, Josefino C., Claire L. Parkinson, Robert Gersten, and Larry Stock. 2008. ‘Accelerated Decline in the Arctic Sea Ice Cover’. Geophysical Research Letters 35 (1): DOI:10.1029/2007GL031972. Cooke, Lisa. 2013. ‘North Takes Place in Dawson City, Yukon, Canada’. In Northscapes: History, Technology, and the Making of Northern Environments, edited by Dolly Jorgensen and Sverker Sörlin, pp. 223–47. Vancouver: UBC Press. Dagens Nyheter. 2016. Gruvindustrin bilaga, 19 February. Dittmer, Jason, Sami Moisio, Alan Ingram, and Klaus Dodds. 2011. ‘Have You Heard the One about the Disappearing Ice? Recasting Arctic Geopolitics’. Political Geography 30 (4): pp. 202–14. Dodds, Klaus, and Mark Nuttall. 2016. The Scramble for the Poles: The Geopolitics of the Arctic and Antarctic. Cambridge: Polity. Edwards, Paul N. 2014. A Vast Machine: Computer Models, Climate Data, and the Politics of Global Warming. Cambridge, MA: MIT Press. European Council. 2008. Communication, 763 final, ‘The European Union and the Arctic Region’. European Council. 2009. Conclusion on Arctic Issues, 8 December. Gautier, Donald L., Kenneth J. Bird, Ronald R. Charpentier, Arthur Grantz, David W. Houseknecht, Timothy R. Klett, Thomas E. Moore, Janet K. Pitman, Christopher J. Schenk, John H. Schuenemeyer, Kai Sorensen, Marilyn E. Tennyson, Zenon C. Valin, and Craig J. Wandrey. 2009. ‘Assessment of Undiscovered Oil and Gas in the Arctic’. Science 324 (5931): pp. 1175–9. Gerhardt, Hannes, Phillip E. Steinberg, Jeremy Tasch, Sandra J. Fabiano, and Rob Shields. 2010. ‘Contested Sovereignty in a Changing Arctic’. Annals of the Association of American Geographers 100 (4): pp. 992–1002. Holmes, Douglas R. 2013. Economy of Words: Communicative Imperatives in Central Banks. Chicago, IL: University of Chicago Press. Hough, Peter. 2013. International Politics of the Arctic: Coming in from the Cold. London: Routledge. International Union for Conservation of Nature and Natural Resources (IUCN). 2017. ‘Natural Marine World Heritage in the Arctic Ocean. Report of an Expert Group and Review Process’. April 2017. Jakobson, Linda. 2010. ‘China Prepares for an Ice-Free Arctic’. SIPRI Insights on Peace and Security no. 1020/2. Jasanoff, Sheila, and Sang-Hyun Kim. 2015. Eds. Dreamscapes of Modernity: Sociotechnical Imaginaries and the Fabrication of Power. Chicago, IL: University of Chicago Press. Jorgensen, Dolly, and Sverker Sörlin. 2013. ‘Introduction: Making the Action Visible, Making Environments in Northern Landscapes’. In Northscapes: History, Technology, and the Making of Northern Environments, edited by Dolly Jorgensen and Sverker Sörlin, pp. 1–17. Vancouver: UBC Press.

99

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures Josephson, Paul. 2016. ‘Putin in the Arctic’. Paper presented at the Futurepol Prediction Conference, Paris, 15–17 April. Josephson, Paul R. 2014. The Conquest of the Russian Arctic. Cambridge, MA: Harvard University Press. Keskitalo, Eva. 2002. Constructing ‘the Arctic’—Discourses of International Region-Building. Rovaniemi: University of Lapland. Koselleck, Reinhart. 1981. ‘Modernity and the Planes of Historicity’. Economy and Society 10 (2): pp. 166–83. Kraska, James. Ed. 2011. Arctic Security in an Age of Climate Change. Cambridge: Cambridge University Press. Lasserre, Francois. 2010. ‘Géopolitiques arctiques: pétrole et routes maritimes au cur des rivalités régionales?’. Critique internationale 4 (49): pp. 131–56. MacKenzie, Donald. 2008. An Engine, Not a Camera: How Financial Models Shape Financial Markets. Cambridge, MA: MIT Press. Mallard, Gregoire, and Andrew Lakoff. 2012. ‘How Claims to Know the Future are Used to Influence the Present: Techniques of Prospection in the Field of National Security’. In Social Knowledge in the Making, edited by Charles Camic, Neil Gross, and Michelle Lamont, pp. 339–78. Chicago, IL: University of Chicago Press. Mirowski, Philip. 2014. Never Let a Serious Economic Crisis Go to Waste: How Neoliberalism Survived the Financial Meltdown. London: Verso. Nordic Council. 2008. Declaration of Solidarity in the Arctic region. http://www. norden.org/en/nordic-council/cases-and-documents/b-253-presidiet. Oechel, Walter C., Steven J. Hastings, George Vourlitis, Mitchell Jenkins, George Riechers, and Nancy Grulke. 1993. ‘Recent Change of Arctic Tundra Ecosystems from a Net Carbon Dioxide Sink to a Source’. Nature 361 (6412): pp. 520–3. Oreskes, Naomi, and Erik M. Conway. 2010. Merchants of Doubt: How a Handful of Scientists Obscured the Truth on Issues from Tobacco Smoke to Global Warming. New York: Bloomsbury. Regeringskansliet (Swedish Government Office). 2011. ‘Sweden’s Arctic Strategy, Sveriges Arktiska strategi’. http://www.government.se/country-and-regional-strategies/ 2011/10/swedens-strategy-for-the-arctic-region/. Rockström, Johan, Will Steffen, Kevin Noone, Åsa Persson, F. Stuart Chapin, III, Eric F. Lambin, Timothy M. Lenton, Marten Scheffer, Carl Folke, Hans Joachim Schellnhuber, Björn Nykvist, Cynthia A. de Wit, Terry Hughes, Sander van der Leeuw, Henning Rodhe, Sverker Sörlin, Peter K. Snyder, Robert Costanza, Uno Svedin, Malin Falkenmark, Louise Karlberg, Robert W. Corell, Victoria J. Fabry, James Hansen, Brian Walker, Diana Liverman, Katherine Richardson, Paul Crutzen, and Jonathan A. Foley. 2009. ‘A Safe Operating Space for Humanity’. Nature 461 (7263): pp. 472–5. Ruel, Genevieve. 2011. ‘The (Arctic) Show Must Go on: Natural Resource Craze and National Identity in Arctic Politics’. International Journal 66 (4): pp. 825–33. Shadian, Jessica. 2006. ‘Remaking Arctic Governance: The Construction of an Arctic Inuit Polity’. Polar Record 42 (3): pp. 249–59. Shadian, Jessica. 2014. The Politics of Arctic Sovereignty: Oil, Ice, and Inuit Governance. London: Routledge.

100

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Arctic Futures Sörlin, Sverker, and Johan Lajus. 2013. ‘An Ice Free Arctic Sea? The Science of Sea Ice and its Interests’. In Media and the Politics of Arctic Climate Change. When the Ice Breaks, edited by Miyase Christensen, Annika Nilsson, and Nina Wormbs, pp. 70–92. London: Palgrave Macmillan. Sörlin, Sverker, and Nina Wormbs. 2010. ‘Rockets and Reindeer: A Space Development Pair in a Northern Welfare Hinterland’. In Science for Welfare and Warfare: Technology and State Initiative in Cold War Sweden, edited by Per Lundin, Niklas Stenlas, and Johan Gribbe, pp. 131–51. Canton, MA: Science History Publications. Sörlin, Sverker. 1989. Land of the Future: Norrland and the North in Swedish and European Consciousness. Umea: Centre for Arctic Culture Research. Steinberg, Phillip E., Jeremy Tasch, and Hannes Gerhardt. 2015. Contesting the Arctic: Rethinking Politics in the Circumpolar North. London: IB Tauris. Stroeve, Julienne, Marika M. Holland, Walt Meier, Ted Scambos, and Mark Serreze. 2007. ‘Arctic Sea Ice Decline: Faster than Forecast’. Geophysical Research Letters 34 (9): DOI: 10.1029/2007GL029703. Stuhl, Andrew. 2016. Unfreezing the Arctic: Science, Colonialism, and the Transformation of Inuit Lands. Chicago, IL: University of Chicago Press. Swyngedouw, Erik. 2010. ‘Apocalypse Forever? Post-Political Populism and the Spectre of Climate Change’. Theory, Culture & Society 27 (2–3): pp. 213–32. UNESCO. 2013. ‘Final Report of the Resolutions Adopted during the 19th Session of the General Assembly of States Parties to the World Heritage Convention’. US Energy Information Administration, Oil and Gas Division. 2009. ‘Arctic Oil and Natural Gas Potential’. 19 October. Young, Oran R. 1998. Creating Regimes: Arctic Accords and International Governance. Ithaca, NY: Cornell University Press. Young, Oran R. 2009. ‘Whither the Arctic? Conflict or Cooperation in the Circumpolar North’. Polar Record 45 (1): pp. 73–82.

101

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Section II The Strange World of Economic Forecasting

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

5 The Interactional Foundations of Economic Forecasting Werner Reichmann

How Economic Forecasts are Produced Modern capitalist economies are future-oriented. To be successful in such an economy, economic actors produce so-called ‘imaginaries’ concerning possible future states of the economy and bring their actions in line with these imaginaries. They often experience the economic future as open and uncertain (Beckert 2016). This chapter asks how one specific kind of actor—the scientific economic forecaster—forms expectations and produces economic forecasts. Drawing on the notion of ‘epistemic participation’ (Reichmann 2013), the chapter argues that interaction between economic actors, economists, and policy-makers is the basis for the production of possible economic futures. At least temporarily, these futures are taken to be as if they were real representations of the future. The notion of interaction underlines that the formation of economic expectations takes place in a social environment. The production of economic futures is embedded in various networks consisting of different interaction partners. This interaction network sharpens economic forecasts in three ways: it brings to light novel imaginaries about the economic future; it ensures the forecasts’ social legitimacy; and it increases their epistemic quality. This argument will be illustrated by using empirical data from a case study involving economic forecasters in German-speaking countries. Their forecasts are a special case of the fictional expectations that actors hold about the economic future. Economic forecasts are made under the constraints of (and in alignment with the rules of) scientific work. They are expectations based on theoretical approaches and methods from economics.

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertain Futures

Two assumptions underlie this chapter. First, strictly speaking, we cannot predict the economic future with any accuracy at the macroeconomic level. Whereas some elements of the economy count as ergodic and are governed by known causal mechanisms and stable regularities, other parts of the economy are non-ergodic, transmutable, and undetermined (Davidson 1996). To date, no scientific discipline has developed an empirically robust method of accessing these non-ergodic and hence unknowable elements of the future and thereby replicating its success in collecting traces from the present or the past. The second assumption is that—irrespective of assumption one—economic agents go to great lengths to predict the future. This chapter asks how economic forecasters fill the gap between these two assumptions and what they do to compensate for the implications of the non-ergodic side of the economy. It does not need much inquiry to see that there are many fields in society that produce forecasts despite radical uncertainty—for example, forecasts about the weather (Daipha 2015; Monmonier 1999), about migration (Bijak 2011), about criminal careers (Seifert 2007), and above all about economies and financial markets. These various fields’ forecasting techniques differ and they depend strongly on whether there are fixed principles underlying a particular field. However, one thing is common to all forecasts: people spend large sums of money on them. This chapter starts by briefly introducing two theoretical concepts—‘mental time travelling’ and ‘foretalk’—that stem from different scientific fields. These concepts help us to understand how actors produce assumptions about the future by emphasizing the underlying interactional element of forecasting.

The Role of Interaction In his classic definition, Erving Goffman states that ‘[s]ocial interaction can be identified narrowly as that which uniquely transpires in social situations, that is, environments in which two or more individuals are physically in one another’s response presence’ (Goffman 1983, 2). In the twenty-first century, Goffman’s ‘body to body starting point’ (Goffman 1983, 2) of interaction must be reformulated because new technologies enable humans to interact and form social situations without being bodily co-present. Nevertheless, Goffman’s main point remains useful: interaction is a reciprocal social action of two or more individuals. Each interaction partner orients his or her actions towards the past, present, or future actions of the other partner(s). In Goffman’s understanding, interaction does not have to be reduced to communication in the sense of oral speech; although speaking is a common element of interaction, it is not a prerequisite. However, human interaction for him includes a consensus on a common immediate goal of action, a common 106

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Interactional Foundations of Forecasting

understanding of the situation, and it is embedded in a complex interaction order. It also plays a significant role in the process of producing expectations about possible futures.

Mental Time Travelling Thomas Suddendorf ’s work on the development of mental capacities in young children and in animals provides an interesting view on how humans interact to imagine the future. Initially, his approach may seem to be slightly a-sociological, but, on closer inspection, it acquires an interactional element. Suddendorf focuses on the question ‘What makes humans unique?’ In his book The Gap (Suddendorf 2013), he identifies eight main differences between humans and animals: one of them is that humans are able to do what he calls ‘mental time travelling’, that is, mentally form expectations and stories about the future. It is one of the fundamental human capabilities to imagine the future; and no other being in the world is able to ‘recall past episodes and imagine future events, including entirely fictional scenarios (such as the invention of an actual time machine)’ (Suddendorf 2013, 89).1 Suddendorf argues that ‘mental time travel into the past and mental time travel into the future are two aspects of the same faculty’ (Suddendorf 2013, 90). He refers to brain imaging studies that ‘have found that when participants are asked to recall past events and imagine future situations, the same areas of the brain . . . are involved’ (Suddendorf 2013, 94). In a second step, he argues that the human imaginative capacity, no matter whether about past or future events, is divided into three systems: a memory for how to do things (procedural memory), a memory for facts (semantic memory), and a memory for events (episodic memory). Episodic memory is not just responsible for our remembering past experiences; it also produces and imagines futures (Suddendorf 2013, 91). Humans use episodic memory in several ways to produce imaginaries. Of course, they use experiences from the past to produce futures. However, they are also able to imagine situations they have never experienced before. There is almost no limit to possible situations humans can imagine and, interestingly enough, humans can even evaluate these fictional situations (Suddendorf 2013, 95). The problem is that episodic memory is well known to be error-prone, no matter whether we use it oriented towards the past or the future (Suddendorf 2013, 98ff). But—and this is the more sociological aspect of Suddendorf ’s argument—humans have developed a unique technique to increase the

1 The claim that only humans have the preconditions for ‘mental time travelling’ is challenged by biologists and animal researchers, such as Clayton et al. (2008).

107

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertain Futures

quality of their episodic memory and their ‘mental time travels’, namely interaction. However, we have radically improved our chances of getting it right through a wonderfully effective trick: we share our plans and predictions with others. We can transmit our mental plays and reflections to audiences around us and, in turn, consider their thoughts. . . . We can learn from others’ memory and foresight, and listen to comments on ours. Indeed, we have a deep-seated drive to broadcast our minds and to read what is on the minds of others . . . And we have an extraordinarily effective way of exchanging our mind travels through language . . . Language is ideally suited for this mental exchange, and much of human conversation is indeed about past events (who did what to whom, and what happened next) and future events (what will happen to whom, and what we are going to do about it). By exchanging our experiences, plans, and advice, we have vastly increased our capacity for accurate prediction. (Suddendorf 2013, 99)

Suddendorf is an evolutionary psychologist. As such, he argues that both the ability to mind-travel and the ability to share real and fictitious stories about the past and the future with others interactionally increase the chance of survival. For him, it is an advantage in evolutionary competition to be able to create mental images for possible futures and thereby achieve better control of the future (Suddendorf 2013, 101–3).

Foretalk David Gibson (2011b; 2012) also emphasizes the interactional element of imagining the future and, by asking how this interaction is shaped in microsociological and conversational detail, he comes to two conclusions that enrich Suddendorf ’s argumentation. Gibson refers to interaction about possible futures using the term ‘foretalk’— a combination of forecasting and talk (Gibson 2012). He focuses on conversation and decision-making under extreme circumstances—in other words, on ‘talk at the brink’. As an example, he analyses the process of decision-making during the Cuban Missile Crisis in 1962, when President Kennedy and his top advisers had to decide within a couple of hours how to react to the Soviet Union’s installation of nuclear missiles on the island of Cuba (Gibson 2011a). In such extreme situations, people create possible future scenarios together by ‘foretalking’ (Gibson 2011a). This group foretalk shapes decisions through two mechanisms. First, foretalk brings to light possible futures that might not otherwise have been imagined. Thus, foretalk is an epistemic resource that enables us to produce new imaginaries of the future. Second, decision-makers anticipate the need to legitimate their decisions afterwards. The foretalk helps to justify decisions and improves their legitimacy. 108

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Interactional Foundations of Forecasting

Both Suddendorf and Gibson emphasize the interactional foundations of producing knowledge about the future. They show that the production of possible futures—for example, about economic development—does not take place in a social vacuum; it is not a purely mind-centred skill. It follows that concepts such as fantasy, creativity, mathematics, or cognition alone are not enough to provide an understanding of how fictional expectations are constructed. There are social and interactional aspects of producing economic futures that go beyond the ‘reserve stock of knowledge’ (Schutz 1967, 77) that individual people have accumulated and can access. Economic forecasts are based on an interactional process.

Interaction in Economic Forecasting The ways in which economic forecasters generate a common view by constantly negotiating their views with each other and with external groups— how they foretalk and how they exchange ideas from their mental time travels—can be elucidated empirically. Economic forecasters produce their forecasts using several channels of interaction as part of their epistemic process. To avoid misunderstandings, this chapter focuses on forecasting institutes in German-speaking countries, which operate quite differently, for example, from forecasting institutes in the United States. There are national differences between forecasting systems and the political uses of the forecasts, especially between the United States and Europe (Campbell and Pedersen 2014). In general, one could say that American forecasters are more commercially oriented whereas European forecasters operate closer to the state (Friedman 2009, 2014).

Economic Forecasters in German-Speaking Countries Numerous organizations publish economic forecasts: banks, financial institutes, rating agencies, academic research units, and so on. The institutes examined in this chapter share at least five common characteristics. First, they earn their money exclusively by producing economic expertise (for example, forecasts) and do not use forecasts to sell something else. Banks, for example, also produce forecasts, but they use them to sell other services or as part of their customer relationship management. Second, the institutes are in a way ‘semi-official’: their work is partly financed by the government and it is institutionalized within the policy-making process (Reichmann 2009). Third, they are ‘independent’ in a specific way: they do not belong to any political movement, company, interest group, or political party and have neither commercial nor political aims. And fourth, the forecasting institutes’ 109

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertain Futures

members consider themselves to be part of academia: they have an identity as academic scholars and do things only scholars do (for example, giving courses at universities, earning their Habilitation,2 and so on) and their practices stick to the rules of economics (Evans 1997, 408). However, despite their academic identity, the vast majority of the forecasting institutes analysed in this chapter are organized outside universities. In German-speaking countries the growth rate of the Gross Domestic Product (GDP) stands at the centre of every economic forecast and, especially in public discussions, economic forecasts are often reduced to it. The forecasts contain between seventy and 700 pages and are summarized in short press releases, showing the main economic indicators and a few points summarizing the main messages. The institutes publish economic forecasts two to four times a year and present them to the public at press conferences.3

Interaction and Econometrics Textbooks show different ways of producing economic forecasts (for example, Döhrn 2014; Tichy 1994). They differ mainly in terms of whether forecasters have more trust in numbers, quantitative data, mathematics, and econometric models or whether they rely more on qualitative data gathered from representatives of the economy (Evans 1997, 1999; see also McNees 1990). In practice, forecasters never rely solely on calculation. Econometric models are used merely as a starting point. And these models are increasingly taking a back seat in the process of manufacturing a forecast. In fact, econometric models play a fairly minor role in producing economic forecasts, and the interviewees for this study agreed with Robert Evans’s claim that ‘macroeconomic models support forecasting activity, but do not actually produce forecasts’ (Evans 1997, 426). Instead of econometrics, the more important parts of the forecasting process consist of various forms of interaction with various interaction partners. Interaction can be either informal or more institutionalized (see also Reichmann 2013, 861–7), and the interaction includes both internal partners (such as colleagues from their institute) and external ones (such as academic 2 In German-speaking countries, the Habilitation is an academic degree obtained after a doctorate. In some scientific disciplines, it is a requirement for becoming a full professor. 3 The data used in this chapter were collected between 2004 and 2012 and consists of thirtyfive qualitative interviews (approximately 30–90 minutes) with economists directly engaged in producing the forecasts and with users of the forecasts from national, regional, and local governments, special interest groups, and labour unions. In addition, the author spent some time at different forecasting institutes taking notes and has collected a large volume of documents from all forecasting institutes in the German-speaking countries. The interviews were conducted in German and were translated by the author. Quotes from the interviews are marked ‘INT’, followed by the number of the interview and the line number.

110

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Interactional Foundations of Forecasting

economists and representatives of ‘the economy’). Forecasters have developed numerous formal and informal interaction channels and a permanent communication flow enabling them to contact those who represent, in one way or another, ‘the economy’. They build formal and informal platforms where they meet these representatives to gather data and information and thus jointly produce an image of the economic future. Economic forecasters supplement the human capacity for ‘mental time travelling’ to imagine possible futures by using the ‘trick’ (Suddendorf 2013) of sharing their predictions with others to obtain information about their respective views of, and alternative perspectives on, the future. Furthermore, forecasters ‘foretalk’ (Gibson 2011b) with selected interaction partners in several ways, thereby ensuring that economic forecasting does not take place in a social vacuum. This chapter emphasizes three reasons why forecasters engage in foretalking with various representatives of economics and the economy, coming under the headings of novelty, legitimacy, and quality. First, foretalking enables forecasters to entertain possible futures and spot emerging developments they would have missed without the foretalk. They use interaction as a resource for novel imaginaries. Second, foretalk increases the social legitimacy of the forecasts in the sense that they are more likely to be believed. As Douglas Holmes (2013) shows, central bankers develop strategies to increase their legitimacy by intensive communication with the public and the economy. Holmes’ argument is parallel to the way in which forecasters increase the legitimacy of their forecasts by involving those who use forecasts in the process of producing them. Users become co-producers of forecasts and thereby have less reason to reject them. Third, foretalk improves what could be called forecasts’ epistemic quality. Foretalk helps to bridge the gap between the knowable and unknowable elements of economic futures by providing (highly) unstandardized data, including judgements that econometric models could not process. The comprehensive interaction process may not make economic forecasts more accurate in a numerical sense. Nevertheless, it increases the range of knowledge about the intentions and assumptions of economic and political actors and therefore builds a more reliable basis for creating forecasts.

Patterns of External Interaction The forecasters are embedded in a network that includes several groups of interaction partners, such as other economists from universities, entrepreneurs, policy-makers, and members of the government and the state administration. This interaction network is a constitutive part of the epistemic process of economic forecasting. The members of this network are transformed 111

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertain Futures

from ordinary interaction partners into co-producers of the economic forecasts. This network is called here an ‘epistemic network’ because it is an active part of the forecasters’ epistemic process. The forecasters do not just interview, survey, or observe the others in the network; they want them to actively co-produce the forecasts. In this sense, forecasters give them the opportunity to participate in the epistemic process of forecasting—a form of ‘epistemic participation’ (Reichmann 2013). In the German-speaking countries, this epistemic network includes a lively interaction between economic forecasters from different institutions. The forecasting institutes may follow conflicting scientific paradigms and they compete for funding, but they frequently interact and cooperate, both formally and informally. On the more formal side, the institutes’ members attend meetings and workshops to discuss economic topics; they talk in advance about their views on current economic developments; they meet at conferences, political hearings, and public discussions. On the more informal side, the forecasters know each other from a variety of activities and relationships developed outside their formal work, whether from their time together as university students, previous cooperations, co-authoring articles, or spending leisure time together. Within the community of forecasters, all forecasters have individually formed networks of ‘foretalkers’ (see Gibson 2011b) and personal sources of information. Furthermore, economic forecasters are part of a network of scholars working at academic institutions. They hold lectures and seminars at universities; they work on common research projects; and they co-author papers and books with researchers from universities. These close ties to universities not only sustain the forecasters’ identity as scientists (Evans 1997, 408), but also give them the chance to exchange ideas, share new insights, and discuss problems; or, in Gibson’s (2011b) words, to ‘foretalk’ with academic economists. As Evans (2007, 691) argues, these ‘professional networks’ are the source of certain types of expertise that help overcome the uncertainties of econometric models and allow judgement between models. Exchanging ideas with colleagues is something familiar to most scientists. But the forecasters’ epistemic networks include not just other economists who have more or less similar knowledge they can bring into the foretalk. In particular, their external networks include policy-makers and business representatives. The policy-makers with whom they interact—for example, members of government units, federal banks, interest groups, lobby organizations, labour unions, and social partners, and so on—provide a different stock of knowledge and a fresh view on ‘the economy’. This part of the external interaction network enables forecasters to interact with ‘the economy’ to gather information about ‘the economy’s’ plans. In practice, forecasters are able to interact only with a limited number of representatives of ‘the economy’. Still, for the forecasters, their interaction partners 112

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Interactional Foundations of Forecasting

are like intermediaries for ‘the economy’. When forecasters talk about their network, they rhetorically reify ‘the economy’ and utter sentences such as: ‘It is really important to speak with the economy.’ Of course, they are aware that they cannot really speak to ‘the economy’ as such, but interpret their intermediaries as windows on it. Forecasters describe this part of their network as the most important one. Indeed, they say it is more important than econometric models or academic conferences. It is a place where those who forecast economic developments meet to foretalk with those who create economic policy, shape the economic policy frame, and actually make economic decisions. And it is a place where two quite different groups of mental time travellers exchange their imagined futures. The business representatives in their networks (such as CEOs, businessmen, and industrial lobbyists) consider forecasters to be scientific consultants, conducting studies to answer their questions. But forecasters also give informal advice that helps the business representatives get an idea of what others think about recent economic developments and of the expectations in other economic sectors. Forecasters allow them to leave the ‘fog of uncertainty’ (INT 10, 454) and get a ‘bird’s eye view’ (Gilbert and Jaszi 1954, 52) on the economy. For that purpose, several economic forecasting institutes conduct regular panel studies. To obtain information about business representatives’ views of the economic future, they gather data from certain groups—for example, financial experts, CEOs, purchasing managers, port executives, and so on—at specific time intervals using standardized questionnaires. This process can also be conceptualized as one part of an ongoing (standardized) interaction between various groups of mental time travellers. The integration of this external group works in many ways. During the forecasting process, the forecasting institutes first autonomously produce a forecast, which is called a ‘draft forecast’ (field term). This first step is dominated by applying econometric models, which are analysed by Evans (1997, 1999) in detail. After that, the continuous formal and informal discussions with the groups start. With an eye to recent problems on the political agenda, forecasters contact specialized policy-makers to discuss the draft forecast, exchange views regarding ongoing economic developments, and explore the perceptions of the members of the policy-maker network. This process is generally not standardized, and it is permanently ongoing. As one member of a special interest group puts it: There are consultations, there are even continuous consultations between us and these forecasting institutes. Of course, we do not influence the results; they are their own. But within this process of consultation, actually we are not the only ones participating in this process: the collective bargaining partners and the most

113

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertain Futures important ministries are involved. In most cases, this is an ongoing process, but one that practically comes to a head when the forecasts are actually produced. In fact, they ask us to give input, to make them more true. Actually, our insights, those of the economic chambers, and those of the Treasury, Federal Reserve Bank, perhaps Ministry of Economic Affairs, are extremely highly valued by the forecasters. Not to say that the insight of the others is less valued, labour unions and so on, but we do indeed have our own data, and we are very liberal with this information and we give it to the forecasters, and when they see that our insights are contrary to their forecast or their capital-investment tests, they have to think of a response. Well, this is how it works. It is an ongoing process that obviously comes together four times a year. But I think that the real value lies in the ongoing consultations. In the official meeting, to be honest, they tell us the forecast, and those of us who already know it and were somehow consulted during the preparations nod and the others watch, that’s it. (INT 17, 317)

Before the forecasts are presented to the public, several meetings take place. They are formal in comparison to the more informal talks described in this section. At these meetings, the final draft forecasts are discussed with a group of policy-makers. Normally, those who participate in these meetings are also involved in the prior talks. The complete preparation of a forecast takes about two to three weeks, but the interaction and the foretalk take place continuously. The mental time travellers keep in permanent contact and ensure that information on economic policy plans, on the political climate, and even on shifts in the economic paradigm are exchanged continuously. We should not misinterpret this dense epistemic interaction network of forecasters and policy-makers as purely a question of political power. Although the interests of particular groups and organizations may influence forecasts in the process of epistemic participation, there is no evidence that ideologically suitable forecasts can be simply ordered by policy-makers. What is more important for the question of how forecasts for the uncertain (and nonergodic) parts of the economy are made is that it is really the economic forecasters who benefit most from being in a process of epistemic networking with policy-makers. The impact of these contacts with political actors on the epistemic process of economic forecasting cannot be overstated: they bring to light new imaginaries about the future; they socially legitimate the forecasts; and they increase the forecasts’ quality in the sense that they are based on better information and more diverse perspectives.

Patterns of Internal Interaction Another part of the epistemic process is much more closed and takes place inside the forecasting institutes. This process of internal interaction enables 114

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Interactional Foundations of Forecasting

different forecasters to harmonize their mental time travels and involves another type of foretalk. There are five discrete internal roles the forecasters have to play.4 Each role is responsible for a specific part of what they call ‘the economy’. One examines public finance and the government’s budget; another focuses on the labour market; a third looks at fiscal policy and inflation; and a fourth studies foreign trade. The fifth role is to integrate the data, the arguments, and the information collected by the other economists: the economist concerned is the one responsible for the national economy and is the ‘single person’ also found in a group of econometric modellers—the one who ‘integrate[s] the disparate inputs and make[s] judgments about the wide range of factors that have impacts on the national and international economy’ (Evans 2007, 688). At the outset, each of the five economist roles individually produces a forecast on their respective topics using both quantitative models and additional information gathered during the external interaction process. Each of them produces calculations, creates interpretations, and thinks about the assumptions underlying these results. In this part of the forecasting process, each forecaster tries to ‘get a feeling for what the present development may cause at the end of the year’ (INT 23, 223; emphasis added). This brings to light that mental time travelling is not just a cognitive but also an emotional activity. After the phase of working alone on the first forecasts, a further interaction process starts. The five types of internal forecasters meet to discuss their individual results, exchange data, discuss their aggregate-related forecasts, and describe and justify their assumptions. They interact and foretalk with each other and try to align their forecasts and harmonize their mental time travels. Their aim is to create a forecast with no internal contradictions. One of the forecasters describes this step in detail: And if someone says ‘Okay, this doesn’t fit here and there’, we just start again and take information from the others and go back to our offices and we begin to recalculate—we cut off the corners to make the calculations fit; we call it Rundrechnung. (INT 25, 408)

The notion of Rundrechnung is an interesting one, as it shows the iterative character of the interaction process. It is barely translatable, but a literal translation may be ‘round-calculation’ or ‘circle-calculation’. It summarizes the process of several re-adjustments of the common forecast until it is a smooth and rounded forecast. This notion describes accurately how economic forecasters adjust, re-adjust, and re-re-adjust their results until they have combined a 4 The teams in the institutes vary and the description provided here is an ‘ideal type’ generalization.

115

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertain Futures

‘rounded image’ of the future. To them, this means that the components of the forecast fit together, that the forecast appears theoretically harmonious, and that there are no internal contradictions, no inconsistent corners, in the image it provides. For about two to three weeks, the forecasters continue to work individually on their special topic. They then meet again with the others to produce a new forecast that is in line with the views of the other four types of internal forecasts. The process of Rundrechnung is based mainly on social interaction and can be understood as the repeated foretalk of mental time travellers, each with a different angle on the economy. Every economist is a specialist in one part of the economy and experiences it from a specific perspective. They come together to produce interactionally a common view that could not be produced individually. This clearly delineates that the forecasters are not passive observers of the economy but active participants in constituting the ‘knowledge’ they create.

Interaction in Formal and Informal Frameworks Economic forecasters are thus embedded in two networks: one includes the external group (scholars, policy-makers, and business representatives) and the other the internal group (forecasters within the forecasting institute with different responsibilities). These epistemic networks provide the social space in which the interactional element of economic forecasting happens. There are various ways in which the interaction process within the different external and internal networks takes place: it can be characterized as either formal or informal, and it is becoming more and more globalized. The foretalk between economic forecasters and others takes place over a wide variety of platforms: in formal meetings and arranged conferences; but also in accidental chats or in impromptu discussions at various occasions that range from formal official hearings to informal sports activities. The formal gatherings are well organized and accurately documented in publications (see, for example, Sachverständigenrat zur Begutachtung der gesamtwirtschaftlichen Entwicklung 2008, iv–v). The informal part of the foretalk is based on the forecasters’ individual ties and contacts with other economists and policy-makers. One forecaster describes how she informally gathers and exchanges information within the research institute like this: Well, we discuss. For example, we have a facility in our institute called the café. And we have lunch in-house and, after lunch, we meet each other or we go for a walk here in [name of a park next to the institute]. But after lunch, we meet to have a coffee and this is the real discussion round. There one is informed; there is

116

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Interactional Foundations of Forecasting someone telling us what is happening in politics and the next one perhaps what he is currently working on. Actually, this coffee is a very, very important institution. Normally every academic sits in their own room, but for a coffee everyone meets and hears what is going on. This is, well, I think this is very important. (INT 10, 337)

These informal contacts have grown over a long period of time and have their point of origin in past cooperations or common research activities. Sometimes informal interaction situations are set up spontaneously. For example, when the author was invited to take part in a public discussion at an economic forecasting institute, he suddenly became part of the interaction network. An extract from the field note describes the situation: After the panel discussion had ended, I had to wait because [head of one forecasting institute] wanted to show his new office to his colleague before going to a restaurant together. I stood around with [director of a European central bank], [head of the economics section in the ministry of finance], [head of the forecasting section of a forecasting institute], [another head of the forecasting section of another forecasting institute] and two PhD-candidates. It was one of those situations that is not easy to manage—one that everyone who is often at conferences knows. What should we talk about? . . . It took about 30 seconds to agree interactionally on our topic: How will the economy develop for the rest of the year? And how will the Greek crisis develop? . . . Everyone brought in their respective view and a lively discussion about the economic future was set in motion. . . . What impressed me most was that all participants (except me) could recall a lot of quantitative data by heart. . . . Suddenly, I became part of an epistemic participation situation. (Fieldnote, 20 October 2015)

No one planned this situation. The interaction was conducted in a highly informal tone; there was a lot of laughter about this and that, and a good part of the interaction situation was chitchat. This is a typical example of an informal emergence of epistemic participation. Such situations have many advantages for the forecasters, putting them at the heart of economic forecasting’s epistemic work: they are a fast way to exchange information; they are based on the trust produced in face-to-face-situations; and they involve the harmonization of the mental time travels. The social settings of the ‘backstage’ of economic forecasting are characterized mainly by this sort of informal contact. Economic globalization also affects the nature of the interaction processes described here. The number and different kinds of mental time travellers with varying knowledge at hand have increased enormously. This makes forecasting more difficult. Forecasters have had to adapt their foretalking strategies and shift them to the global level. For example, there are new associations, institutionalized platforms, and common research projects 117

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertain Futures

where international forecasters meet both each other and representatives from the economy and policy-makers (for more detail, see Reichmann 2018, 98–100). The forecasters’ interaction networks have gone global and have transformed the epistemic process into big business.

Beauty Contests, Herding, and the Dominant Mood The dominance of the social in the analysis of the forecasting process raises critical questions about the forces of conventionality and social conditioning: how can forecasters form expectations independently in such a dense interaction network? How do they avoid falling into a version of what Keynes (1936) called the ‘beauty contest’? Do forecasters rely on average expectations when they produce their own forecasts? Or, to stretch this argument, do they consider what the average may think the average may think about what the mean forecast is? These concerns have in common the assumption that the social interaction and the observation of others’ expectations may not only influence but also ‘distort’ the future imaginaries that any particular forecaster will entertain. The fieldwork on which this chapter relies shows that sociality and social interaction is an inevitable condition of creating economic forecasts. Confronted with a high degree of uncertainty in many elements of the economy, forecasters develop the social technique of epistemic participation to handle this uncertainty—that is, to bridge the gap between the elements of the economy that are ergodic (and therefore capable of being modelled and predicted) and those that are non-ergodic (and hence not amenable even to probability forecasting). Nevertheless, within the interaction order and institutional structures of the process of epistemic participation, there can occur social forces that are ungovernable and may unintentionally influence the forecasts’ outcome. At least three of these social forces are theoretically well known: Keynes idea of ‘failing conventionally’, herd behaviour, and the ‘dominant mood’. One of Keynes’ most popular metaphors—as already mentioned—is that of the beauty contest (Keynes 1936, 156), which casts light on one of the problems that may arise when conventions replace substantive knowledge on markets—a phenomenon that emerges especially under the condition of radical uncertainty. Keynes likened the stock market to a beauty contest in which the winners are those who anticipate the average opinion about who is the most beautiful. Participants in such a contest, that is, do not follow their own knowledge or opinion about beauty; instead, they choose the face they think the crowd expects will be judged the prettiest by the average participant. So the question for the argument in this chapter is this: are the forecasters analysed in this chapter in danger of falling into such a beauty-contest trap of 118

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Interactional Foundations of Forecasting

tending, or having an incentive, to produce forecasts that are close to other forecasters’ predictions? The analogy between the market actors in Keynes’ work and the economic forecasters analysed in this chapter does not fit perfectly. At least for the German forecasting institutes, the market metaphor is of limited application: this is closely connected to the specific kind of academic independence and freedom they enjoy from both markets and policy-makers. History shows that forecasting institutes do not automatically lose their financial funding when they produce ‘wrong’ forecasts5, even when that happens over a long time period. The more important market for forecasting institutions is the academic one, in which they compete for what Bourdieu (1975) calls ‘scientific capital’, which is primarily reputation. Keynes’s (1936, 158) assertion that ‘it is better for reputation to fail conventionally than to succeed unconventionally’ may suggest by analogy that scientific capital is at risk when forecasters create an ‘outlier forecast’, one that deviates from the forecasts of other institutions. Thus, one of the unintended consequences of the dense sociality in the forecasting process may be the social pressure that can lead to conventional forecasts. In the economic forecasting literature, the question of whether interaction and social embeddedness lead to more consensual results and a tendency to ‘fail conventionally’ is raised alongside the notion of the ‘herd’. Herding is, together with political bias and conservatism, often discussed as the ‘judgmental bias’ of forecasting. For example, Grömling (2002), who is a forecaster himself, states that economic forecasters are not at all isolated from each other. They are embedded in an information network with other forecasters and thus they are in danger of herding. He argues that especially in phases characterized by high uncertainty, the phenomenon of ‘herding’ emerges, which is in line with Keynes’s argument. As in the beauty contest, forecasters ‘herd’, when they follow their forecasting colleagues and tend towards a consensus instead of considering the information they themselves could get from econometric data and from epistemic participation. Grömling (2002) speculates that only well-established forecasters are in a position to break out of the herd. Being a novice increases the risk to ‘fail conventionally’. The empirical studies on ‘herding’ show highly inconclusive results. For example, Grömling (2002, 14) empirically tests the herding hypothesis for the German GDP forecasts in 2001, but cannot find evidence for it. Batchelor and Dua (1992) test whether US forecasters herded when forecasting real GDP, unemployment, inflation, and interest rates in the 1980s. They found evidence that forecasters are not consensus- but ‘variety-seeking’ (Batchelor and Dua 5 Though there are several ways to evaluate the quality of economic forecasts, this chapter uses the most common one, ex post analysis.

119

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertain Futures

1992, 169) in their forecasts. The same has been found for forecasters on equity markets who are even ‘anti-herding’ (Pierdzioch and Rülke 2012). In contrast, Gallo et al. (2002) identified that forecasters do orient their forecasts to be in line with the expressions of other forecasters and do tend to follow the ‘herd’. To complete the empirical confusion, there is an extensive literature showing that financial market forecasters display strong herding tendencies. (For an overview of this literature, see Rangvid et al. 2013.) One of the problems posed by these divergent empirical results is that the samples, the statistical tests used, and the analysed time periods are different and impede the studies’ comparability. However, the fact that herding is a highly discussed topic shows that it is a theoretical threat that forecasters themselves are well aware of. Whereas herding is defined by the forecaster’s orientation to other forecasters, there is another distinctive term that may unintentionally influence forecasts, the so-called ‘dominant mood’. This is a field term used in an interview by an experienced and well-established forecaster and university professor: Forecasts are strongly conditioned by sociological factors—the mood at the moment. It is terribly hard to do something against the present mood. Funnily enough, in some moments all, or nearly all, forecasters are in the same mood and, thus, have the same opinion. (INT 12, 691)

The ‘dominant mood’ goes beyond the group of forecasters and also includes the mood in policy-making, within the academic community of economics, and in the public—for example, expressed in journalists’ commentary pages. Forecasters describe it as difficult to forecast against the dominant mood. They say that deviant expectations about the future are stigmatized and sanctioned and labelled deviant behaviour. As another forecaster puts it: I think no one wants to be subject to Cassandra’s curse.

(INT 23, 762)

The mythological figure of Cassandra was often mentioned in interviews with forecasters. In this quote, the figure of Cassandra is used to express the idea that the range of possible forecasts is—at least if they are to be seen as credible—not limitless but bounded by the ‘dominant mood’ (see Introduction to this volume, p. 12, for a similar point). The notion of the beauty contest, the phenomenon of herding, and the forces of the ‘dominant mood’ are inherent in the interactional foundations of forecasting, which may lead to unintended outcomes. As the forecasting process is full of sociality and interaction, there can occur hidden and unintended social pressure from both the internal and the external groups involved in the epistemic participation process. However, it seems that the advantages of the interactional foundations of forecasting exceed the possible costs that such unintended and ungovernable social forces may cause. 120

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Interactional Foundations of Forecasting

Conclusion Producing economic forecasts involves social interaction. Expectations about economic futures are the result neither of simply feeding econometric models and rational computation, nor of cognitive processes in social isolation that disgorge a legitimate result everyone is happy with. Rather, imagining economic futures is based on various interactions between different interaction partners. This interactional process enriches and sharpens the imaginaries of the economic future by increasing their responsiveness to novelty, their social legitimacy, and their quality. Social interaction is first of all a resource for economic forecasters to discover novel imaginaries of the future they would otherwise have missed. It also boosts the social legitimacy of imaginaries or forecasts. Forecasters are confronted with the fact that the non-ergodic character of the economic future increases the need to legitimate imaginaries about the future. By including as many relevant actors as possible, the interaction process helps to justify forecasts (and the decisions that are deduced from this knowledge), even if they turn out to be ‘wrong’ afterwards. Finally, the epistemic participation involved in foretalk improves the epistemic quality and content of economic forecasts. The forecasts are more useful when they correctly reflect the likely plans of policy-makers and other key economic actors. Hayek (1937) argued that no one standing at the centre of an economy can forecast the unknown future or even know all the existing relevant information and beliefs of economic actors because they are highly decentralized (Bronk 2013). Forecasters use their interaction network to help counter this critique and collect as much decentralized information as possible—based on the diverse plans and perspectives of various actors. In this way, the network allows forecasters to overcome some of the problems of knowledge facing all economic actors by enabling them to access the beliefs of a myriad of key players and to pick up signals of emerging trends and the novel imaginaries entertained by actors that have a significant chance of performing the future. (A similar point is made by Holmes in his chapter in this volume [p. 189] when he discusses the role that a network of agencies plays in the formation of Bank of England policy.) Forecasts remain imaginaries, of course, but they are rationally improved by interaction with various actors. The imaginative process of placing ‘oneself in the shoes of another’ (Beckert and Bronk, this volume, p. 3) is facilitated by social interaction and enables forecasters to align their own imaginaries with the imaginaries of others likely to have an impact on economic development and shape the economic framework. Interaction improves forecasters’ ability to see the world through the eyes of others, understand their plans, and discover their influential expectations. 121

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertain Futures

Bibliography Batchelor, Roy, and Pami Dua. 1992. ‘Conservativism and Consensus-Seeking Among Economic Forecasters’. Journal of Forecasting 11 (2): pp. 169–81. Beckert, Jens. 2016. Imagined Futures: Fictional Expectations and Capitalist Dynamics. Cambridge, MA: Harvard University Press. Bijak, Jakub. 2011. Forecasting International Migration in Europe: A Bayesian View. Dordrecht: Springer. Bourdieu, Pierre. 1975. ‘The Specificity of the Scientific Field and the Social Conditions of the Progress of Reason’. Social Science Information 14 (6): pp. 19–47. Bronk, Richard. 2013. ‘Hayek on the Wisdom of Prices: A Reassessment’. Erasmus Journal for Philosophy and Economics 6 (1): pp. 82–107. Campbell, John L., and Ove K. Pedersen. 2014. The National Origins of Policy Ideas: Knowledge Regimes in the United States, France, Germany, and Denmark. Princeton, NJ: Princeton University Press. Clayton, Nicola S., Sérgio P.C. Correia, Caroline R. Raby, Dean M. Alexis, Nathan J. Emery, and Anthony Dickinson. 2008. ‘Response to Suddendorf & Corballis (2008): In Defence of Animal Foresight’. Animal Behaviour 76: pp. e9–e11. Daipha, Phaedra. 2015. Masters of Uncertainty: Weather Forecasters and the Quest for Ground Truth. Chicago, IL: University of Chicago Press. Davidson, Paul. 1996. ‘Reality and Economic Theory’. Journal of Post Keynesian Economics 18 (4): pp. 479–508. Döhrn, Roland. 2014. Konjunkturdiagnose und -prognose: Eine anwendungsorientierte Einführung. Berlin: Springer. Evans, Robert. 1997. ‘Soothsaying or Science? Falsification, Uncertainty and Social Change in Macroeconomic Modelling’. Social Studies of Science 27 (3): pp. 395–438. Evans, Robert. 1999. Macroeconomic Forecasting: A Sociological Appraisal. London: Routledge. Evans, Robert. 2007. ‘Social Networks and Private Spaces in Economic Forecasting’. Studies in History and Philosophy of Science Part A 38 (4): pp. 686–97. Friedman, Walter A. 2009. ‘The Harvard Economic Service and the Problems of Forecasting’. History of Political Economy 41 (1): pp. 57–88. Friedman, Walter A. 2014. Fortune Tellers: The Story of America’s First Economic Forecasters. Princeton, NJ: Princeton University Press. Gallo, Giampiero M., Clive W.J. Granger, and Yongil Jeon. 2002. ‘Copycats and Common Swings: The Impact of the Use of Forecasts in Information Sets’. IMF Staff Papers 49 (1): pp. 4–21. Gibson, David R. 2011a. ‘Avoiding Catastrophe: The Interactional Production of Possibility during the Cuban Missile Crisis’. American Journal of Sociology 117 (2): pp. 361–419. Gibson, David R. 2011b. ‘Speaking of the Future: Contentious Narration during the Cuban Missile Crisis’. Qualitative Sociology 34 (4): pp. 503–22. Gibson, David R. 2012. Talk at the Brink: Deliberation and Decision during the Cuban Missile Crisis. Princeton, NJ: Princeton University Press.

122

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Interactional Foundations of Forecasting Gilbert, Milton, and George Jaszi. 1954. ‘National Product and Income Statistics as an Aid in Economic Problems’. In Readings in the Theory of Income Distribution: Selected for the American Economic Association, edited by William Fellner and Bernard F. Haley, pp. 44–57. 2nd impr. London: Allen & Unwin. Goffman, Erving. 1983. ‘The Interaction Order’. American Sociological Review 48 (1): pp. 1–17. Grömling, Michael. 2002. ‘Konjunkturprognosen: Methoden, Risiken und Treffsicherheiten’. Iw-Trends 2: pp. 1–17. Hayek, Friedrich A. von. 1937. ‘Economics and Knowledge’. Economica 4 (13): pp. 33–54. Holmes, Douglas R. 2013. Economy of Words: Communicative Imperatives in Central Banks. Chicago, IL: University of Chicago Press. Keynes, John M. 1936. The General Theory of Employment, Interest and Money. London: Macmillan. McNees, Stephen K. 1990. ‘Man vs. Model? The Role of Judgment in Forecasting’. New England Economic Review July/August: pp. 41–52. Monmonier, Mark. 1999. Air Apparent: How Meteorologists Learned to Map, Predict, and Dramatize Weather. Chicago, IL: University of Chicago Press. Pierdzioch, Christian and Jan-Christoph Rülke. 2012. ‘Forecasting Stock Prices: Do Forecasters Herd?’. Economic Letters 116 (3): pp. 326–9. Rangvid, Jesper, Maik Schmeling, and Andreas Schrimpf. 2013. ‘What Do Professional Forecasters’ Stock Market Expectations Tell Us about Herding, Information Extraction and Beauty Contests?’ Journal of Empirical Finance 20: pp. 109–29. Reichmann, Werner. 2009. ‘Dauerhaft, gefestigt und erfolgreich: Kommunikative Einigkeiten zwischen Konjunkturforschung, Öffentlichkeit und Politik’. Geographische Revue 11 (2): pp. 39–45. Reichmann, Werner. 2013. ‘Epistemic Participation: How to Produce Knowledge about the Economic Future’. Social Studies of Science 43 (6): pp. 852–77. Reichmann, Werner. 2018. Wirtschaftsprognosen: Eine Soziologie des Wissens über die ökonomische Zukunft. Frankfurt am Main: Campus. Sachverständigenrat zur Begutachtung der gesamtwirtschaftlichen Entwicklung. 2008. Die Finanzkrise meistern—Wachstumskräfte stärken. Jahresgutachten 2008/2009. Wiesbaden: Sachverständigenrat zur Begutachtung der gesamtwirtschaftlichen Entwicklung. Schutz, Alfred. 1967. The Phenomenology of the Social World. Evanston, IL: Northwestern University Press. Seifert, Dieter. 2007. Gefährlichkeitsprognosen: Eine Empirische Untersuchung über Patienten des psychiatrischen Maßregelvollzugs. Darmstadt: Steinkopff. Suddendorf, Thomas. 2013. The Gap: The Science of What Separates Us from Other Animals. New York: Basic Books. Tichy, Gunther. 1994. Konjunktur: Stilisierte Fakten, Theorie, Prognose. Berlin: Springer.

123

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

6 Escaping the Reality Test How Macroeconomic Forecasters Deal With ‘Errors’ Olivier Pilmis

Introduction ‘Why did no one see it coming?’ H.M. Queen Elizabeth II reportedly asked during a visit to the London School of Economics in November 2008.1 The question echoes the criticisms economic forecasters often face for not having ‘anticipated’ the emergence of a particular economic crisis. The 2008 subprime mortgage crisis and its aftermath were no exception and gave birth to an important literature that takes the inaccuracy of forecasts as its starting point. Springing from sociology (Fligstein et al. 2017), economics (Galbraith 2014), popular science (Orrell 2010), or the media (Turin 2015), critical assessments of macroeconomic forecasters’ records rejuvenate the well-known saying: ‘economists are experts who know tomorrow why the things they predicted yesterday didn’t happen today’.2 N. N. Taleb’s best-selling book The Black Swan goes further and calls ‘predictions’ a ‘scandal’ (Taleb 2007, 137–64). Whatever their sources, criticisms of forecasting usually present a mirror image of praise for whistleblowing: economists and forecasters are blamed for being overly optimistic and consequently failing to warn of future crises. These criticisms endorse a systematic comparison between ‘what actually happened’ and ‘what had been predicted’ as the only ‘reality test’3 possible to warrant forecasts’ accuracy and eventually pronounce them ‘right’ or Chris Giles, ‘The Vision Thing’, Financial Times, 26 November 2008, p. 13. The success of this saying makes it difficult to trace its precise origins. 3 Although tackling issues of legitimacy and critique, the expression ‘reality test’ does not, here, strictly match ‘pragmatic’ conceptions (Boltanski and Thévenot 2006; Boltanski 2011). 1 2

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

How Forecasters Deal With ‘Errors’

‘wrong’.4 Repeated experiences seem to have brought clear-cut evidence that predicting future economic reality is impossible. The mere existence of an activity such as forecasting therefore turns out to be puzzling: how do forecasters manage to fend off criticisms and persuade themselves and others of their own credibility and that of their activities? This attention to the relationship between beliefs and practices on one hand and ‘reality’ on the other is reminiscent of early anthropologists’ discussion of magic, religion, and science, all of which have been defined as belief-based practices. Building bridges between anthropology, economics, and sociology, Durkheim’s enduring conception of belief offers an understanding of the persistence of forecasting despite recurring ‘errors’. The Durkheimian ground of economic conventions and fictions has long been emphasized in both heterodox economics and sociology: while Orléan (2014, Chapter 5) has shown that following Keynes’s track eventually leads back to Durkheim, Beckert (2016, 192–204) demonstrates how Durkheim’s notion of totem and mana (Durkheim 1912 [1965]) sheds light on the kind of fictionality that consumption encapsulates. However, the issue of whether beliefs are still held in adverse situations encourages us to turn to Henri Hubert and Marcel Mauss’s seminal theory of magic. As disciples of Durkheim, they claim that the authority of magic and magicians originates in collective forces, so much so that beliefs and confidence in magic are a priori rather than evidence-based. A priori confidence in magic arises from the collective need to find explanations to phenomena that would otherwise remain inexplicable. Thus, magical beliefs stand firm when challenged. Moreover, the ‘will to believe’ is the origin of magicians’ belief in magic: at the very least, they keep faith in their own ability to perform magic because they believe in the magic of others (Hubert and Mauss 1902 [1972]). In the case of macroeconomic forecasting, a similar somewhat functionalist argument posits that forecasting has more to do with coordination than with prediction. When uncertainty prevails, actors’ decisions are necessarily anchored in ‘fictions’, requiring actors a priori to ‘suspend disbelief ’ and adopt an ‘as if ’ convention. When the future has yet to be created and cannot be known at present (Shackle 1972 [1992]; and the introduction to this volume), economic actors can base their action only on ‘fictional expectations’—that is, ‘pretended representations of a future state of affairs’ (Beckert 2013, 226) drawing upon actors’ imagination. In this perspective, fictional expectations may be supported by ‘instruments of imagination’, such as forecasts or economic theories, but a ‘right’ forecast is one that is shared within a large In order to underline that a statement will here only be considered ‘erroneous’ with respect to how actors define it, the nouns ‘error’ and ‘mistake’, as well as the adjectives ‘right’, ‘wrong’, and so on will be kept in quotes throughout the chapter. 4

125

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

community, thus enabling economic action. The ex post judgement of the accuracy of ex ante forecasts does not matter, as forecasts are judged according to their practical credibility at the time made (Beckert 2016). The notoriously poor track record of macroeconomic forecasting does not put its legitimacy at risk, because actors need forecasts anyway to design strategies. A rather provocative Hughes-inspired insight might even suggest that forecasters are to a certain extent hired by economic actors to ‘make their mistakes for them’ (Hughes 1951, 321). Nevertheless, a series of ‘failures’ may lead to the persistence, or return, of disbelief. ‘Fiction in economic contexts is vulnerable to contradictory experiences in the real world’ (Beckert 2013, 225). Such tension within the forms of beliefs in which future-oriented activities engage is shared by both macroeconomic forecasting and magic. The belief may vanish if the expected benefits from magical acts do not come ‘true’. Magic, like religion, is viewed as a totality; either you believe in it all, or you do not. . . . Conclusions are immediately generalized, and a belief in a single case of magic implies the belief in all possible cases. Conversely one negative instance topples the whole edifice; magic itself then comes under suspicion. We have examples of obstinate credulity and deeply rooted faith crumbling before a single experience. (Hubert and Mauss 1902 [1972], 113)

Even though Hubert and Mauss pay little attention to it, they implicitly suggest a distinction between faith in ‘general’ and ‘singular’ beliefs—that is, between magic as a whole and local magical acts, or between forecasting per se and some forecasts or forecasters. Their work also stresses that, in some circumstances, a ‘will to believe’ does not suffice and that ‘make-believe’ practices are needed. Magical acts thus encompass ‘simulation’—a notion, the authors remark, which should not be confused with that of ‘fraud’, especially as it applies not only to the public or clients of magicians, but also to magicians themselves. Magicians’ practices and discourses are consequently aimed at convincing both others and themselves. In this chapter, a similar inside approach is proposed for the case of forecasting ‘errors’. Indeed, while forecasters freely acknowledge that ‘errors’ and ‘mistakes’ are their bread-and-butter, they also devote significant time to replying to critiques, sometimes even highlighting them in public5 or making fun of them, as this case-study observation makes clear:

5 In early 2017, the chief economist of the Bank of England provided an example of the necessity for forecasters to address public criticisms (‘Chief Economist of Bank of England Admits Errors in Brexit Forecasting’, Philip Inman, The Guardian, 5 January 2017. https://www. theguardian.com/business/2017/jan/05/chief-economist-of-bank-of-england-admits-errors, retrieved 11 January 2017).

126

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

How Forecasters Deal With ‘Errors’ After we met, the interviewee walks me back through the building. We come across another economist who, I am told, joined the Institute in the 1980s. I am encouraged to benefit from his long experience: ‘Didn’t you [the other economist] tell me [the interviewee] that, in the 80s, forecasts were as right as they are today?’ The other instantly replies: ‘You mean I said “as wrong”?’ Both laugh. (Institute, Observation, 1 September 2015)

In practice, forecasters’ discourses question the relevance of a ‘reality test’ in its most common-sense form. Because fully endorsing such a test would jeopardize the whole project of forecasting, forecasters’ justifications often design alternative definitions of forecasting ‘rightness’ that rely on the procedures it follows rather than its substance. Scrutinizing how ‘errors’ are dealt with internally provides insights into forecasters’ activity and the organization of the forecasting world. This shift implies a move from front stage, where forecast clients stand and public discussions take place, to back stage, wherein lie interpretative and discursive processes, as well as the actual production of forecasts (Reichmann 2013). It underlines the role of key features found in both magic and macroeconomic forecasting, such as rituals, professional identity, and representations of causal mechanisms. Ultimately, forecasters’ de facto replies to criticisms delineate a singular ‘regime of veridiction’—that is, ‘the set of rules enabling one to establish which statements in a given discourse can be described as true or false’ (Foucault 2008, 35). Paying attention to insiders’ perspective on this issue also makes it possible to understand how forecasters remain motivated to perform an activity that is bound to be ‘mistaken’ if its purpose is to depict correctly future economic reality. Competing definitions of ‘rightness’ are connected to differing representations of ‘errors’: one is defined externally, and the other internally. In line with Hubert and Mauss’s argument that magicians truly believe in their magic, including their own, even if it involves simulation to some degree, forecasters’ discourse should not be considered to be merely an expression of dishonesty— even though excluding bad faith a priori would obviously pose mirroring methodological problems. In a market for symbolic goods, actors’ justifications are for the most part rooted in sincere beliefs as to the worth of what is done, rather than solely expressing denial or concealing ‘real’ economic motivations (Bourdieu 1977, 5, 21–2).6 Going ‘back stage’ allows us to shed light on forecasters’ reflections on their own craft and their judgements on their own performance. These reflections and judgements revolve around three major 6 By contrast, Philip Tetlock’s analysis of political forecasts regards forecasters’ justifications as mere belief defence strategies (Tetlock 2005, 129–37). While stressing similar mechanisms to those that are tackled in this chapter, Tetlock incorporates them as elements for testing forecasters’ reliability in a psychology-inspired positivist perspective. However, the distinction he makes between ‘Getting it Right’ and ‘Thinking the Right Way’ undoubtedly applies to macroeconomic forecasters as well, as will be shown in this chapter.

127

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

arguments: the ontological indeterminacy of economies, the constructivist epistemology of forecasting, and the claim of professionalism.

Data and Method The material drawn upon in this chapter consists of interviews and observations from ongoing research on macroeconomic forecasting that focuses mainly, but not exclusively, on France (that is, macroeconomic forecasts produced about France). Thirty-six interviews have been conducted since June 2014 (average duration: eighty minutes) with economists and forecasters from public (either national or international) and private (banks, insurance companies, and so on) institutions. Among them, ten interviews were conducted with members of the same prominent French forecasting institution, which will be referred to here as the Institute, and where data from observation have been collected upon four successive forecast-production cycles between 2015 and 2017. During the interviews, interviewees were encouraged to describe the actual practice of forecasting, rather than just expressing their own views about forecasting in an abstract sense. The observations used in this chapter include those obtained at twenty-two work meetings (around three hours each), two press conferences (one hour each), and two debates during which the newly produced forecasts were discussed by a pair of economists from other French institutions (two hours each). The purpose of work meetings varies along with the process of forecasting, within which three stages may be roughly identified: post-mortem (the analysis of ‘errors’ in previous forecasts and of recent economic trends) and scenario and point forecasting. A ‘forecast’ indeed comprises two distinct elements. One is the ‘point forecast’, a precise figure depicting the value an economic variable may take in the future (for example, ‘next year’s GDP growth rate will be x %’), and whose calculation relies on econometric modelling and economic expertise. The other is the ‘scenario’—a narrative that exhibits a set of causal mechanisms leading from the present to the forecast horizon (the next quarter, semester, year, and so on), established through the combination of national accounts and judgemental assessments.7 Often considered a summary of the ‘scenario’, the ‘point forecast’ is determined at the 7 In the world of forecasting, two kinds of ‘scenario’ may be distinguished, along with two different kinds of operation. As described in this paragraph, the first refers to the ‘baseline scenario’, namely the ‘story’ (to quote numerous interviewees) that forecasters tell once causal mechanisms are identified and a set of hypotheses is settled upon. Another refers to ‘variants’—that is, changes in the set of hypotheses and assumptions in order to build ‘alternative scenarios’ and fathom the risks associated with the baseline scenario (Carnot et al. 2005, 148–54). To avoid confusion between these two meanings, the word ‘scenario’ will be used in this chapter only in the sense of a baseline scenario or narrative.

128

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

How Forecasters Deal With ‘Errors’

end of the process, once a scenario is agreed upon. It attracts public attention, if only because it fits formats in use in the media, such as newspaper headlines.

Is It Even Possible to Know? The Ontological Indeterminacy of Economies A first line of argument proffered by forecasters seeking to absolve themselves from forecasting ‘failure’ is related to the nature of economic reality and the tools forecasters use to apprehend it. The unknowability of the future is a constraint on economic decision-making and gave rise to the profession of forecasting. But uncertainty is also a resource for forecasters when facing contradiction. Their poor results in the context of a reality test may be qualified as consequences of the shortcomings of econometric models rather than as ‘errors’. By combining equations, models reflect the relationships between major economic variables (prices, exports, imports, public spending, employment, and so on) and describe the behaviour of ‘representative’ economic agents (Evans 1999, 50–76) in a way that is taken to be symbolic of the whole economy. In this respect, models hold a position in the world of forecasting similar to that of instruments and tools in magical rituals: specifically, they belong to the ‘class of objects which appear to be used for their own sakes by virtue of their real or imaginary properties, or . . . because they coincide with the nature of the rite’ (Hubert and Mauss 1902 [1972], 59). By definition, models focus only on chosen variables: completeness is not an option. It is thus no surprise that a significant part of the economic literature devoted to macroeconomic forecasting conveys a statistical representation of forecast ‘errors’ through the notions of ‘residuals’, ‘error terms’, or ‘loss functions’ (see, for example, Elliott and Timmermann 2008). Residuals capture all elements left outside the equations, such as the unexpected events (e.g. surprising weather conditions or accidents) that dramatically change the course of events and render the former assumptions, reasoning, and hence forecasts obsolete. These shocks are treated as external or ‘exogenous’ to the system—and therefore as legitimate excuses for forecasting ‘errors’. In practice, though, at least some of these shocks—the 2008 financial crisis for example—are ‘endogenous’, as they originate from innovations and other features within the economy (Bronk 2011). More importantly still, although economists discuss plausible stock market movements and possible changes in central bank policies, most, if not all, econometric models do not include the financial sector and its impact on the real economy: this is a consequence of the dominance of a macroeconomic framework within which the ‘real’ and the ‘financial’ spheres of the economy are unconnected (Fligstein et al. 2017; Goodhart et al. 2013). 129

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

All shocks or decisive factors not captured by models are treated in the same way, whether endogenous or exogenous. Obviously, the further out the forecast horizon, the greater the odds that shocks will occur. Whereas macroeconomic forecasting—along with other activities such as credit rating (Carruthers 2013)—purports to turn uncertain futures into calculated probabilities or ‘risks’, in classical terminology (Keynes 1921; Knight 1921 [1985]), shocks ensure the persistence of the very uncertainty that forecasting aims at reducing: the economy remains ontologically undetermined and the future remains open. The dream of a perfect elimination of uncertainty can never be realized and forecasts can thus never be ‘right’ in a realistic sense. Forecasters’ emphasis on radical uncertainty implicitly acknowledges that economic reality is nonergodic, that is, non-stationary and not determined by the past. Because ‘future outcomes are [not] merely the statistical shadow of past and current market signals’ (Davidson 1996, 480; see also the introduction to this volume), forecasting—which implies assessing economic regularities—is threatened at a microeconomic level by innovations, and at a macroeconomic level by shocks. A review of the ECB Survey of Professional Forecasters (SPF),8 for example, relates the ‘errors’ panel members made regarding the inflation rate to the various unexpected shocks that affected the newly created eurozone (Garcia 2003, 16). Economic crashes have a special place among such shocks. The case of the Great Recession, regularly evoked by forecasters in the interviews, illustrates this, all the more as the crisis exploded at the end of a forecasting process. Since forecasts are produced on a quarterly (end of March, June, September, and December) or biannual (end of March and September) basis, the collapse of Lehman Brothers on 15 September 2008 took place when point forecasts had been settled. ‘Error’ then brings into play both the shock itself and the organization and timing of the forecasting process, eventually making forecasts and forecasters ‘run off the road’, to quote a French economist. We had our scenario frozen on September 15th, 2008. The very same day Lehman Brothers collapsed! The forecast was not published yet, but the numbers were set: actually, to be released around October 15th, numbering has to be frozen earlier, next comes a phase of writing and harmonizing. [ . . . ] We actually had ended up, by mid-September 2008, with a growth scenario that was pretty much settled on and stated ‘well, there is a crisis, a liquidity crisis and a rather severe financial crisis’ but, back then, it seemed to us that . . . Big mistake we made was thinking the 8 The ECB-SPF is a quarterly survey of expectations concerning inflation rates, real GDP growth, and unemployment in the euro area for several horizons. Explicitly inspired by a similar survey carried out in the United States in 1968 by the American Statistical Association and the NBER, and taken over by the Federal Reserve in Philadelphia in 1990, the ECB-SPF was launched in 1999 to coincide with the introduction of the euro.

130

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

How Forecasters Deal With ‘Errors’ economy would be more or less resilient and there wouldn’t be another shock like this. So, what happened that day is that we didn’t change our forecasts. We are committed, though unofficially: people expect us to release our forecasts around mid-October and, at that moment, it was impossible to start all over again and make a brand new scenario. . . . We decided to go public anyway but said we’d make new forecasts by the end of the year. We warned our readers that the forecasts had been made before [the Lehman Brothers collapse] and would be updated. . . . Indeed, one bank going bankrupt is not enough for us to say that our scenario for growth in the Eurozone is ‘minus 4 %’, ‘4 % recession’. It needs the bankruptcy and a set of later developments. But it was embarrassing because we knew we were going to release something that ran off the road. (Economist, Institute, 28 April 2015)

The ontological indeterminacy of economies is a key issue for forecasters, who regard it as a major cause of ‘errors’ and ‘mistakes’. External critics argue that this brings discredit on forecasting per se. ‘Black Swans’—that is, highly improbable events with devastating consequences—underline experts’ inability to accurately predict what departs from the ordinary (Taleb 2007, 206–14). But for most economists, either orthodox or heterodox, the ‘Black-Swan theory’ is merely a wake-up call that emphasizes the limits of forecasters’ reasoning and modelling in the face of indeterminacy (Galbraith 2014, 4). Significantly, even ninety years after the Great Crash, the economic literature continues to address the possibility of being able to predict it, sometimes leading to posthumous reassessments of the importance and significance of economists with respect to their ‘prescience’ (Irwin 2014). For instance, Dominguez and her colleagues (1988) argue that the Great Depression could not have been foreseen either by contemporary forecasters (the then competing Harvard Economic Service led by Warren Persons and the team Irving Fisher directed at Yale9) or, more interestingly, by forecasters making use of late 1980s methods and data. In the same breath, the authors thus both refute the assertion that Harvard forecasters actually had anticipated the Crash (Bullock and Crum 1932, 137; Schumpeter 1954 [1997], 1131) and exculpate all forecasters from any period, on the grounds that ‘error’ is the consequence of unexpected events outside all models. On this line of reasoning, the number of ‘mistaken’ forecasters serves as a proxy measure of the historical uniqueness of an event, and hence of its unexpectability. As McGoey (2012, 563–71) argues, collective ignorance offers an ‘alibi’ that forecasters use to deflect individual accountability: the wider the extent of ignorance, the more easily ‘mistakes’ are denied. The very fact that no expert knew about or understood what was coming proves it was impossible

9 This early period of forecasting in the United States has been vividly recounted by W. A. Friedman (2014).

131

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

to know. Collective ignorance does not make the reality test ineffective as such but it makes forecasters immune to its verdict as they are not accountable for not having predicted what was unpredictable: forecasts were ‘wrong’ but no error was made. Contrary to mainstream economics, whose perfect-competition hypothesis assumes predictability as a methodological feature of reasoning, forecasting as a form of applied economics deals only with actual economic conditions and thus must cope with a reality that is partly indeterminate and uncertain. The activity of forecasting is, consequently, doomed to be at least a partial failure: by definition, it cannot capture ex ante what a dynamic and unstable ‘reality’ will be. This provides forecasters with a rationale to discard a supposedly stable ‘reality’ as a relevant benchmark to assess the accuracy of forecasts. At the same time, acknowledging the difficulties of forecasting involves shedding new light on the process of forecast production and suggests methodological refinements. In this respect, ‘errors’ mark the shifting frontier of knowledge about the economic environment, the relationship between variables in an econometric model, and the values they are expected to take. Hence the statement that forecasts are usually ‘erroneous’ leads to the apparently paradoxical conclusion that forecasting methods and models should be improved rather than withdrawn once and for all (Evans 1997). As part of a never-ending trial-and-error process, ‘error’ may then even be of some value by pointing towards the limits and flaws of forecasting models and reasoning.

What Is It All About? The Epistemology of Forecasting Emphasizing the way in which forecasts are produced offers forecasters another argument for proclaiming the irrelevance of the reality test. Criticisms addressed at macroeconomic forecasting usually adopt a ‘realistic perspective’: they focus on numerical output to compare it ex post to the ‘reality’ it was supposed to anticipate ‘correctly’. ‘Mistakes’ then come from mismeasurements (Desrosières 1998). A stress on the nature and epistemological basis of forecasting helps forecasters to challenge this standpoint on two different, yet related, grounds: first, by focusing on the nature and consequences of forecasting as a practical activity based around scenarios; and secondly, by insisting that a ‘realist’ perspective on forecasting is misguided, because the use of statistical data requires the adoption of a ‘constructivist’ (Desrosières 2001) or ‘constructionist’ (Hacking 1999) perspective. To a large extent, narrative forecast scenarios matter more to forecasters than calculated point forecasts. Forecasters often describe the establishing of scenarios as the ‘true’ purpose of their activity. Macroeconomic evolutions and causal mechanisms are then described in relation to a statistical and 132

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

How Forecasters Deal With ‘Errors’

qualitative assessment of the current macroeconomic situation. Qualitative information is distributed along a wide range of phenomena, from public policy announcements to scandals involving major firms, through elections, weather conditions, and geopolitical environment. Built from a set of assumptions and guiding the subsequent econometric modelling, narratives illustrate the blend of statistical data and judgement that characterizes macroeconomic forecasting. Scenarios rely on experience, expertise, and intuition, all of which are assumed to endow forecasters with an ability to spot economic processes as they unfold: while economic reality is too erratic for accurate point forecasts, it remains predictable enough for regularities to be discerned. Such scenarios comprise representations of causal mechanisms in the economy. As emphasized in the Durkheimian tradition, religion, science, and magic all rely on representations of causality: whether in Newtonian science, primitive religion, or magic, unexplained ‘forces’ ultimately operate as explanatory principles (Durkheim 1912 [1965]). A more or less comprehensive analysis shows how scientific, religious, or magical actions can set these forces in motion in order to produce effects (Hubert and Mauss 1902 [1972], 75, 78). In macroeconomic forecasting, representations of causality deal only with the analysis of economic causes and consequences. Differences between forecasts from different institutions, or from the same institution at different times are thus often related to differences between scenarios that, in turn, may be understood with reference to the various possible depictions of economies. Baseline scenarios remain interpretations of the present economic environment, and one of the key challenges facing forecasters is selection of the relevant information. Besides, scenarios are thoroughly discussed within forecasting institutions and among forecasters from different institutions. As Reichmann (2013, 857; see also Evans 2007) notes, ‘economic forecasters are permanently negotiating their views with each other and with others to come to a certain view’. Since interpretations based on judgements and shaped through collective deliberation are central to scenarios, they are often revised as new information and elements are taken into account. In this way, they combine facts (what is known) and bets regarding developing situations in a way reminiscent of the way in which uncertainty is managed in the case of journalism (Pilmis 2014). This can be illustrated by the case of ‘Brexit’. Until the vote (23 June 2016), forecasters from the Institute bet that the United Kingdom would remain in the EU, in accordance with a common ‘nothing changes’ assumption, and hence Brexit was not regarded as a key issue for forecasting. By contrast, once the referendum results were known, Brexit gained major importance in the process of forecasting. This can be measured by the amount of time devoted to it: Brexit was not covered in any of the five observed meetings of the spring 2016 forecasting session, whereas the autumn 2016 work meetings usually began 133

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

with Brexit, with up to two hours devoted to it. It may also be measured in terms of coverage in the published forecasts: some two and a half pages out of 180 in spring 2016 (with the word ‘Brexit’ appearing on ten occasions) as against twenty-one pages out of 206 in autumn 2016 (with the word ‘Brexit’ appearing on sixty occasions). Forecasters often present narratives as the ‘real’ forecasts, those that should be paid attention to rather than ‘point’ forecasts. In this perspective, forecasting has more to do with storytelling than with calculation. Whereas it seems easier to assess the accuracy of forecasts, and thus their ‘errors’, with reference to calculated numbers, the quality of storytelling is a more appropriate, though less convenient, test of forecasts.10 To fellow forecasters, a ‘good’, or ‘right’, forecast ‘correctly’ describes macroeconomic progression in such a way that the ‘story’ matches the coming string of events, even if the anticipated figures are ‘wrong’. The often-quoted expression ‘being right for the wrong reasons, and being wrong for the right reasons’ suggests that forecasting is not subject to simple falsification because it leads to assertions unsuited to a process of refutation. In short, forecasters argue for a more nuanced epistemological position in relation to their craft. The partial construction by contingent theoretical frames of the data used in forecasts also introduces indeterminacy. This suggests that it is appropriate to adopt a constructivist (rather than realist) epistemology, which deals with the work routines, deliberations, and conventions involved in the process of forecasting. Conventions characterize not only forecasting institutions but also those surrounding them. A detailed investigation of the principles according to which data are built and knowledge is produced is often described as a prerequisite for judging forecasting: the classical statement that data do not pre-exist their collection or gathering is a common caveat among forecasters. This is obviously the case for abstract notions, such as potential GDP or the output gap, with which actual economic performances are compared and whose estimations vary widely across forecasting institutions. It is also true of the more or less formal information that forecasters collect from the economic actors they meet, as well as of the raw material that statistical bureaus supply them with. Forecasting ‘errors’ may originate in misleading information or erroneous statistical data, as they may lead to ‘inaccurate’ depictions of the economy. The case of economic indices exhibits the constructivist point of view that forecasters adopt: whereas they are regularly considered to be reliable measures of economic activity (as shown by their use as the

10 Forecasters regularly complain about what seems to them an excessive focus on numbers, leading to situations that are described as ‘absurd’: calling a 1.5% GDP growth rate forecast ‘wrong’ because the observed rate was 1.3% would just be a matter of the ‘thickness of the line’ (Economist, asset manager, 22 October 2015).

134

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

How Forecasters Deal With ‘Errors’

unquestioned ‘reality’ against which forecasts are compared), their regular revisions have a strong impact on macroeconomic forecasting. You need people with a bit of knowledge about public stat[istics], people who know how key figures, those everyone pays attention to, are constructed. When you’ve been in the public sector, you know in what condition figures are brought to you, because you have practiced their production, so you’re a bit better than others since you know what may play. I, for instance, since I used to forecast inflation, I know that in May and December, revisions are large because basic welfare benefits rise. So, these two figures are a bit more important. I know that rises in the price of natural gas (by 2.3% as announced today) or transportation will have an impact on inflation on the following month . . . (Chief economist, insurance company, 13 February 2017) Statcan [Statistics Canada] is for me a model. And why does it work great? They estimate GDP monthly, and it is a very good proxy for the upcoming quarterly data, with few revisions. That’s why I told you [earlier] that ‘the [Japanese] data is no good’. Japanese data is tremendously revised. And when you use a model to forecast GDP in the short run, you not only check the last statistical point, but also the preceding quarters. If the quarters are revised a lot, you’ll be mistaken a lot. That’s why I called them ‘disastrous’, because of the revisions. (Economist, international economics agency, 9 November 2015)

Discussing the nature and purpose of forecasting gives forecasters the opportunity to dispute the reality test as a relevant assessment of quality. First, they argue that a focus on calculation misses the ‘true’ locus of forecasting, which consists in establishing scenarios. Though less apparent, scenarios are presented as a decisive outcome of forecasting. Secondly, forecasters’ epistemological arguments make it possible to qualify the results of a reality test as consequences of the data used in the process of forecasting. As data come at the conclusion of a series of conventions and operations, they may not accurately measure today’s economic situation; and, consequently, neither can forecasts envision tomorrow’s situation. Like magical rituals, forecasting implies that ‘the confection or preparation of [the] materials, the ritual ingredients, is a long and finicking business’ (Hubert and Mauss 1902 [1972], 58). ‘Error’ springs not only from the centre of the universe of forecasting, but also from its margins, where raw data are produced at the earliest stage of the forecasting process. In any event, forecasters should not take the blame. The emphasis on using second-hand data and information to produce forecasts ultimately depicts the universe of forecasting as a ‘social world’, or as a segment of a broader ‘social world of economic analysis’. Like the art worlds that H. S. Becker (1982) studied, the ‘forecasting world’ cannot be reduced to those who are located at its heart: beyond the most obvious practitioners (forecasters themselves), support personnel—such as statistical agencies or 135

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

economic actors who ‘epistemically participate’ (Reichmann 2013) in the activity—have to be taken into account.

What Makes a ‘Good’ Forecaster? Forecasters’ Claims of Professionalism The notion that forecasting is a ‘social world’ implies that it is ruled by a set of conventions according to which there is an established boundary between what ought to be done and what not, enabling cooperation between actors (Becker 1982, 28–34). Conventions delineate a set of rules and methods that ‘professional’ forecasters are expected to comply with. Although central bank staff are unlikely to be familiar with the sociology of professions and occupations, their actual definition of what the forecasting ‘profession’ might be when surveying ‘professional forecasters’ is noteworthy. They identify the implemented methods as a key feature of professionalism: in the case of the ECB-SPF, the selection and inclusion of panellists are delegated to each member country’s central bank. The criteria applied include a ‘formalized approach’ (by which is meant the use of macroeconomic modelling), which is taken as crucial evidence that a given forecasting institution ‘fits the . . . description [of] professional forecasters’ (Economist, European Central Bank, 26 June 2014). Generally speaking, according to forecasters, a major aspect of ‘professional’ forecasting is consistency—that is, the mutual compatibility of all macroeconomic aggregates that are included in the analysis, whether it relies on narratives or statistical models. Specifically, macroeconomic forecasts have to fit national accounting principles, and for example make sure the total exports from all countries equal total imports. Some widely used terms, such as ‘closure’ (French: ‘bouclage’), express this concern. Both econometricians dealing with large-sized statistical models and economists who only make use of spreadsheets to implement small-scale models consisting of supply-use tables insist on the importance of consistency for forecasting. A national accounts basis for forecasting professionalism delineates the legitimate forms of forecasting and operates as a key criterion to judge the quality and validity of forecasts. Mavericks’ claims to have foreseen what other forecasters missed are often met with scepticism if not hostility due to their perceived lack of professionalism: it is no surprise that they are regarded by ‘professional’ forecasters as mere outsiders whose reasoning lacks basic economic principles and who should be kept beyond the boundaries of the world of macroeconomics. The skills of economic ‘visionaries’ are regularly disputed. By stressing their lack of training in macroeconomics and challenging the outlined causal sequence, forecasters question the expertise of people who were probably just ‘lucky’.

136

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

How Forecasters Deal With ‘Errors’ Sometimes you read something and think ‘this is rubbish’, because it doesn’t fit, there is no closure with respect to macroeconomics. I do not pretend mine are always perfect but still . . . Actually a first simple closure is accounting identity: make sure national accounts are a closed system, which means one cannot see an increase in households saving rate and an increase in consumption if income is lower. Unfortunately, it happens sometimes. . . . Well, it’s a bit more subtle. If you take a look, you see the income doesn’t rise enough to allow what has been reported. It’s not that blatant, but still. Accounting identities have to be checked. Then another thing is . . . for example, if you report a decrease in growth in China, in the US, and in France, you cannot see a rise in the foreign demand for French goods. There is an identity at the international level which is actually pretty weighty because world demand has to be taken into account. (Economist, research institute, 23 October 2015) Predicting the worst is a good strategy because as time goes by . . . See, nowadays, X [another French observer who, the interviewee assures us, is a professor of finance and ‘knows nothing of macroeconomic matters’] made a business out of it. He keeps repeating ‘the crisis lies ahead’. And he is right: the crisis is ahead! Inevitably there will a crisis at some point. If you want to win the media, you have to say the crisis lies ahead because afterwards people think you should have been listened to. . . . Last summer, X said (we know each other well, he is a nice guy by the way) ‘there is a crisis ahead’. ‘Sure, you’re right, date it’. ‘This summer.’ And, once the Chinese thing occurred [the 2015 Chinese slowdown], he came back: ‘See? I told you so.’ ‘No, you didn’t say that. You said the stock markets would collapse in the United States and Europe for intrinsic reasons that had nothing to do with emerging countries’. (Economist, Institute, 7 September 2015)

Conversely, forecasters who go by the book are exonerated of forecasting ‘errors’: they are not to be held responsible for ‘mistakes’ since none have been made. Forecasters advocate an approach to ‘error’ that relates it exclusively to the degree of compliance with internally defined standard practices and formal rules. In line with the Durkheimian view that magicians represent an example of early professionals, forecasters’ emphasis on procedures seem connected to their claim that their professional identity is first and foremost that of economists who shall be judged according to the narratives they set up rather than the calculations they end up with. Like other occupations, protection from blame is obtained by proving that the ritual of forecasting has been followed: insofar as the process has been correctly handled, there is no point in criticizing its outcomes and only ‘professionals’ are, consequently, authorized to assess that ‘errors’ have been committed (Hughes 1951, 324–5). As noted in the case of magic, the general belief in forecasting remains to a certain extent immune to contradictory experiences because it may be argued that the dubious ritual was not properly executed, rendering it null and void. The conditions required for magic or forecasting to be successful are numerous

137

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

‘to such an extent that it seems to be providing itself with loopholes, and often successfully’, so that ‘it is natural for a magician to take refuge behind questions of procedures and technicalities, to protect himself in case of failure of magical prowess’ (Hubert and Mauss 1902 [1972], 62). Forecasters, then, like magicians, argue that all that really matters is that they follow due process. The outcome can be that front stage (public discussion) is dismissed as a place for disputing the accuracy of forecasts and that discussions remain confined to the back-stage area (the profession itself): while economic actors may provide forecasters with relevant information, debates and disputes are kept between ‘professionals’. As mentioned earlier, that does not mean that forecasters do not acknowledge that forecasts are frequently ‘erroneous’ from a realist point of view, but a firm distinction is made between ‘erroneous forecasts’ and ‘faulty forecasters’. As a social world based on shared conventions, the forecasting world also displays some of the properties of ‘professional’ ecologies. The combination leads to a representation of the world of forecasting as autonomous. Social boundaries protect forecasters both from competitors whose abilities are contested and from criticisms originating from outside the profession. However, forecasting cannot achieve full autonomy, because it necessarily remains connected to the economy, which takes part in the process of forecasting and within which forecasters’ clients are to be found. At some point, forecasters are bound to turn back to the external sphere and front stage. Ironically, then, the distance that forecasters can put between themselves and the reality test should not be overestimated. Provided the general credibility of forecasting is maintained, the credibility of specific forecasts or forecasters can be questioned. Indeed, regarding its connection to the outside, the forecasting world shares features with a market within which actors compete with each other to access economically defined opportunities with clients and customers (Weber 1922 [2013], 635–40). Such competition necessarily takes place in the foreground and uses the ‘reality test’—that is, the accuracy of ex ante forecasts—to assess forecasters’ reliability: ‘errors’ may be put forward as ‘evidence’ that some forecasters are more trustworthy than others. The reality test is then commonly used to increase the symbolic capital or prestige of individual forecasters or forecasting institutions. [At the end of the interview, the interviewee shows me a couple of files on his computer] I can’t help the sweet treat: you’ll get why. [He opens a spreadsheet: two tables display data from a dozen forecasting institutions – French government, OECD, IMF, research institutes and banks – for two different periods: 1999–2016 and 2013–2016]. So I look at what has been forecasted for Year y in September of Year y 1, then I look at the first GDP growth estimate for Year y, which is released in January of Year y+1. Obviously I consider the difference in absolute value. Here it is: our average difference for 2013–2016 is 0.15 point [he waves the computer

138

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

How Forecasters Deal With ‘Errors’ mouse to stress this figure is lower than that of other institutions]. For example, we forecasted an average 1.15 GDP growth and it came out at 1. . . . I fill the form myself. So, here is what my little fellows have marked for 2017 so, when it’ll be January 2018, I will check it out. . . . Everyone is convinced [they] have the best forecast except [he says it as if whispering a secret] that I have the evidence. (Chief economist, research institute, 22 March 2017)

Like many other social spaces, the forecasting world is not characterized by an opposition between symbolic and economic rewards: on the contrary, both coalesce to a large extent. Indeed, retaining clients and gaining new ones is a key concern for privately funded forecasting institutions. In this perspective, the accuracy of forecasts exhibits ‘evidence’ of excellence in economic analysis, which ‘clients remember’ (Economist, private bank, 30 September 2015), even though they often value the narrative element of forecasting as well. Dismissed as being irrelevant to assess the quality of forecasts in general terms, the reality test remains the most commonly used instrument when specific macroeconomic forecasts come to be publicly debated.

Conclusion The contradiction between what had been anticipated and what actually happened is a common challenge for all belief-based practices, and serves as the basis for comparing present-day macroeconomic forecasters to the magicians that Hubert and Mauss analysed in the early twentieth century. Rather than presenting the condescending claim that sophisticated macroeconometric models barely conceal primitive forms of reasoning, the aim of this chapter is to suggest that, though different, magic and forecasting share important traits. Both are activities whose results become apparent somewhere down the line, and in each case their performance relies on precise forms of representation, purposively designed materials, defined sets of procedures, and trained professionals. The analysis presented in this chapter of the discourse that forecasters use to legitimate ‘errors’ and preserve forecasting from a reality test that could endanger its very existence involves three arguments. Taken together, they serve to reduce the relevance of a comparison between ex ante and ex post figures to assess the ‘quality’ of forecasts and to mitigate the impact of forecasters’ overall poor records in this respect. In the first place, forecasters argue that forecasting focuses on narratives rather than on calculations: figures and numbers are thus discarded as the appropriate criterion by which forecasts should be judged. Instead, narratives identify the causal mechanisms through which economic processes unfold, 139

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

and make representations about causality a key feature of forecasting. In this perspective, a ‘reality test’ based on estimating the accuracy of ex ante point forecasts is revoked. The causal representations may also account, however, for the discrepancies between forecasts and ‘reality’ in order to dispute their qualification as ‘errors’. Indeed, forecasters contend that many representations of ‘reality’ are misleading, as they assume that economies are more stable and predetermined than they actually are. The causal sequences that economic theory posits, for example, are often pulled off track by contingent—hence unanticipated— phenomena. A proper appreciation of the indeterminacy of economic futures, consequently, absolves forecasters of blame for ‘errors’. The materials used to produce forecasts pose a second set of challenges. Whether they are made up of qualitative information or statistical data, they are considered a possible source of ‘error’ as a result of the contingent conventions involved in their construction. While, on an ontological level, economic ‘reality’ is regarded as shaped by events over which forecasters have no control, this epistemological argument suggests that attention should be paid to the series of operations leading to the depiction of such ‘reality’. The implication, again, is that forecasters should not necessarily be held responsible for ‘errors’. A reality test is not, in this case, dismissed in principle, but its results are framed in such a way that they may not qualify as ‘errors’. A third argument puts forward the notions of ritual and profession. Forecasters claim that they are mostly, if not exclusively, accountable for their compliance with sets of standards and practices. In this perspective, the process of forecasting matters more than its results: provided forecasts are produced according to the proper methods, they may not be considered faulty. A ‘professional’ representation of the forecasting world disputes the relevance of an external adjudication of the reality test: instead of the test being part of a public debate, the emphasis on professionalism tends to confine discussions about forecasting ‘quality’ within the limits of the forecasting world. Another shared feature between magic and forecasting is thus their separation from the ordinary, or ‘profane’, world. Ontological, epistemological, and professional claims resonate to challenge the appropriateness of the reality test in general. Nevertheless, forecasting cannot stand entirely apart from the rest of the economic world. Front stage does not lie so far away from back stage. Since its actual purpose remains to guide economic actions, forecasting takes part in a division of labour that makes it necessary for forecasters to engage in relations with ‘outsiders’. Moreover, while forecasts are regularly used to depict the future evolution of the economy, especially in the private sector they are also used competitively to assess the quality of economic analysis and gain clients. Forecasts and forecasters’ credibility thus remain at stake. Despite 140

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

How Forecasters Deal With ‘Errors’

forecasters’ efforts to frame ‘errors’ as an issue solely for insiders, the reality test cannot be dismissed once and for all. But the resurgence of a reality test presupposes the narrowing of its scope: maintaining a general belief in forecasting is a prior condition for disputing particular forecasts and forecasters.

Acknowledgements Earlier versions of this text were presented at the SASE mini-conference ‘Uncertain Future in Economic Decision Making’ (LSE, July 2015), and at the workshop ‘Genèse des Futurs Économiques’ at the University of Toulouse-Jean Jaurès (Toulouse, December 2015). I thank the participants in these sessions for their comments and suggestions. I am also grateful to Vincent Cardon (CURAPP), Sidonie Naulin (PACTE), Étienne Nouguez (CSO), and the editors of this volume for their remarks.

Bibliography Becker, Howard S. 1982. Art Worlds. Berkeley, CA: University of California Press. Beckert, Jens. 2013. ‘Imagined Futures: Fictional Expectations in the Economy’. Theory and Society 42 (3): pp. 219–40. Beckert, Jens. 2016. Imagined Futures: Fictional Expectations and Capitalist Dynamics. Cambridge, MA: Harvard University Press. Boltanski, Luc. 2011. On Critique: A Sociology of Emancipation. Cambridge: Polity. Boltanski, Luc, and Laurent Thévenot. 2006. On Justification: Economies of Worth. New edn. Princeton, NJ: Princeton University Press. Bourdieu, Pierre. 1977. ‘La production de la croyance: Contribution à une économie des biens symboliques’ [The Production of Belief: Contribution to an Economy of Symbolic Goods]. Actes de la recherche en sciences sociales 13 (1): pp. 3–43. Bronk, Richard. 2011. ‘Uncertainty, Modelling Monocultures and the Financial Crisis’. Business Economist 42 (2): pp. 5–18. Bullock, Charles J., and William L. Crum. 1932. ‘The Harvard Index of Economic Conditions: Interpretation and Performance, 1919–31’. Review of Economic Statistics 14 (3): pp. 132–48. Carnot, Nicolas, Vincent Koen, and Bruno Tissot. 2005. Economic Forecasting. Basingstoke, UK: Palgrave Macmillan. Carruthers, Bruce G. 2013. ‘From Uncertainty toward Risk: The Case of Credit Ratings’. Socio-Economic Review 11 (3): pp. 525–51. Davidson, Paul. 1996. ‘Reality and Economic Theory’. Journal of Post Keynesian Economics 18 (4): pp. 479–508. Desrosières, Alain. 1998. The Politics of Large Numbers: A History of Statistical Reasoning. Cambridge, MA: Harvard University Press. Desrosières, Alain. 2001. ‘Entre réalisme métrologique et conventions d’équivalence: les ambiguïtés de la sociologie quantitative’ [Between Metrological Realism and

141

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures Conventions of Equivalence: the Ambiguities of Quantitative Sociology]. Genèses 43: pp. 112–27. Dominguez, Kathryn M., Ray C. Fair, and Matthew D. Shapiro. 1988. ‘Forecasting the Depression: Harvard versus Yale’. American Economic Review 78 (4): pp. 595–612. Durkheim, Émile. 1912 [1965]. The Elementary Forms of the Religious Life. New York: Free Press. Elliott, Graham, and Allan Timmermann. 2008. ‘Macroeconomic Forecasting’. Journal of Economic Literature 46 (1): pp. 3–56. Evans, Robert. 1997. ‘Soothsaying or Science? Falsification, Uncertainty and Social Change in Macroeconomic Modelling’. Social Studies of Science 27 (3): pp. 395–438. Evans, Robert. 1999. Macroeconomic Forecasting: A Sociological Appraisal. Abingdon, UK: Routledge. Evans, Robert. 2007. ‘Social Networks and Private Spaces in Economic Forecasting’. Studies in History and Philosophy of Science 38 (4): pp. 686–97. Foucault, Michel. 2008. The Birth of Biopolitics: Lectures at the Collège de France, 1978–1979. Basingstoke, UK: Palgrave Macmillan. Fligstein, Neil, Jonah S. Brundage, and Michael Schultz. 2017. ‘Seeing Like the Fed: Culture, Cognition, and Framing in the Failure to Anticipate the Financial Crisis of 2008’. American Sociological Review 82 (5): pp. 879–909. Friedman, Walter A. 2014. Fortune Tellers: The Story of America’s First Economic Forecasters. Princeton, NJ: Princeton University Press. Galbraith, James K. 2014. The End of Normal: The Great Crisis and the Future of Growth. New York: Simon & Schuster. Garcia, Juan Angel. 2003. ‘An Introduction to the ECB’s Survey of Professional Forecasters’. ECB Occasional Paper Series 8. http://www.ecb.europa.eu/pub/pdf/scpops/ ecbocp8.pdf. Goodhart, Charles, Dimitrios Tsomocos, and Martin Shubik. 2013. ‘Macro-Modelling, Default and Money’. LSE Financial Markets Group Special Paper no 224. http://www.lse. ac.uk/fmg/workingpapers/specialpapers/pdf/sp224.pdf. Hacking, Ian. 1999. The Social Construction of What?. Cambridge, MA: Harvard University Press. Hubert, Henri, and Marcel Mauss. 1902 [1972]. A General Theory of Magic. London: Routledge. Hughes, Everett C. 1951. ‘Mistakes at Work’. Canadian Journal of Economics and Political Science 17 (3): pp. 320–27. Irwin, Douglas A. 2014. ‘Who Anticipated the Great Depression? Gustav Cassel versus Keynes and Hayek on the Interwar Gold Standard’. Journal of Money, Credit and Banking 46 (1): pp. 199–227. Keynes, John M. 1921. A Treatise on Probability. London, UK: Macmillan. Knight, Frank H. 1921 [1985]. Risk, Uncertainty and Profit. Chicago, IL: University of Chicago Press. McGoey, Linsey. 2012. ‘The Logic of Strategic Ignorance’. British Journal of Sociology 63 (3): pp. 553–76. Orléan, André. 2014. The Empire of Value: A New Foundation for Economics. Cambridge, MA: MIT Press.

142

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

How Forecasters Deal With ‘Errors’ Orrell, David. 2010. Economyths: Ten Ways Economics Get It Wrong. Mississauga, ON: John Wiley & Sons Canada. Pilmis, Olivier. 2014. ‘Producing in Urgent Situations: The Management of Unpredictability in the World of Journalism’. Revue française de sociologie 55 (1): pp. 101–26. Reichmann, Werner. 2013. ‘Epistemic Participation: How to Produce Knowledge about the Economic Future’. Social Studies of Science 43 (6): pp. 852–77. Shackle, George. 1972 [1992]. Epistemics and Economics. A Critique of Economic Doctrines. New Brunswick, NJ: Transaction Publishers. Schumpeter, Joseph A. 1954 [1997]. History of Economic Analysis. Abingdon, UK: Routledge. Taleb, Nassim N. 2007. The Black Swan: The Impact of the Highly Improbable. New York: Random House. Tetlock, Philip E. 2005. Expert Political Judgment: How Good Is It? How Can We Know? Princeton, NJ: Princeton University Press. Turin, Michel. 2015. Le bal des aveugles [The Ball of the Blind]. Paris: Albin Michel. Weber, Max. 1922 [2013]. Economy and Society: An Outline of Interpretive Sociology. Berkeley, CA: University of California Press.

143

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

7 Uncertainty in Macroeconomic Modelling Andrew G. Haldane

Introduction The past few years have witnessed a jarring financial crisis as great as any we have experienced since the world wars, a crisis whose aftershocks are still being felt today. Against that dramatic backdrop, this chapter reflects on the state of economics, not least in helping us make sense of such catastrophic phenomena. In particular, it examines whether we need a new approach to modelling uncertain futures—one that moves away from relying on mechanical forecasting towards drawing on simulations using agent-based models.1 This topic has risen in both prominence and urgency since the financial crisis (Battiston et al. 2016; Coyle and Haldane 2014). Indeed, it would probably not be an exaggeration to say the economic and financial crisis has spawned a crisis in the economics and finance profession—and not for the first time. Much the same occurred after the Great Depression of the 1930s when economics was rethought under Keynes’s intellectual leadership (Keynes 1936). Although this crisis in economics is a threat for some, for others it is an opportunity—an opportunity to make a great leap forward, as Keynes did in the 1930s. But seizing this opportunity requires a re-examination of the contours of economics and an exploration of some new pathways. In the light of the crisis, there has been renewed interest in the work of George Shackle as economists have sought new insights into age-old problems. In exploring new pathways, 1 This chapter is an amended and abridged version of the GLS Shackle Biennial Memorial Lecture given by Andrew Haldane on 10 November 2016. The original lecture is available online at: www.bankofengland.co.uk/publications/Pages/speeches/default.aspx. The views expressed are not necessarily those of the Bank of England or the Monetary Policy Committee.

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertainty in Macroeconomic Modelling

this chapter draws inspiration from several features of economic systems that underpinned Shackle’s work. First, there is the importance of recognizing that these systems may often find themselves in a state of near-continuous disequilibrium. Indeed, the notion of equilibrium, stationary through time, may itself be misleading (Shackle 1972). Shackle argued that it was inappropriate to model dynamic and innovative markets using the metaphor of ‘celestial mechanics’ as manifested in ‘the neo-classical conception of general equilibrium’ (Shackle 1972, 4; Bronk 2009, 211). It was perhaps this feature of Shackle’s work that earned him his ‘heterodox’ label. Second, there is the importance of radical uncertainty. Shackle saw uncertainty about the future as fundamental to human decision-making and, thus, to the functioning of social systems (Shackle 1972). Human imagination was a crucial frame for social progress (Shackle 1979). This meant that social systems were inherently unpredictable in their behaviour. Latterly, the importance of radical uncertainty in making sense of social systems has gained new traction (King 2016; Taleb 2014). This chapter explores how we might deal with the prevalence of disequilibria and radical uncertainty in complex economic systems, and examines the potential of agent-based models as a tool for helping us understand the dynamics of these systems and the impact of policy interventions.

From Natural to Social Sciences One of the potential failings of the economics profession is that it may have borrowed too little from other disciplines; in other words, it might be described as a methodological monoculture. In 1999, Professor Nancy Cartwright, a philosopher of science, published a book entitled The Dappled World: A Study of the Boundaries of Science (Cartwright 1999). This passage captures its essence: [s]cience as we know it: apportioned into disciplines, apparently arbitrarily grown up; governing different sets of properties at different levels of abstraction; pockets of great precision; large parcels of qualitative maxims resisting precise formulation; erratic overlaps; here and there, once in a while, corners that line up but mostly ragged edges; and always the cover of law just loosely attached to the jumbled world of material things. (Cartwright 1999, 1)

Cartwright was describing the natural sciences. She describes them as a patchwork of theory and evidence, some of it precisely cut, most of it haphazardly shaped, and pieced together irregularly. And there is, she argues, no shame in that. To the contrary, it may be the best science can do, given limited 145

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertain Futures

knowledge and limited time, in trying to make sense of an often-jumbled world. While Cartwright’s Dappled World aims to dispel the notion that natural science is precise science, it goes further in arguing that the same case can be made for social science. It, too, is a haphazard patchwork of theories and hunches, operating at different levels of abstraction, often loosely held together. The economic and financial world is every bit as dappled as the natural world. And, as with the natural sciences, there is real virtue in an eclectic approach. However, this view jars with the dominant methodological direction of travel in economics over many years. That methodological lead was provided by Karl Popper in the 1930s (Popper 1934 [1959]). In The Logic of Scientific Discovery, Popper argued for a ‘deductive’ approach to advancing knowledge. This involved, first, specifying a clear set of assumptions or axioms. From those were deduced a set of logical propositions or hypotheses. And then, and only then, were these hypotheses taken to the data to be tested and subjected to possible refutation. Newtonian physics was built on the same principle of internal consistency within systems, with energy within the system always preserved. Systems were subject to disturbances that could cause them to oscillate dynamically. Ultimately, however, these systems typically had an equilibrium or steady-state to which they returned. As one of the simplest examples, Newton’s pendulum exhibits damped harmonic motion once displaced before returning to a state of rest. Much of mainstream macroeconomics and finance has essentially followed this intellectual lead. It typically starts with a set of assumptions or axioms— often defining the preferences of consumers and the technology available to firms. From those assumptions are derived equations of motion for the behaviour of consumers and firms—which, rather revealingly, are often called the Euler conditions. Then, and only then, are these first-order conditions taken to the data to be tested. For example, the dominant approach over recent decades to modelling the macroeconomy has probably been the Dynamic Stochastic General Equilibrium (DSGE) model (Smets and Wouters 2003). In its plain-vanilla form, this comprises a set of representative, optimizing households and firms. This model gives rise to an equilibrium that is unique and stationary, and dynamics around that equilibrium that are regular and oscillatory. Various knobs and whistles have been added to this workhorse framework, often involving market frictions in price-setting, competition, and credit provision. These add colour to the model’s dynamics but, by and large, leave its properties intact—stable, stationary, oscillatory. It is not just among academics that this workhorse framework has found favour. The majority of 146

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertainty in Macroeconomic Modelling

central banks also take the DSGE framework as their starting point, including the Bank of England (Burgess et al. 2013). The DSGE approach has many of the hallmarks of Newtonian physics. As every action has an equal and opposite reaction, every shock has an equal and proportional reaction in DSGE models. The economy’s dynamics exhibit the same damped harmonic motion as Newton’s pendulum, or a rocking horse hit with a stick. The rocking horse metaphor is apt, as it was first used by Swedish economist Knut Wicksell almost a hundred years ago to describe the business cycle motion of an economy (Wicksell 1918).2 Mainstream finance and macroeconomics, then, have followed firmly in the footsteps of giants, part Popperian, part Newtonian. It has been heavily indebted, intellectually, to classical physics. That has led some to dub the dominant economic paradigm ‘econo-physics’ (Mirowski 1989). Less kindly, some have described economics as suffering from physics-envy (Hirschman 1991).

Assessing the Pros and Cons of Standard Modelling Approaches Despite recent criticism, which has come thick and fast, it is important not to overlook the benefits that have come from following this path. One benefit, shared with theoretical physics, is that economic theory has well-defined foundations. There are fewer ‘free’—or undefined—parameters floating around the model. Nobel Laureate Robert Lucas said ‘beware of economists bearing free parameters’.3 He was right. A theory of everything is a theory of nothing. The advantages do not stop there. On the assumption that agents’ behaviour is representative—it broadly mirrors the average person’s—these models of microlevel behaviour can be simply summed to replicate macroeconomic behaviour. The individual is, in effect, a shrunken replica of the economy as a whole. These macroeconomic models are, in the jargon, microfounded—that is, constructed bottom-up from optimizing, microeconomic foundations. These advantages carry across into the policy sphere. If the assumptions underlying these models are valid, then the behavioural rules from which they are derived will be unaffected by changes in the prevailing policy regime. These models are then a robust test-bed for policy analysis. They are, in economists’ jargon, immune to the Lucas critique (Lucas 1976). This

2 ‘If you hit a rocking horse with a stick, the movement of the horse will be very different from that of the stick. The hits are the cause of the movement, but the system’s own equilibrium laws condition the form of movement’. Wicksell’s metaphor appears in a footnote to a review of Karl Petander’s ‘Goda och darliga tider’, published in Ekonomisk Tidskrift (Wicksell 1918). 3 Attributed to Robert Lucas in Sargent (2001).

147

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertain Futures

feature, above all others, probably explains these models’ ubiquity in policy organizations. Not least in the light of the crisis, however, the potential pitfalls of these approaches have also become clearer of late. Recently, these models have been subjected to stinging critiques (Romer 2016). One common complaint is that they may not do an especially good job of describing the real world, especially in situations of economic stress. ‘Exhibit one’ is that they offered a spectacularly poor guide to the economy’s dynamics around the time of the global financial crisis. To illustrate, Figure 7.1 plots the range of forecasts for UK GDP growth from 2008 onwards, produced by twenty-seven economic forecasters (including the Bank of England) in 2007, at the dawn of the financial crisis. Three features are notable. First, pre-crisis forecasts were very tightly bunched in a range of one percentage point. The methodological monoculture produced, unsurprisingly, the same crop. Second, these forecasts foresaw a continuation of the gentle undulations in the economy seen in the decade prior to the crisis, the so-called ‘Great Moderation’ (Bernanke 2004). At the time, these damped oscillations seemed to match well the damped harmonic motion of DSGE models. A good crop today foretold an only slightly less good crop tomorrow.

Latest vintage of GDP data

Year-on-year growth 4 3 2

November 2007 Inflation Report forecasts

1 0 –1 –2 –3 –4 –5

2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

Figure 7.1 Range of GDP forecasts in 2007 Q4 Source: Bank of England November 2007 Inflation Report; Bank of England Survey of Economic Forecasters.

148

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertainty in Macroeconomic Modelling

Third, most striking of all, every one of these forecasts was not just wrong but spectacularly so. Few forecasters foresaw even a slight downturn in GDP in 2008 and none foresaw a recession. Yet we witnessed not just a recession but the largest since the 1930s. The one-year-ahead forecast error in 2008 was eight percentage points. The crop failed and the result was economic famine. While forecasting performance has improved since then, there has been a continuous string of serially correlated errors, with the speed of the recovery consistently over-estimated (Figure 7.2). The average forecast error one year ahead has been consistently negative, averaging 0.5 percentage points per year. The average error two years ahead has been over one percentage point per year. At root, these are failures of models, methodologies, and monocultures. It has been argued that these models were not designed to explain such extreme events. To quote Robert Lucas once more: ‘The charge is that the [ . . . ] forecasting model failed to predict the events of September 2008. Yet the simulations were not presented as assurance that no crisis would occur, but as a

Per cent year on year 7 6 5 4 3 2 1 0 –1 –2 2000

2002

2004

Oct-07 Oct-12

2006 Oct-08 Oct-13

2008 Oct-09 Oct-14

2010

2012 Oct-10 Oct-15

2014

2016

Oct-11 Oct-16

Figure 7.2 Forecasts for World Growth by the IMF since 2007 Source: IMF World Economic Outlook.

149

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertain Futures

forecast of what could be expected conditional on a crisis not occurring’ (Lucas 2009). But this is not really a defence. Economics is important because of the social costs of extreme events. Economic policy matters precisely because of such events. If our models are silent about these events, this jeopardizes the very thing that makes economics interesting and economic policy important.

Agent-Based Models (ABMs) Even if some of the post-crisis criticism of workhorse macroeconomic models is overdone, this still raises the question of whether new modelling approaches might be explored that provide a different lens on the world or better match real-world dynamics. One potentially promising strand is so-called agent-based modelling. ABMs are interconnected systems of individual ‘agents’ who follow welldefined behavioural rules of thumb. What makes these models interesting is that these agents are heterogeneous and interactive. In other words, these models relax one of the key assumptions of the standard model, that of a single, representative agent. For social systems, this does not sound too implausible. Humans are social animals. Indeed, the feature that sets humans apart from other animals is their degree of social interaction (Harari 2015). This—on the face of it—modest adaptation gives rise to some fundamental changes in model dynamics. Linear, proportional responses to shocks become the exception; complex, non-linear responses the rule. Single, stationary equilibria become the exception; multiple, evolutionary equilibria the rule. These differences are discussed in the next section, ‘The Costs and Benefits of ABMs’. Before doing so, it is helpful to bring to life some of the uses of ABMs. These models have found widespread application across a broad array of disciplines in both the natural and social sciences. They have been used to address a massive array of problems, big and small: from segregating nuts to segregating races, from simulating the fate of the universe to simulating the fate of a human cell, from military planning to family planning, from flocking birds to herding (fat) cats. Yet in economics, ABMs have been used relatively less (Battiston et al. 2016). Perhaps the most famous application is Thomas Schelling’s work on racial segregation (Schelling 1969, 1971). This demonstrated how, in a simple cellular structure with agents following simple rules of thumb, a pattern of segregation might naturally emerge. The model’s predictions closely matched locational patterns in real cities and communities. In the light of the crisis, interest in ABM methods has blossomed, albeit from a low base. They have been used to study the effects of fiscal and monetary 150

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertainty in Macroeconomic Modelling

policies (Dosi et al. 2015), systemic risk (Geanakoplos et al. 2012), and financial market liquidity (Bookstaber et al. 2015). ABMs of entire economies have also begun to be developed. One example is the Complexity Research Initiative for Systemic Instabilities (CRISIS), an open source collaboration between academics, firms, and policymakers (Klimek et al. 2015). Another is EURACE, a large microfounded macroeconomic model with regional heterogeneity (Dawid et al. 2012). A third is the Complex Adaptive System model, which incorporates bounded rationality and heterogeneity to reproduce business cycles.4 A fourth is the MINSKY model.5 To be clear, ABMs are no panacea for the modelling ills of economics. In discussing them here, the implication is not that they should replace DSGE models, lock, stock, and barrel. Rather, their value comes from providing a different—complementary—lens through which we might make sense of our dappled economic and financial world, a lens that other disciplines have found useful when attempting to understand their worlds or devise policies to improve them.

The Costs and Benefits of ABMs If ABMs have found relatively limited application in economics and finance, despite widespread application elsewhere, does that matter? That depends on the potential benefits, and associated costs, of ABM techniques in understanding economic phenomena. Starting with costs, ABM technology has been transformed over the past decade, for two reasons. First, the cost and speed of running these models has been revolutionized. The largest ABMs can now deal with interactions among 100 billion agents (LLNL 2013). Second, there has been a revolution, every bit as significant, in the availability of data to calibrate these models. At the same time as the costs of developing and simulating ABMs have shrunk, it seems likely that the benefits of developing these models may have become larger. Inter-connections between agents have lengthened and strengthened over recent decades, locally and globally (Haldane 2015). These trends increase the importance of taking seriously interactions between agents when modelling an economy’s dynamics. These benefits can perhaps best be illustrated by drawing out some of the key behavioural differences between ABM and mainstream macromodels. 4

Dilaver et al. (2016). This is being developed by Steve Keen and Russell Standish; further details are available here: http://www.debtdeflation.com/blogs/minsky/. 5

151

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertain Futures

Emergent Behaviour and Uncertainty In standard macromodels, system dynamics are fully defined by the distribution of shocks to the economy and the behavioural parameters determining how they ripple through the system. The classic Frisch/Slutsky impulse-propagation mechanism determines the economy’s fortunes. If the distribution of shocks and the parameters of the model are known and fixed, the dynamics of this system are well-defined and predictable. Complex systems, of which ABMs are one example, do not in general have these properties. The Frisch/Slutsky decomposition is very unlikely to be stable, if it exists at all. The reason is that a complex system’s dynamics do not derive principally from disturbances arising outside the system, but from interactions within the system. Dynamics are endogenously, not exogenously, driven. These feedback effects within the system may either amplify or dampen cycles. They may also give rise to abrupt shifts or discontinuities in system behaviour if pushed beyond a critical threshold or tipping point (Wilson and Kirman 2016). These complex patterns are often referred to as ‘emergent’ behaviours because they ‘emerge’ without any outside stimulus. And because these emergent patterns arise from complex interactions, they are often difficult or impossible to predict. To bring this to life, imagine that instead of a single wooden rocking horse, the system comprised a pack of wild horses. Taking a stick to one of them will generate ‘emergent’ behaviour. It may result in them all staying put, one of them fleeing, or all of them fleeing. And if they do all flee, this will be in a direction, and to a destination, that is impossible to predict. These emergent behaviours depend crucially on behavioural interactions within the system. In the natural sciences, examples of these emergent behaviours are legion.6 These emergent properties of complex systems carry important implications for model-building. In these systems, there is a sharp disconnect between the behaviour of individual agents and the behaviour of the system as a whole. Aggregating from the microscopic to the macroscopic is very unlikely to give sensible insights into real-world behaviour, for the same reason that the behaviour of a single neutron is uninformative about the threat of nuclear winter (Haldane 2015). The simple aggregation of ‘microfounded’ models, rather than being a virtue, may then be a cause for concern.

6 They include the dynamics of sandpiles that ‘self-organize’ as each new grain of sand is added until a tipping point is reached and collapse occurs (Bak et al. 1988); the flocking of migrating birds and fish, whose patterns exhibit complex, and sometimes chaotic, patterns of motion (Macy and Willer 2002); and the dynamics of traffic jams among cars and pedestrians, whose flows are irregular and emergent (Nagel and Paczuski 1995).

152

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertainty in Macroeconomic Modelling

The emergent dynamics of these systems are likely to exhibit significant degrees of uncertainty and ambiguity, as distinct from risk (Knight 1921; Shackle 1979). This uncertainty is intrinsic to complex systems, and makes forecasting and prediction in these systems extremely difficult. There is unlikely to be any simple, stable mapping from shocks through to outcomes, from causes to consequences, from stick to rocking horse. Indeed, in these systems you do not need any shocks to generate variability in the system. This, too, stands in sharp contrast to existing macroeconomic models, which tend to emphasize the identification of exogenous shocks as a key factor in understanding system dynamics.

Heuristic Behaviour Mainstream models in macroeconomics and finance tend to have a fairly sophisticated treatment of risk. Provided the distribution of possible outcomes is reasonably well understood, this risk can be priced and hedged in financial markets. Saving and investment behaviour can then be analysed under the assumption that agents optimize their risk-adjusted decisionmaking (Haldane 2012). A world of radical uncertainty, the like of which arises in a complex system, changes that perspective fundamentally. Uncertainty means it may sometimes be impossible to compute future outcomes. In the language of computer science, behavioural decisions are no longer ‘Turing computable’ (Beinhocker 2006; Velupillai 2000). The relevant Euler conditions, familiar from mainstream macromodels, may not even exist. Facing such uncertainty, many consumers appear to follow simple rules of thumb when deciding their spending, rather than solving a complex intertemporal optimization (Allen and Carroll 2001). Likewise, rather than solving a complex mean-variance optimization, many investors appear to invest passively or to equally weight assets in their portfolios (Gigerenzer 2014). And rather than solve a complex inter-temporal trade-off, monetary policy in practice seems to mimic simple rules of thumb (Taylor 2016). Some would interpret these simple decision rules as irrational, in the sense of being inconsistent with the Euler conditions from standard macromodels. But even the concept of rationality needs careful reconsideration in an environment of radical uncertainty. Rationality can be defined only in relation to the environment in which decisions are made—what some have called ecological rationality (Gigerenzer 2014). Heuristics can be the ecologicallyrational response to radical uncertainty. In ABMs, the behaviour of agents is characterized, not by Euler conditions, but by behavioural rules of thumb. These systems also exhibit radical uncertainty. That means there is a degree of model-consistency in ABMs—heuristics 153

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertain Futures

and uncertainty are mutually consistent. In that sense, the behavioural rules embedded in ABMs are neither as irrational, nor as prone to the Lucas critique, as some critics might imply.

Non-Normal Behaviour In many standard models, the equilibrium of the system is both singular and stationary. There is a natural and unique state of rest towards which the model converges following a disturbance. Wicksell’s rocking horse is not a perpetual motion machine, nor does it turn somersaults. While many models of multiple equilibria exist in economics and finance, they tend, as it were, to occupy the suburbs rather than the city centre. In ABMs, the equilibria that emerge are often non-stationary or multiple, sometimes both. The equilibrium may often be an evolutionary one, a type that often arises in ecological and biological models. The dynamics around this equilibrium are also often highly non-linear, and sometimes discontinuous, with a degree of non-linearity that is state-dependent (Taleb 2014). The combined effect of non-stationary, multiple equilibria and highly nonlinear dynamics makes for non-standard, and often highly non-normal, distributions for the variables in these systems. For example, they are more likely to exhibit excess sensitivity in their fluctuations relative to fundamentals. And they may also be subject to large dislocations or discontinuities. In consequence, they are liable to have much fatter tails than the Gaussian distributions that often emerge from linearized, DSGE models.

Matching the Dappled World It is interesting to ask whether the behaviour exhibited in real-world economic and financial systems is broadly consistent with the patterns that emerge from complex ABMs. One simple, reduced-form way to assess that is to look at the statistical distribution of various economic and financial timeseries for evidence of discontinuity, non-normality, and fat tails. These are properties that, we know, exist in other natural and social systems (Barabási 2005; Turrell 2013). The short answer is ‘yes’—the properties of economic and financial systems differ little from other social—and many natural—systems (Haldane and Nelson 2012). To provide a few illustrations, Figure 7.3 looks at the distribution of a set of economic and financial variables over a time-series dating back at least 150 years, including equity prices, bond prices, GDP, and so on. These historical distributions can be compared with a normal curve calibrated to the same data. 154

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertainty in Macroeconomic Modelling

(a)

Probability density

10–1

10–2

–30

–20

–10

0

10

20

30

Year-on-year change, % Best Gaussian fit

House prices

Probability density

(b)

10–1

10–2

10–3 –40

–20

0

20

40

Year-on-year change, % Best Gaussian fit

GDP (UK, US, DE, Japan)

Figure 7.3 Long-run distributions of economic and financial variables a. UK house prices: 1846–2015, annual b. World GDP growth (across the UK, US, Germany, Japan): 1871–2015, annual

155

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertain Futures

Probability density

(c)

10–1

10–2

–20

–10

0

10

20

Year-on-year change, % Best Gaussian fit

Probability density

(d)

GDP-UK

100

10–1

10–2

10–3

–30

–20

–10

0

10

20

Year-on-year wage growth, % Best Gaussian fit

Figure 7.3 Continued c. UK GDP growth: 1701–2015, annual d. UK wage growth: 1751–2015, annual

156

Wage growth

30

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertainty in Macroeconomic Modelling

(e)

100

Probability density

10–1

10–2

10–3 10–4

–40

–20

0

20

40

Month-on-month change, % Best Gaussian fit

Equity prices

(f)

Probability density

10–1

10–2

10–3

10–4

–150

–100

–50

0

50

100

150

Basis point changes Best Gaussian fit

Corporate bond spreads

Figure 7.3 Continued e. Equity prices: 1709–2016, monthly f. Corporate bond spreads: 1854–2016, monthly

  2 is Notes: For each empirical distribution f(x), a Gaussian fit to f(x) of the form gðxÞ ¼ Aexp  ðxμÞ 2σ2 sought in which A, μ and σ are varied and optimized using the Levenberg–Marquardt non-linear least squares fitting algorithm. Note that to achieve the best fit to the empirical data, the fitted distributions do not integrate to unity. Source: Hills et al. (2016); Bank calculations.

157

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertain Futures

In each case, there is strong evidence of non-normality in the empirical distribution. Specifically, the tails of the historical distribution are often significantly larger than normality would imply. For real variables such as GDP, around eighteen per cent of the data across the United Kingdom, United States, Germany, and Japan fall outside the ‘bell’ curve described by a best-fit normal distribution. For financial variables such as equity prices, around ten per cent of the data fall outside the normal distribution. In both cases, there is evidence of out-sized dislocations in variables, which occur with far higher frequency than normality would imply. This evidence is consistent with, if not proof of, complex system dynamics.

Real-World Applications of Agent-Based Models Given that ABMs potentially better match the moments of real-world data, at least in some situations and in some markets, and given their seeming success in other disciplines, the Bank of England has recently made an investment in them as part of its One Bank Research Agenda (Bank of England 2015). Two pieces of Bank research are described here that have drawn on ABMs in an attempt to better understand two markets and how policy might reshape dynamics in these markets.

UK Housing Market The housing market has been one of the primary sources of financial stress in a great many countries (Jordà, et al. 2014). Not coincidentally, this market has also been characterized by pronounced cyclical swings. Figure 7.4 runs a filter through UK house price inflation in the period since 1846. It exhibits clear cyclicality, with peak-to-trough variation often of around twenty percentage points. Mortgage lending exhibits a similar cyclicality. House prices, like other asset prices, also exhibit out-sized booms and busts. Figure 7.3a plots the distribution of UK house price growth since 1846. It has fat tails, with the probability mass of big rises or falls larger than implied by a normal distribution. For example, the probability of a ten per cent movement in house prices in any given year is twice as large as normality would imply. Capturing these cyclical dynamics, and fat-tailed properties, of the housing market is not straightforward using aggregate models. These models typically rely, as inputs, on a small number of macroeconomic variables, such as incomes and interest rates. They have a mixed track record in explaining and predicting housing market behaviour. One reason for this poor performance may be that the housing market comprises not one, but many sub-markets: a rental market, a sales market, a 158

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertainty in Macroeconomic Modelling Per cent year on year

25 20 15 10 5 0 –5 –10

1846

1866

1886

1906

1926

1946

1966

1986

2006

Figure 7.4 Long-run UK house price growth, 1846–2015 Notes: The figure shows the Hodrick–Prescott trend in annual house price growth data (where lambda = 6.25). Data during WWI and WWII are interpolated. Source: Hills et al. (2016); Bank calculations.

mortgage market, and so on. Moreover, there are multiple players operating in these markets—renters, landlords, owner-occupiers, mortgage lenders, and regulators—each with distinctive characteristics, such as age, income, gearing, and location. It is the interaction between these multiple agents in multiple markets that shapes the dynamics of the housing market. Aggregate models suppress these within-system interactions. The housing market model developed at the Bank of England aims to unwrap and model these within-system interactions and use them to help explain cyclical behaviour (Baptista et al. 2016). Specifically, the model comprises households of three types: • Renters who decide whether to continue to rent or attempt to buy a house when their rental contract ends and, if so, how much to bid; • Owner-occupiers who decide whether to sell their house and buy a new one and, if so, how much to bid/ask for the property; and • Buy-to-let investors who decide whether to sell their rental property and/or buy a new one and, if so, how much to bid/ask for the property. They also decide whether to rent out a property and, if so, how much rent to charge. The behavioural rules of thumb that households follow when making these decisions are based on factors such as their expected rental payments, house price appreciation, and mortgage cost. These households differ not only by type, but also by characteristics such as age and income. An important feature of the model is that it explicitly includes a banking sector—a feature often missing from off-the-shelf DSGE models. The banking sector provides mortgage credit to households and sets the terms and conditions 159

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertain Futures

available to borrowers in the mortgage market, based on their characteristics. The banking sector’s lending decisions are, in turn, subject to regulation by a central bank or regulator. They set loan-to-income (LTI), loan-to-value (LTV), and interest cover ratio policies, with the objective of safeguarding the stability of the financial system. These so-called macroprudential policy measures are being used increasingly by policy authorities internationally (IMF-FSB-BIS 2016). The various agents in the model, and their inter-linkages, are shown schematically in Figure 7.5. This multi-agent model can be calibrated using micro datasets. This helps to ensure that agents in the model have characteristics, and exhibit behaviours, that match those of the population at large. For example, the distribution of LTI or LTV ratios on mortgages are calibrated to match the UK population using data on over a million UK mortgages; and the impact on the sale price of a house of it remaining unsold is calibrated to match historical housing transactions data. One of the key benefits of the ABM approach is that it provides a framework for drawing together and using—in a consistent way—data from a range of sources to calibrate a model. For example, the Bank of England has been making use of the Financial Conduct Authority’s Product Sales Database to get a more granular picture of the mortgage position of households. This is a very detailed database, covering over 13 million financial transactions by UK households since 2005. One of the key features of an agent-based model is that it is able to generate complex market dynamics, without the need for exogenous shocks. In other words, within-system interactions are sufficient to generate booms and busts

Central bank sets caps on LTV, LTI and ICR ratios, and affordability tests

Bank gives mortgages

Ownership market

Rental market

BTL investors

Renters

Owner-Occupiers

Social Housing

Households

Figure 7.5 Agents and interactions in the housing market model Source: Baptista et al. (2016). Reprinted with permission from the Bank of England.

160

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertainty in Macroeconomic Modelling

Value

in the housing market. Cycles in house prices and in mortgage lending are, in that sense, an ‘emergent’ property of the model. Figure 7.6 shows a simulation run of the model, looking at the dynamic behaviour of listed prices, house prices when sold, and the number of years a property is on the market. The model exhibits large cyclical swings, which arise endogenously as a result of feedback loops in the model. Some of these feedback loops are dampening (‘negative feedback’), others amplifying (‘positive feedback’). Baptista et al. (2016) use the simulated data from Figure 7.6 to construct distributions of house price inflation over time. This simulated distribution exhibits fat tails, although not as heavy as the historical distribution. Nonetheless, the model goes some way towards matching the moments of the realworld housing market. The same approach can also be used to examine the impact of various macroprudential policy measures, whether hard limits (such as a LTV limit of eighty per cent for all mortgage contracts) or soft limits (such as a LTI cap for some fraction of mortgages). These policies could also be state-contingent (such as an LTV limit if credit growth rises above a certain threshold). As an example, we can simulate the effects of introducing a LTI limit of 3.5, where fifteen per cent of mortgages are not bound by this limit. This simulation is similar, if not directly comparable, to the macroprudential intervention made by the Bank of England’s Financial Policy Committee (FPC) in June 2014. 1.00 0.95 0.90 0.85 0.80 0.75 0.70 0.65 0.60 0.55 0.50 0.45 0.40 0.35 0.30 0.25 0.20 0.15 0.10 0.05 0.00 77.5 80.0 82.5 85.0 87.5 90.0 92.5 95.0 97.5 100.0 102.5 105.0 107.5 110.0 112.5 Time (years)

Figure 7.6 Model simulations of the housing market Notes: The top line is the list price index, the middle line is the house price index, and the bottom line is the number of years a house is on the market. Source: Baptista et al. (2016). Reprinted with permission from the Bank of England.

161

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertain Futures 45% 40%

Share of total loans

35% 30% 25% 20% 15% 10% 5% 0% 0

0.5

1

1.5

2 2.5 3 3.5 Loan-to-income band Baseline

4

4.5

5

5.5

LTI 3.5/15%

Figure 7.7 Simulated effect of a loan-to-income policy Source: Adapted from Baptista et al. (2016). Reprinted with permission from the Bank of England.

Figure 7.7 looks at the simulated impact of this policy on the distribution of LTI ratios across households, relative to a policy of no intervention. The incidence of high LTI mortgages (above 3.5) decreases, with some clustering just below the limit. With some borrowers nudged out of riskier loans, a greater degree of insurance is provided to households and the banking system. Another advantage of this class of models is that they allow one to simulate the longer run impact once the second round and feedback loops have taken effect. Figure 7.8 shows that the distribution of house price growth narrows under the scenario relative to the baseline.

An Agent-Based Model of Financial Markets A second ABM project looks at behaviour in financial markets (BraunMunzinger et al. 2016). As in the housing market, this involves complex interactions between multiple agents. Again, as in the housing market, these interactions are prone to abrupt dislocations in prices, fattening the tails of the asset price distribution. Figure 7.9 looks at the empirical distribution of daily corporate bond price movements over various intervals, pre- and post-crisis. It is clearly fat-tailed. The dynamics of financial markets are also an area of active policy interest, not least in light of the financial crisis. During the crisis, there were sharp 162

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertainty in Macroeconomic Modelling 40% 35%

Share of total loans

30% 25%

20% 15% 10% 5% 0% –5

–4

–3

–2 –1 0 1 2 House price growth rate (%)

Baseline

3

4

5

LTI 3.5/15%

Figure 7.8 Simulated effect on house price growth Source: Adapted from Baptista et al. (2016). Reprinted with permission from the Bank of England.

100

Probability density

10–1

10–2

10–3

10–4 –8

–6

–4

Simulation

–2 0 2 Daily log-price return Pre-crisis

Crisis

4

6

8 ×10–3

Post-crisis

Figure 7.9 Distribution of corporate bond price returns Notes: The distribution function of daily log-price returns of three periods shown against data generated by a single model run. Source: Braun-Munzinger et al. (2016). Reprinted with permission from the Bank of England.

163

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertain Futures

swings in asset prices and liquidity premia in many financial markets. Since the crisis, there have been concerns about market-makers’ willingness to make markets, potentially impairing liquidity. Such policy questions are not easily amenable to existing asset pricing models. Braun-Munzinger et al. (2016) build a model that seeks to capture some of the interactions between market players that might give rise to these asset price patterns. In particular, the model comprises three classes of agent: a market maker, making two-way prices in the asset; a set of funds trading in the asset, but pursuing distinct trading strategies; and end-investors in these funds. Funds are, in turn, assumed to be one of three types: value traders—who assume that yields converge over time to some equilibrium value, buying/ selling when the asset is under/over-valued; momentum traders—who follow short-term trends on the assumption they persist; and passive funds—who trade only in response to in- and outflows from investors. These interactions are shown schematically in Figure 7.10. The model is based on, and calibrated against, the corporate bond market using microlevel data on 1000 mutual funds. These data can be used to calibrate the size distribution of funds, their trading behaviour, and the links between their performance and in- and outflows from the fund. The interactions among these players give rise to interesting dynamics, some of which are shown schematically in Figure 7.11. For example, imagine a shock to the expected loss on a bond. This reduces demand for that bond by funds holding it and causes a re-pricing by the market maker and momentum selling by funds, generating a further fall in the bond’s price and in the wealth of the funds holding it.

Investor

$ Corporate bond market Value traders

Momentum traders

Market maker

Passive funds

Noise

Figure 7.10 Overview of the corporate bond trading model Source: Braun-Munzinger et al. (2016). Reprinted with permission from the Bank of England.

164

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertainty in Macroeconomic Modelling

Shock to expected loss rate

Funds’ demand for bonds is reduced

Funds’ wealth is reduced

Investors reduce allocation of cash to funds

Price falls as market maker sees reduced demand

Returns to investors fall as a consequence of price drop

Figure 7.11 Transmission mechanism of expected loss rate shock Notes: A schematic showing the feedback loops following a shock to the value of the expected loss rate. The shading in the feedback loop indicate the different market players; funds, the market maker, and the investor pool. Source: Braun-Munzinger et al. (2016). Reprinted with permission from the Bank of England.

This fall in fund performance then gives rise to a second feedback loop, inducing investor withdrawals from those funds that have under-performed, further reducing demand for the bond and amplifying the fall in its price. It is only after some time that the influence of value investors stabilizes the market. Each individual run of the model is like hitting a wild horse once and has an unpredictable outcome. But we can get an idea of the general behaviour of the funds by running the same scenario repeatedly—if you like, hitting the wild horses hundreds of times and looking at their most likely response. The most likely behaviour of funds in this model market is oscillatory, with shocks amplified in the short run and only damped after a period of several hundred days (Figure 7.12). By rolling the dice over and over again, we can also look at the distribution of possible outcomes, as the average may hide extreme behaviour in the tails. For example, if the fraction of passive investors increases, this dampens average changes in bond yields (Figure 7.13). But it also increases the chances of much bigger changes in yields—the tails of the distribution fatten. How can this model help us to understand the dynamics of real-world financial markets and the appropriate setting of policy in these markets? 165

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertain Futures

Change in yield subsequent to shock, %

60 50 40 30 20 10 0

100

500 300 400 Trading days since shock

200

0.35% loss rate p.a.

0.22% loss rate p.a.

600

700

800

0.08% loss rate p.a.

Figure 7.12 Impact of a shock to the expected loss rate Notes: What happens after a shock to the expected loss rate; a sudden change in firms’ expected loss rate on bonds causes both short-term fluctuations in yield, and a new, higher long-term yield. Results presented are the median of 100 individual simulations runs; individual model runs exhibit a range of outcomes. Source: Braun-Munzinger et al. (2016). Reprinted with permission from the Bank of England.

Median change in yield post shock, %

45 40 35 30 25 20 15 10 10

20

30 40 50 60 Passive investment, %

95th percentile

70th percentile

70

80

50th percentile

Figure 7.13 Distribution of outcomes after a shock to the expected loss rate Note: The outcomes for median yield over 100 trading days after a 0.36% loss rate shock. Percentiles indicate the distribution of results taken from 250 simulation runs. Source: Braun-Munzinger et al. (2016). Reprinted with permission from the Bank of England.

166

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertainty in Macroeconomic Modelling

Figure 7.9 compares the actual distribution of corporate bond prices changes with the distribution that emerges from ABM simulations. There is a reasonable correspondence between the two, with fatter than normal tails. The model can be used counter-factually—for example, to assess the impact of a rise in the number of passive or momentum traders relative to value investors. This makes for larger and longer-lived oscillations. So, too, does a reduction in the market-making capacity—for example, lower market-maker inventories—as this amplifies the impact on prices of any shock to fund demand. One topical policy issue is whether constraints might be imposed on some funds to forestall investor redemptions in the face of falling prices and performance. For example, US money market mutual funds experienced such redemption runs during the course of the financial crisis. And, more recently, UK property investment funds also exhibited run-like redemptions following the EU referendum result, which depressed asset prices. The model can be used to assess the impact of different approaches to constraining redemption. Results from Braun-Munzinger et al. (2016) suggest that extending the redemption window from one day to one month would, according to the model, have reduced amplitude of the resulting asset price cycle considerably.

Conclusion In one of his most famous metaphors, Shackle (1972) described the economy as a kaleidoscope, a collision of colours subject to ongoing, rapid, and radical change. Many of our existing techniques for modelling and measuring the economy invoke a rather different metaphor, with the economy a rather colourless, inanimate rocking horse. Both approaches have their place in making sense of the dappled economic and financial world. But, to date, the two have not been given equal billing. Standard modelling approaches may be reasonable in normal times, but are unlikely to capture the rich non-linear dynamics of economies in stress. ABM approaches are better-equipped to make sense of economies and markets in flux and in distress. The global financial crisis is an opportunity to rebalance these scales, to take uncertainty and disequilibrium seriously, to make the heterodox orthodox.

Acknowledgements I would like to thank Jeremy Franklin, Marc Hinterschweiger, Andreas Joseph, Matthew Manning, Rajan Patel, Paul Robinson, Arthur Turrell, and Arzu Uluc for their help in

167

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertain Futures preparing the text. I would also like to thank Rafa Baptista, Karen Braun-Munzinger, J. Doyne Farmer, Zijun Liu, Katie Low, and Daniel Tang for their comments and contributions.

Bibliography Allen, Todd W., and Christopher D. Carroll. 2001. ‘Individual Learning about Consumption’. National Bureau of Economic Research (NBER) Working Paper no 8234. Bak, Per, Chao Tang, and Kurt Wiesenfeld. 1988. ‘Self-Organized Criticality’. Physical Review A 38 (1): pp. 364–74. Bank of England. 2015. ‘One Bank Research Agenda: Discussion Paper’. http://www. bankofengland.co.uk/research/Documents/onebank/discussion.pdf. Baptista, Rafa, J. Doyne Farmer, Marc Hinterschweiger, Katie Low, Daniel Tang, and Arzu Uluc. 2016. ‘Macroprudential Policy in an Agent-Based Model of the UK Housing Market’. Bank of England Staff Working Paper no 619. Barabási, Albert-László. 2005. ‘The Origin of Bursts and Heavy Tails in Human Dynamics’. Nature 435 (7039): pp. 207–11. Battiston, Stefano, J. Doyne Farmer, Andreas Flache, Diego Garlaschelli, Andrew G. Haldane, Hans Heesterbeek, Cars Hommes, Carlo Jaeger, Robert May, and Marten Scheffer. 2016. ‘Complexity Theory and Financial Regulation’, Science 351 (6275): pp. 818–19. Beinhocker, Eric D. 2006. The Origin of Wealth: Evolution, Complexity, and the Radical Remaking of Economics. Boston, MA: Harvard Business School Press. Bernanke, Ben S. 2004. ‘The Great Moderation’. Speech given at the Eastern Economic Association, Washington, D.C. https://www.federalreserve.gov/boarddocs/speeches/ 2004/20040220/. Bookstaber, Richard M., and Mark E. Paddrik. 2015. ‘An Agent-Based Model for Crisis Liquidity Dynamics’. Office of Financial Research Working Paper no 15–18. Braun-Munzinger, Karen, Zijun Liu, and Arthur Turrell. 2016. ‘An Agent-Based Model of Dynamics in Corporate Bond Trading’. Bank of England Staff Working Paper no 592. http://www.bankofengland.co.uk/research/Pages/workingpapers/2016/swp592.aspx. Bronk, Richard. 2009. The Romantic Economist: Imagination in Economics. Cambridge: Cambridge University Press. Burgess, Stephen, Emilio Fernandez-Corugedo, Charlotta Groth, Richard Harrison, Francesca Monti, Konstantinos Theodoridis, and Matt Waldron. 2013. ‘The Bank of England’s Forecasting Platform: COMPASS, MAPS, EASE and the Suite of Models’. Bank of England Working Paper no 471. http://www.bankofengland.co.uk/research/ Pages/workingpapers/2013/wp471.aspx. Cartwright, Nancy. 1999. The Dappled World: A Study of the Boundaries of Science. Cambridge, MA: Cambridge University Press. Coyle, Diane, and Andrew G. Haldane. 2014. ‘Financial Crash: What’s Wrong with Economics?’. Prospect Magazine, 9 December. http://www.prospectmagazine.co. uk/economics-and-finance/financial-crash-whats-wrong-with-economics.

168

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertainty in Macroeconomic Modelling Dawid, Herbert, Simon Gemkow, Philipp Harting, Sander van der Hoog, and Michael Neugart. 2012. ‘The Eurace@Unibi Model: An Agent-Based Macroeconomic Model for Economic Policy Analysis’. Bielefeld Working Papers in Economics and Management no. 05–2012. Dilaver, Özge, Robert Jump, and Paul Levine. 2016. ‘Agent-based Macroeconomics and Dynamic Stochastic General Equilibrium Models: Where Do We Go from Here?’. Discussion Papers in Economics 1/16. Dosi, Giovanni, Giorgio Fagiolo, Mauro Napoletano, Andrea Roventini, and Tania Treibich. 2015. ‘Fiscal and Monetary Policies in Complex Evolving Economies’. Journal of Economic Dynamics and Control 52 (c): pp. 166–89. Geanakoplos, John, Robert Axtell, J. Doyne Farmer, Peter Howitt, Benjamin Conlee, Jonathan Goldstein, Matthew Hendrey, Nathan M. Palmer, and Chun-Yi Yang. 2012. ‘Getting at Systemic Risk via an Agent-Based Model of the Housing Market’. American Economic Review 102 (3): pp. 53–8. Gigerenzer, Gerd. 2014. Risk Savvy: How to Make Good Decisions. London: Allen Lane. Haldane, Andrew G. 2012. ‘Financial Arms Races’. Speech given at the Institute for New Economic Thinking, Berlin, 14 April 2012. http://www.bis.org/review/r120426a.pdf. Haldane, Andrew G. 2015. ‘On Microscopes and Telescopes’. Speech given at the Lorentz Centre Workshop on Socio-Economic Complexity, Leiden, NL, 27 March 2015. http:// www.bankofengland.co.uk/publications/Pages/speeches/2015/812.aspx. Haldane, Andrew G., and Benjamin Nelson. 2012. ‘Tails of the Unexpected’. Paper presented at the conference ‘Credit Crisis Five Years On: Unpacking the Crisis’ held at the University of Edinburgh Business School, 8–9 June. https://pdfs.semanticscholar. org/55bd/31941afbbd1072b0b022971b3ab479db6c0d.pdf. Harari, Yuval Noah. 2015. Sapiens: A Brief History of Humankind. London: Vintage Books. Hills, Sally, Ryland Thomas, and Nicholas Dimsdale. 2016. ‘Three Centuries of Data—Version 23’. http://www.bankofengland.co.uk/research/Pages/onebank/ threecenturies.aspx. Hirschman, Albert O. 1991. The Rhetoric of Reaction: Perversity, Futility, Jeopardy. Cambridge and London: Belknap Press of Harvard University Press. IMF-FSB-BIS. 2016. ‘Elements of Effective Macro-Prudential Policies. Lessons from International Experience’. International Monetary Fund, Financial Stability Board, and Bank for International Settlements. http://www.imf.org/external/np/g20/pdf/ 2016/083116.pdf. Jordà, Òscar, Moritz Schularick, and Alan M. Taylor. 2014. ‘The Great Mortgaging: Housing Finance, Crises, and Business Cycles’, NBER Working Paper no 20501. http://www.nber.org/papers/w20501.pdf. Keynes, John Maynard. 1936. The General Theory of Employment, Interest and Money. London: Palgrave Macmillan. King, Mervyn. 2016. The End of Alchemy: Money, Banking and the Future of the Global Economy. London: Little, Brown Publishers. Klimek, Peter, Sebastian Poledna, J. Doyne Farmer, and Stefan Thurner. 2015. ‘To BailOut or to Bail-In? Answers from an Agent-Based Model’. Journal of Economic Dynamics and Control 50: pp. 144–54.

169

OUP CORRECTED PROOF – FINAL, 31/5/2018, SPi

Uncertain Futures Knight, Frank H. 1921. Risk, Uncertainty, and Profit. Hart, Schaffner, and Marx Prize Essays, no 31. Boston and New York: Houghton Mifflin. Lawrence Livermore National Laboratory. 2013. ‘Record Simulations Conducted on Lawrence Livermore Supercomputer’. Press release, 19 March. https://www.llnl.gov/ news/record-simulations-conducted-lawrence-livermore-supercomputer. Lucas, Robert E. 1976. ‘Econometric Policy Evaluation: A Critique’. Carnegie-Rochester Conference Series on Public Policy 1 (1): pp. 19–46. Lucas, Robert. 2009. ‘In Defence of the Dismal Science’. The Economist, 6 August. http:// www.economist.com/node/14165405. Macy, Michael W., and Robert Willer. 2002. ‘From Factors to Actors: Computational Sociology and Agent-Based Modeling’. Annual Review of Sociology 28: pp. 143–66. Mirowski, Philip. 1989. More Heat than Light: Economics as Social Physics, Physics as Nature’s Economics. Cambridge, MA: Cambridge University Press. Nagel, Kai, and Maya Paczuski. 1995. ‘Emergent Traffic Jams’. Physical Review E 51 (4): pp. 2909–18. Popper, Karl. 1934 [1959]. The Logic of Scientific Discovery, London: Hutchinson & Co. Translation of Logik der Forschung: Zur Erkenntnistheorie der modernen Naturwissenschaft. Mohr Siebeck, 1934. Romer, Paul. 2016. ‘The Trouble with Macroeconomics’. Commons Memorial Lecture of the Omicron Delta Epsilon Society, delivered on 5 January 2016. https://paulromer. net/wp-content/uploads/2016/09/WP-Trouble.pdf. Sargent, Thomas J. 2001. The Conquest of American Inflation. Princeton, NJ: Princeton University Press. Schelling, Thomas C. 1969. ‘Models of Segregation’. American Economic Review 59 (2): pp. 488–93. Schelling, Thomas C. 1971. ‘Dynamic Models of Segregation’. Journal of Mathematical Sociology 1 (2): pp. 143–86. Shackle, George L.S. 1972. Epistemics and Economics. Cambridge: Cambridge University Press. Shackle, George L.S. 1979. Imagination and the Nature of Choice. Edinburgh: Edinburgh University Press. Smets, Frank, and Raf Wouters. 2003. ‘An Estimated Dynamic Stochastic General Equilibrium Model of the Euro Area’. Journal of the European Economic Association 1 (5): pp. 1123–75. Taleb, Nassim N. 2014. Antifragile: Things That Gain from Disorder. New York: Random House Trade Paperbacks. Taylor, John B. 2016. Policy Stability and Economic Growth: Lessons from the Great Recession. London: Institute of Economic Affairs. Turrell, Arthur E. 2013. ‘Processes Driving Non-Maxwellian Distributions in High Energy Density Plasmas’. PhD thesis, Imperial College London. Velupillai, Kumaraswamy. 2000. Computable Economics: The Arne Ryde Memorial Lectures. Oxford: Oxford University Press. Wicksell, Knut. 1918. ‘Goda och darliga tider’. Ekonomisk Tidskrift 19: pp. 66–73. Wilson, David S., and Alan Kirman. 2016. Complexity and Evolution: Toward a New Synthesis for Economics. Cambridge, MA: MIT Press.

170

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Section III The Role of Narratives and Planning in Central Banking

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

8 A Tractable Future Central Banks in Conversation with their Publics Douglas R. Holmes

Introduction This chapter examines how central bankers seek to endow the future with discernible features that we—the public—can reflect and act upon, animating or curtailing our propensities to produce, consume, borrow, and lend. It argues that central bankers, rather than predicting the future, seek to create elements of a tractable future. They do this with words. They use language to explore, promulgate, and sustain the ideas that animate our economic future, as well as the structures of feeling, the sentiments, expectations, and desires that make them real. The future is, in the first instance, a technical problem for central bankers—‘the inter-temporal problem’—upon which the basic challenge of monetary affairs hinges: by what means is the value of money to be anchored over time? The chapter builds on material from Economy of Words: Communicative Imperatives in Central Banks (Holmes 2014a) and subsequent publications (2014b, 2015, 2016) in an effort to stabilize a series of analytical perspectives on the evolving operation of central banks and the shifting nature of monetary affairs. It revisits some reliable illustrative examples—ethnographic vignettes—that are presented in abbreviated form as paradigmatic cases. While the research on which this chapter is based has been informed by conversations with senior officials of central banks, it is focused predominantly on the people who do the routine work within these institutions. It was from conversations with them, and from the speeches, reports, graphic images, and other written documents they produce, that key insights were drawn and refined. The research demonstrates how these people apply

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

rhetorical acumen to the high-level mathematical reasoning and quantitative analysis that impart institutional authority to the work of central banks; and, how they model the economy and the financial system with language establishing a radically communicative and relational dynamic at the centre of monetary affairs. Here is a brief aside by Ben Bernanke—one of the key architects of the contemporary monetary regime—that captures the rhetorical nature of monetary policy, while simultaneously posing the key questions of, and for, a narrative economics:1 When I was at the Federal Reserve, I occasionally observed that monetary policy is 98 percent talk and only two percent action. The ability to shape market expectations of future policy through public statements is one of the most powerful tools the Fed has. The downside for policymakers, of course, is that the cost of sending the wrong message can be high. Presumably, that’s why my predecessor Alan Greenspan once told a Senate committee that, as a central banker, he had ‘learned to mumble with great incoherence’. (Bernanke 2015)

In the informal genre of a blog post—indeed his first blog post—the former chair of the Federal Reserve system broached, albeit wryly, the deepest questions about the nature of contemporary monetary affairs. He asserted that monetary policy is managed not solely or necessarily by conventional levers that central bankers employ to set interest rates and regulate the availability of money and credit, but by ‘talk’. This assertion, of course, begs a series of other questions: what is the nature of this talk, where does it come from, how does it work? Further still, it opens vertiginous questions of how markets, particularly financial markets, operate as a function of language (relatedly, see Riles 2011). Talk is action. But who is listening? Bernanke is suggesting that there is an audience for this talk, an audience that is not merely overhearing policy pronouncements, but enacting them prospectively. This talk is not simply a descriptive genre for the representation of economic and financial conditions; it is the substance of policy (Yellen 2013). By drawing attention to the now famous aside by his predecessor, Alan Greenspan, Bernanke contrasts his attitude towards communication with the traditional stance of central bankers as resolutely secretive, as figures who cultivated opacity and viewed talk, clear unambiguous talk, as antithetical to the effective exercise of their statutory duties. Bernanke’s modest blog post also disrupts fundamental assumptions of economics, not least of which is how economic ideas are created and by whom, as well as how these ideas are integral to the operation of the economy and not sequestered from it in the realm of academic scrutiny. 1

174

This vignette is drawn from Holmes (2016).

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

A Tractable Future

Talk implies a conversation and thus, from the standpoint of neoclassical economics, it represents an anomaly insofar as it spawns forms of social relations as instruments for both gleaning information and the exercise of policy (McCloskey 1985). This talk has an unusual reach, demarcating conversations sustained by vast, global networks of interlocutors in which distinctive forms of knowledge are circulated relentlessly. The talk also reaches down to the deepest levels of quantitative research within central banks, to the technical operation of the large macroeconomic models and the scenarios they generate: variables are critically scrutinized, theory continually reassessed, and layer upon layer of contextual information added discursively. The economy and financial systems undergo a continuous and relentless linguistic intermediation. Perhaps most importantly for this research, language is used experimentally to explain and articulate the novel contingencies defining central banks’ relationship to the market and to the public. Far more than the price of money is at stake in central bankers’ narratives: talk is their two-way bridge to the sociological and political, and to the entrepreneurial relationships within which creative economic action is planned and orchestrated (Miyazaki 2013; Nelson et al. 2008; Nelson and Katzenstein 2010). Thus coextensive with the market is an expansive communicative field across which words and ideas circulate, and within which the policy pronouncements thereby informed are reflected and acted upon (Akerlof and Shiller 2009; Beckert 2016; Bronk and Jacoby 2016; Searle 1969; and see also Reichmann on the interactional basis of forecasting in this volume).

A Shift in Both Governability and Communication Paradigms In the midst of the financial crisis and its aftermath, central bankers found themselves compelled to address one of the fundamental questions in the contemporary social and behavioural sciences: how far do we need to redefine the distinctions and discursive relationships between various categories of ‘market participants’ and various strata and segments of ‘the public’? This question coincides with a fundamental shift in the nature and operation of monetary policy that commenced in late 2008. Benjamin Braun (this volume) has analysed these transformations: ‘Keynesian’ fiscal demand management and monetary inflation targeting can be conceptualized as separate ideal-types of macroeconomic state agency: the former operates in a hydraulic manner, while the latter is strategic and performative. Crucially, however, recent quantitative easing programmes mark a return to the hydraulic style of macroeconomic governance, albeit via the monetary rather than the fiscal authority, and via the market for financial assets rather than the markets for goods and services. (p. 195–6).

175

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

He further notes: Keynesian demand management policy involves the government expanding its balance sheet to purchase goods and services in order to stimulate the (real) economy directly; QE involves the central bank expanding its balance sheet to purchase financial assets in order to stimulate the financial economy directly (with an intended second-round stimulus effect for the real economy). In other words, central banking has acquired what had previously been the exclusive domain of fiscal policy—hydraulic macroeconomic agency. (p. 209).

This tectonic shift in monetary affairs has fundamentally altered the conversations between central bankers, market participants, and the public. Crucially, discursive practices have been repurposed to facilitate precisely the transformations—the shifts in the ‘governability paradigm’ from ‘performative’ to ‘hydraulic’—that Braun so persuasively describes (see also Borio 2017; Gabor and Jessop 2015). The transmission of monetary policy operates via two communicative channels: broadly speaking, information on interest rates—the policy rate— format expectations on the pricing of risk, while the information conveyed by the ‘monetary-policy story’ creates the relational elements of a tractable future. The former largely speaks to market participants; the latter underwrites enduring two-way relationships—based on confidence and trust—with both investors and the public (Haldane 2015; Peter J. Katzenstein, personal communication). More broadly, this chapter seeks to demonstrate why members of the public and market participants should be viewed as ‘protagonists’ in central banking, who simultaneously enact the economy according to policy pronouncements and enliven (or inform) the public sphere by engaging in conversation with policymakers. The economy changes conceptually and practically over time in profound ways. The conceptual tools—the variables—by which policymakers explain, interpret, and model economic and financial phenomena are transitory, if not fugitive. As a result, central bankers—like other policymakers—can reconfigure and stabilize their own understanding of the uncertain futures they face only through sustained conversational interaction with the market and the public. Such understanding gained through exploratory conversation is a precursor to the making of successful policy pronouncements designed to guide the expectations of the public in a desirable direction.

Prototype The New Zealand experiments with inflation targeting, which commenced in the late 1980s, were remarkable insofar as a group of young economists 176

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

A Tractable Future

working within the Reserve Bank of New Zealand sought in the midst of an acute financial crisis to design a central bank and its monetary-policy framework ‘from scratch’ (Arthur Grimes, personal communication). These economists sought to address the key problem posed for discretionary monetary policy: can policymakers influence expectations not merely about the future, but in the future and thereby shape and format economic behaviour prospectively? Their answer to this ‘inter-temporal problem’ rested on communication (Bernanke et al. 1999). The New Zealand framework and the logic impelling its experimental dynamics can be summarized as follows: if the behaviour of prices is ‘expectational’—as Irving Fisher, J. M. Keynes, Knut Wicksell, and others had proposed many decades earlier—then an anticipatory policy that projects central bank action into the future becomes a means to influence these sentiments. The instruments developed to manage expectation are expressed most concisely in official statements—typically running to between 500 and a few thousand words—which the major central banks of the world publish periodically in support of their interest-rate decisions. Rounds of speeches and press conferences by senior personnel of central banks elaborate and explain policy statements in relation to research and analysis on the trajectory of economic and financial conditions. These ‘macroeconomic allegories’, as Alan Blinder and Ricardo Reis (2005) term them—in a clear evocation of the persuasive labour these narratives are called upon to perform—draw on the full intellectual resources of these institutions, namely, the research acumen, the judgement, the experience, and the rhetorical skill of their personnel. They project a forecast of economic and financial conditions over a time horizon of approximately one to two years, along with an explanation of how the respective banks’ interest-rate policy will achieve particular outcomes. As economic agents assimilate policy intentions as their own personal expectations, they do the work of the central bank. Expectations (guided by central bank rhetoric) can thus influence the course of inflationary and deflationary processes independent (or in anticipation) of conventional interventions on interest rates, in a form of self-fulfilling prophecy (Merton 1948). The bridge to the ephemera of expectations—expectations that shape economic behaviour prospectively—is constructed with words, demonstrating, inter alia, how markets themselves are discursive phenomena—essentially ‘markets in stories’ (Beckert 2016, 148). Rational expectations theory assumes that economic agents effectively internalize the correct model of the economy (Muth 1961, 316; Beckert and Bronk, this volume). According to this theory, if central banks have a role, it is in coordinating the efficient gravitation of agents’ expectations towards this correct model. This chapter, by contrast, seeks to demonstrate that the narratives (and other policies) that central banks use to cajole expectations 177

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

in a particular direction are themselves based in part on contingent and exploratory discursive interactions with the public they serve. That is to say, the economic narratives (and associated policies) of central bankers are creatures of repeated conversation with others used as a discursive input to dynamic judgements about how to interpret the fluid and indeterminate socio-economic context that policymakers face. In other words, what we are dealing with here is not the performativity of theoretical models (as in MacKenzie 2006; Callon 2007), but rather the performativity of contingent narratives that are themselves the product of sustained conversation with a range of socio-economic interlocutors.

Forward Guidance Since the onset of the financial crisis, the major central banks of the world have undertaken a series of interventions, most notably using their balance sheets to address the fundamental conditions of the crisis, while maintaining interest rates at or close to zero for extended periods. ‘Quantitative easing’ (QE), the much commented-on effort to buttress vulnerable banks and financial institutions, support asset prices, thwart deflationary pressures, and thereby encourage economic stability and growth, has been the most prominent and controversial of these interventions (see, for example, Blinder 2013, 248–56). These policy actions were preceded by what is termed, prosaically, ‘forward guidance’, encompassing systematic communications that carefully explained the nature of policy interventions and their intended purpose (Fischer 2016; Woodford 2012). The rationale for forward guidance in the wake of the financial crisis is well explained in the following: [W]hen interest rates reach their effective lower bound, central banks should indicate their future intentions more precisely than they do under normal circumstances. Because bondholders are exposed to a risk of capital loss due to the uncertainty about the timing and magnitude of the rise in short-term interest rates that would accompany a future recovery, it is argued that central bank communication could provide insurance against that risk. (Coeuré 2013)

Communications were thus not just intended as an adjunct to policy, but as the decisive means to achieve the ends of policy. While interest rates were held unchanged for years, a series of ‘monetary stories’ evolved to address shifting policy priorities, to explain policy interventions, and, above all, to address the mercurial substance that is glossed, perhaps too simplistically, as ‘confidence’. In this process, models such as Michael Woodford’s Interest and Prices (MWIP) model attain rhetorical expression for the purposes of influencing 178

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

A Tractable Future

economic outcomes. By translating Woodford’s theoretical insights, rendering them intelligible to members of the public and to market participants, a distinctive means of managing monetary affairs prospectively is created. By articulating the central bank’s ‘reaction function’—its intentions to act in specific ways in response to continuously updated data and information—a system for shaping and formatting expectations is set in motion. Again, as we assimilate these intentions as our own basis of planning and action, we do the work of the monetary authority, and we participate in a particular form of collaborative governance. The Bank of England, for example, made the contingencies of its reaction function explicit by communicating its intention in 2013 to hold temporarily in abeyance its statutory obligation to target inflation and substitute a measure of unemployment as a target and thus as the fulcrum for determining the future course of monetary policy. The decisive point is that the Bank made this forward guidance part of a wide-ranging public conversation; and the efficacy of the model depended in part on the power of communications, like this one, to shape expectations and format behaviour. Monetary policy trade-offs and forward guidance, August 2013 At its meeting on 1 August 2013, the Monetary Policy Committee (MPC) agreed its intention not to raise Bank Rate from its current level of 0.5% at least until the Labour Force Survey (LFS) headline measure of the unemployment rate had fallen to a ‘threshold’ of 7%, subject to the conditions below. The MPC stands ready to undertake further asset purchases while the LFS unemployment rate remains above 7% if it judges that additional monetary stimulus is warranted. But until the unemployment threshold is reached, and subject to the conditions below, the MPC intends not to reduce the stock of asset purchases financed by the issuance of central bank reserves and, consistent with that, intends to reinvest the cash flows associated with all maturing gilts held in the Asset Purchase Facility. This proposition linking Bank Rate and asset sales to the unemployment threshold would cease to hold if any of the following three ‘knockouts’ were breached: • in the MPC’s view, it is more likely than not that CPI inflation 18 to 24 months ahead will be 0.5 percentage points or more above the 2 per cent target; • medium-term inflation expectations no longer remain sufficiently well anchored; • the Financial Policy Committee (FPC) judges that the stance of monetary policy poses a significant threat to financial stability that cannot be contained by the substantial range of mitigating policy actions available to the FPC, the Financial Conduct Authority and the Prudential Regulation Authority in a way consistent with their objectives. In essence, the MPC judges that, until the margin of slack within the economy has narrowed significantly, it will be appropriate to maintain the current exceptionally

179

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures stimulative stance of monetary policy, provided that such an approach remains consistent with its primary objective of price stability and does not endanger financial stability. (Bank of England 2013)

This brief communiqué, running to 300 words, recapitulates many of the key innovations in monetary policy over the past four decades; indeed, the statement itself is an instrument of these innovations. Why are these communications so vital? Because the audiences for these pronouncements are not merely served by monetary policy, they enact it.

Monetary-Policy Story In the early 2000s, while observing the analytical labour performed by modellers working in the Bundesbank, the author noticed the dynamic interplay between high-level quantitative analyses and the crafting of macroeconomic allegories. The array of sophisticated quantitative models employed for the purposes of policy formulation were, in fact, designed to tell stories in which the economy was modelled discursively. These quantitative instruments— the ‘machineries of knowing’, as Karin Knorr Cetina (1999) termed them— also operated as ‘machineries of relating’ capable of articulating policy in relation to the distinctive and shared predicaments of various segments of the market and diverse strata of the German public. The ‘monetary-policy story’ was able to bridge, as it were, the multiple registers of knowing and relating. Graham Smart described the social life of the monetary-policy story as the template for collaboration within the Bank of Canada: The monetary-policy story is constructed in three stages, over time and across a set of written genres, with each successive version offering a broader knowledge claim in the form of a more comprehensive account of the state of the Canadian economy. The narrative appears in the first stage as a cluster of what I refer to as sector stories, specialists’ analysis of developments in different sectors of the economy; in the second stage, as a more encompassing, although still somewhat circumscribed, narrative about the Canadian economy as a whole, produced by a team of economists during a quarterly activity known as the Projection Exercise and inscribed in a document called the White Book; and then in the final stage, as a fully elaborated institutional story, constructed by executives from the White Book and other sources of information. (Smart 1999, 257)

But in its final manifestation the monetary-policy story, he notes, ‘is nowhere completely articulated in written form in any internal document; rather, it resides in the executives’ discourse, surfacing in meetings and informal 180

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

A Tractable Future

conversations . . . and underlying certain assumptions . . . in the texts they produce’ (1999, 266). These actors freely acknowledge storytelling as implicit in negotiating various levels of formal empirical analysis, notably in reference to the Bank’s econometric model of the Canadian economy, the QPM. This ecology of discourses fosters the assimilation of feelings, intuition, discretion, and judgement, reaching into the reserves of experience within these institutions, and sustaining ‘the intersubjectivity—the ground of shared understandings—that makes possible the intellectual collaboration of the bank’s economists’ (1999, 256). Storytelling thus constitutes a creative and relentless simulation of the economy orchestrating inter-temporally the thinking both shared and contested among personnel of the Bank. To what end? Insofar as storytelling constitutes the machinery of relating, it establishes a critical purview on the social foundations of the contemporary monetary regime. As Benoît Coeuré, Member of the Executive Board of the European Central Bank, has noted: ‘Commitment to monetary stability is not only grounded in its economic merits but is also a cornerstone of the social contract’ (Coeuré 2013).

Machineries of Relating In early 2009, when the author was in Stockholm conducting research at the Riksbank, there was a glimmer of hope that the first dire challenges of the financial crisis had been addressed and that a range of catastrophic outcomes, at least for the time being, had been avoided.2 Unsurprisingly, among six members of the Board of the Riksbank—composed of the Governor and five deputy Governors—critical discussion was developing about interest rates. But these discussions turned on an unusual question: how are monetary affairs managed at the ‘zero lower bound’; that is, the point when the bank’s policy rate, the ‘repo rate’, approaches zero? In other words, if stimulative monetary policy is achieved by lowering interest rates, what alternatives are available when interest rates are at or close to zero? Can a policy rate be set below zero? The key figure in this discussion was Lars Svensson, at the time a Deputy Governor of the bank. Svensson had written extensively about this problem while a member of the economics department at Princeton University, focusing in particular on the monetary conundrum of Japan and its ‘lost decades’ mired in deflation and a banking system caught in a ‘liquidity trap’. Already well-versed in Svensson’s scholarly works and the speeches he had delivered, 2 For a full account of the Riksbank case study, see chapters 9 and 10 of Economy of Words (2014) and Holmes (2015).

181

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

the author was fortunate enough to have a conversation with him on his ‘foolproof ’ strategy for avoiding deflation (Svensson 2009). It was by any standards a remarkable and creative application of macroeconomic theory to an overriding concern posed by the financial crisis. He had a series of policy remedies to hand if Sweden was to encounter a serious episode of deflation and he articulated them forcefully. When Professor Svensson entered the boardroom of the Riksbank to discuss policy matters his analytical work in monetary economics took on a different significance, his confidence in his ‘foolproof ’ method notwithstanding. When he introduced his academic insights to the discussions with other members of the bank’s board, engaging their perspectives, a different means and method for modelling the Swedish economy unfolded. The deliberations of the Riksbank’s Executive Board represent not merely a simulation of abstract economic conditions but, again, an articulation of the social foundations of the contemporary monetary regime. Technical matters became acute social concerns, as the six members of the board confronted an audience, the Swedish public—an immanent public—whose members had to be persuaded by board members’ insights and commentary. Controlling the parameters of the monetary-policy story and communicating them with care and circumspection served as the vehicle for imparting stability at a time of considerable uncertainty. The conversation among the members of the committee was unusually expressive at the April 2009 meeting. No particular data series, no particular variables, were decisive in their deliberations. Rather, members carefully contextualized the overall economic situation as a framework for instilling confidence among themselves and hence, in due course, the public. Indeed, because they were performing this intellectual labour inside one of the key institutions charged with managing economic and financial affairs, their representational labour, their conversation, would itself constitute the material shaping a dynamic contextual and situational field (Abolofia 2010). A key task of this storytelling was to stabilize sentiments and expectations rather than fully eradicate fear. Fear and uncertainty were entirely appropriate elements of the story at that time. But what was needed was a story—or, more accurately, multiple narratives—within which a broad range of more nuanced sentiments and expectations could be evoked and brought to bear on the particular circumstances that prevailed in early 2009. The experience and judgement of board members was key to providing a range of interpretative insights to bear on the available data—data that were unusually difficult to read at the time. Significantly, board members recapitulated the history of similar crises and the means and manner by which they were resolved. Casting the monetary-policy story prospectively was decisive. Teasing out how the circumstances of the moment could be read in multiple ways 182

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

A Tractable Future

consistent with information, judgement, and experience meant that the issues at the heart of the crisis became tractable. Their forward-looking appraisals were capable of orchestrating those leaps of faith by which investment, employment, and consumption plans by firms, household, and individuals could become the basis of action, when rational appraisals of risk and of rewards were largely or entirely incalculable. The troubled economy and the dysfunctional financial system assumed the guise of shared predicaments, which the Riksbank’s Board understood were ultimately in the hands, the visible hands, of its target groups—the public—to creatively resolve. During this period, the author also had a conversation with the Governor of the Riksbank, Stefan Ingves. Ingves provided an account of the crisis and the bank’s efforts to grapple with various contingencies. During the conversation there was an awkward moment when the governor asked the author for his view of the bank’s policy. Initially taken aback by the question, the author replied that he was not in a position to critically appraise or evaluate the technical aspects of the bank’s monetary policy, not least because he understood the deliberative procedures by which these decisions were made. Setting the repo rate was not a matter of mere opinion, but of systematic research and carefully orchestrated conversations. At this point, the author realized quickly that he was actually being asked a different question. Governor Ingves was not soliciting the author’s position on the interest-rate policy per se. Rather, he was asking: was the policy legible? Was the policy story consistent with the relationship the bank had carefully cultivated with the Swedish public over the course of decades? These were central questions for the research project elaborated in this chapter. Ingves was also asking implicitly if his colleague’s ‘foolproof strategy’—with its hydraulic style of macroeconomic management at that zero lower bound— exceeded the load-bearing capacity of the monetary-policy story.3 The parsing of Governor Ingves’ question presented here was possible only because of other conversations the author had had at the bank, specifically on the nature of the Riksbank’s political mandates and accountabilities. Ingves was expressing a concern that was not reducible to economic theory, but predicated on essentially constitutional accountabilities aligned with broader public interests. Addressing the public demanded a story that acknowledged the public’s role in the monetary drama. The bank had more than mere information or theory to offer; it had a relationship with the public, cultivated over many years.

3

In February 2015, the Riksbank shifted policy to a negative repo rate along with a program of quantitative easing involving the purchase of Swedish government bonds, a strategy broadly in line with Svensson’s longstanding recommendations (Sveriges Riksbank 2015).

183

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

The public entered the conversation at precisely the moment when the Riksbank’s Board was contemplating measures that encompassed an alternative style of macroeconomic governance, based on negative interest rates and quantitative easing. Discursive practices were thus rearticulated to address the circumstances of the crisis and the dramatic changes in policy—the shift in the governability paradigm—that members of the board were contemplating at the time. The narrative innovations overheard initially by the author in Stockholm—innovations that encompassed the Riksbank’s legal and regulatory accountabilities—took on a profound ‘constitutional’ significance in subsequent months, notably at the height of the euro crisis.

‘Whatever it Takes’ When Mario Draghi, President of the ECB, in July 2012, made his now legendary declaration to ‘rescue’ the euro, he did it with three words: ‘whatever it takes’ (2012a). The power of words to restore the viability of a distressed currency in a matter of minutes, if not seconds, was, however, inseparable from a constitutional gambit.4 Coherence was key. Draghi made it crystal clear that he had trillions of euros at his disposal, if needed, to stem the euro crisis and by communicating this message he achieved his goal: restoring confidence without—at least initially—the expenditure of a single euro. Draghi avowed that there was a paramount authority, the ECB’s constitutional mandate— formally inscribed in the Maastricht Treaty—to manage monetary policy across the eurozone, which overruled any restrictions on ‘unorthodox’ interventions. Perceived convertibility risk, as expressed by the interest premia imposed by the bond market, was a clear and undeniable challenge to the ECB’s management of interest rates. Restoring the ECB’s authority demanded that these differentials be eliminated, not merely to relieve the financial burdens imposed on specific member states, but to restore the ECB’s control over the transmission of monetary policy, which would thereby re-establish the credibility of the common currency. The purchase of short-dated bonds by the ECB under what came to be known as ‘outright monetary transactions’ (OMTs) was deemed by Draghi (2012b) as the means to accomplish this constitutional gambit. The ECB, in Draghi’s view, had the right—indeed, the constitutional obligation—to undertake any initiatives necessary to retain its control over monetary policy, and this assertion of the Bank’s legal authority was translated into a mantra that began to be repeated by senior officials of the Bank, 4 This section is an abbreviated account drawn from chapter 13, ‘Manifesto for a Public Currency’, Economy of Words (Holmes 2014a).

184

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

A Tractable Future

that the ECB would ‘take whatever measures necessary’ to preserve and protect the viability of the common currency. The ECB struggled repeatedly to resolve or at least defer vexing constitutional matters in order to formulate meaningful and effective policy; the euro crisis was and is inseparable from a series of constitutional predicaments. But this is hardly a surprise; the interplay between monetary affairs and fundamental constitutional exigencies was integral to the development of the Bundesbank, the central bank upon which the institutional architecture of the ECB was modelled. The founding principles of the Bundesbank—most notably its political independence—were intended explicitly to address the constitutional exigencies at the heart of the Bundesrepublik: the governance of monetary affairs by means of an enduring relationship with the German people. The notion of performativity, in the scholarship of Donald MacKenzie and Michel Callon, famously raises the question: how does economic theory shape and format behaviour prospectively, rather than merely represent it (MacKenzie 2006; MacKenzie et al. 2007; Muniesa and Callon 2007)? By seeking to influence expectations and thereby behaviour through their chosen stories (and the models embedded in those stories), central bankers embrace a variant of performativity (though they typically do not use the term) as foundational to monetary policy. It is a variant that draws on particular strands of macroeconomics to achieve explicit policy outcomes.5 As Benjamin Friedman notes ‘the language in which . . . [the monetary policy] debate takes place exerts a powerful influence on the substance of what participants say, and eventually even over what they think’ (2002, 6–7). Draghi, however, was not invoking macroeconomic theory as the primary rationale for his intervention. He had another, perhaps more formidable narrative. To re-establish the viability of the euro Draghi had to craft a constitutional story to underwrite the currency with faith and credit. He targeted those provisions of the Maastricht Treaty—establishing the euro as a matter of European law—as a pretext for his intervention. It was not a macroeconomic allegory per se, but a legal narrative that restored the currency’s viability in the blink of an eye.

‘Primordial Responsibility’ Shortly after Mark Carney was named Governor of the Bank of England in 2012, he embarked on a wide-ranging reorganization of the Bank in response ‘Goodhart’s law’ (Goodhart 1975) is related to how performativity is subverted in the practice of monetary policy; ‘any observed statistical regularity will tend to collapse once pressure is placed upon it for control purposes’. 5

185

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

to a series of reports regarding the Bank’s failures and shortcomings in the wake of the crisis (Bank of England 2014). The reorganization—The One Bank agenda—was predicated on the following critical insights and reflections. In particular, it [the Bank of England] failed to recognise that financial stability is as important an objective of macroeconomic policy as price stability, and it downplayed the interrelationships between the two. And it failed to recognise that central banks have a vital role to play in maintaining financial stability because of the deep underlying connection between it and monetary stability. Both are fundamentally about maintaining the public trust and confidence in money and financial intermediation that are essential for them to oil the wheels of commerce. That trust and confidence can be undermined through a loss of certainty about the future value of money, a loss of confidence in financial intermediaries, or ultimately a loss of faith in the financial system. Central banks have a primordial responsibility to act as guarantors of trust and confidence in money because of their status as monopoly issuers of currency. This naturally gives them control over the quantity of money and interest rates – monetary policy. It also means that a core part of financial stability policy – acting as lender of last resort to private financial institutions at times of financial stress – falls naturally to central banks. (Carney 2014, 4)

What Governor Carney’s reorganization achieved was the coordination of financial stability, macroprudential management, and monetary policy within a single, independent monetary authority. It represented a major organizational (and political) achievement for Carney and a significant aggrandizement of power by the Bank of England. It was in the first instance justified in relation to the conventional rationale for central bank independence: The case for the independent operation of monetary policy is firmly established around the time-inconsistency of governments with horizons dictated by the electoral cycle. That time-inconsistency argument applies even more strongly to both microprudential and macroprudential policy given the large potential size and long duration of credit cycles. The avoidance of potentially unpopular measures to boost the resilience of the financial system today can have disastrous consequences many years later. Moreover, being tough and avoiding crises has no obvious reward. It is hard to be given credit for a counterfactual. (Carney 2014, 12)

But this conventional technocratic argument for independence, based on resolution of the ‘time-inconsistency problem’, was inadequate and Governor Carney and his colleagues sought to craft a very different, ‘primordial’ basis of legitimacy:

186

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

A Tractable Future The best answer to the question of what the Bank of England is for is given by the original 1694 founding charter: ‘promoting the good of the people of the United Kingdom’. . . . That mission is timeless. The understanding of what we should do to achieve it has evolved. In 1694 promoting the good of the people meant financing a war with France. During the Great Moderation, it meant price stability. Today, reflecting the lessons of the ensuing financial crisis, it means maintaining both monetary and financial stability. (Carney 2014, 4)

The Bank thus began a search, so to speak, for ‘the people’ of the United Kingdom—an effort to articulate a relationship of accountability between the Bank and various segments and strata of the British public. They pursued this agenda as part of a wide-ranging public discussion, to one part of which the author was party, namely ‘The Open Forum’, held in November 2015. The following is an excerpt from the agenda, which sought as its primary conceptual challenge to reconcile the public and the market: To promote prosperity, markets must meet two conditions: they must be effective – ensuring competitive pricing and proper allocation of capital and risks; and they must maintain their social licence – the consent of society to operate and innovate. . . . The foundations of social licence are fairness and accountability. It is not enough for market participants to meet the letter of regulations – they must act, and be seen to act, in accordance with the spirit of standards and codes, and the values of society, if they are to merit its trust in turn. (Open Forum June 2015, Bank of England, 4–5)

The Bank sought to establish a robust basis of governance by means of a ‘social licence’, one which, unsurprisingly, ratified its autonomy. The paradox of central bank independence is that these institutions must articulate a meaningful rationale for the perpetuation of their independence to the public, a public that the Bank itself must continually address (Dewey 1927 [1991]; Lippmann 1927 [2002]). The Open Forum was impressive—and fully archived on the Bank’s website—drawing together virtually all the Bank’s ‘stakeholders’ in serious discussions on its role, namely, the responsibilities and accountabilities of its officials. One might be inclined to dismiss these intense, short-lived conversations as mere public relations, as a PR exercise to highlight the Bank’s new organizational profile. Perhaps that is true, but the author heard something else, as a participant. The forum recapitulated a key mode of economic thinking, which is registered in one of the lesser-known domains of the Bank’s research activities, its use of regional agencies, an area in which discursive practices are decisive and conversations assume the status of empirical facts, in a manner foreshadowed by Deirdre McCloskey (1985).

187

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

The Network Central bankers employ carefully constructed networks as a means to glean backstories incorporating an alternative epistemic framework to render the economy and the financial system legible. In this way, they draw on stories continually generated outside the central bank from situated actors who are themselves orchestrating and evaluating economic and financial conditions. Central bankers have developed sophisticated discursive techniques to assimilate these stories—and the contextual information they contain—into their communicative regimes and their policy frameworks. The Bank of England’s ‘network’ is accessed by means of its staff of ‘agents’ spread across twelve regional agencies. The network is composed of approximately 9000 contacts in the business and financial communities, as well as in governmental and non-governmental agencies, and the regional agents interview 700 or so of these contacts each month. The contact pool is selected ‘with a cross-section of companies in terms of sector, location and size, in order to get a reasonably balanced view’ of the UK economy as a whole (Ellis and Pike 2005, 424). There is an amplification effect that ramifies across this communicative field. Each of the nine thousand contacts, the moving parts of the network, are continually in conversation with scores of their own contacts, creating an enormous epistemic apparatus of secondary and tertiary actors that extends the field of intelligencegathering far beyond the shores of the UK, yielding a system for gleaning information with a global reach. (Holmes 2015, 23)

These reports are summarized and presented to the Monetary Policy Committee of the Bank just prior to its deliberations on interest rates. Senior policymakers from the Bank—including the Governor and deputy governors—regularly accompany the agents on these forays into the field. These senior officials communicate central bank policy during these visits, but they also actively solicit stories—anecdotal data—from the employees, managers, and owners of these enterprises. They talk numbers; they talk trends; and they talk outlooks. In these interchanges, they glean contemporaneous reports on the UK economy, and they also garner from their interlocutors the details and contradictions typically lost or suppressed by economic statistics. The discursive exchanges enable them to put words both to the ephemera of local expectations and sentiments and to the rapidly changing competitive pressures unfolding in global markets, particularly among the UK’s trading partners. This network of interlocutors provides technical representations of the British economy, imparting (or restoring) social mediation and social context to economic analysis (Holmes 2014a, 108; see also Riles 2000). The diverse groups of contacts that make up the network perform descriptive, 188

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

A Tractable Future

explanatory, and interpretive labour, refining the discursive nature of economic phenomena in real time. In these face-to-face conversations, officials draw on the creative insights of individuals, who are making, remaking, and unmaking the economic drama prospectively under conditions in which ‘ceteris paribus’ does not obtain (Rudnyckyj 2014). These conversations—the epistemic basis of monetary affairs—can align policy with the shifting challenges of the present situation and the means and methods by which they are addressed by firms, households, and individuals. The actors engaging the Bank in conversation model the economy and the financial system on their terms and for their purposes. Their ideas—their configurations of belief—thereby play a decisive role in the economic and monetary drama by which investment, employment, and consumption plans by firms, households, and individuals become the basis of action or inaction. The forward-looking appraisals of these contacts—articulated in a language that may or may not be congruent with conventional economic theory—are capable on their own of orchestrating the transformations by which plans become deeds. But by accessing and internalizing this wide range of appraisals in a structured fashion, the Bank’s own policy decisions and pronouncements are informed by the sort of dissonance of diverse interpretive frames that is necessary for the disruption of groupthink (Bronk and Jacoby 2016, 19–20; Stark 2009). This increases the Bank’s chances of spotting newly emerging trends when forecasts of future outcomes and related risk and rewards would—thanks to uncertainty—otherwise be largely or entirely incalculable.

Conclusion This chapter provides an overview of a series of narrative experiments operating across numerous written and spoken genres that enliven the work of central banks. It seeks to demonstrate the creative possibilities of these narratives, their capacity to articulate new metaphors, new relationships, and new ways of knowing, thinking, and understanding. These discursive practices— the ethnographic basis of a narrative economics—can help policymakers and public alike address fundamental questions that are unanswerable (or unthinkable even) from most conventional economic perspectives or purviews (Bronk 2009, 273–87). The chapter also examined several cases—the Riksbank, the Bank of England, and the European Central Bank—in which the transformations of the ‘governability paradigm’ that Benjamin Braun depicts were fully registered in the re-articulation of central bank policy. To accomplish the shift in macroeconomic governance that Braun identifies, central bankers were obliged to address a fundamental conceptual challenge regarding the identity 189

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

of participants in these communicative experiments. The central bankers involved have been compelled to develop means and methods to speak to various segments and strata of the public—no longer merely to market participants—broaching a decisive transformation in the relational basis (the ‘primordial responsibilities’) that animate an emerging social and constitutional order. More importantly, central banks have seen their discursive role shift from merely talking at the public in order to corral their expectations (as in forward guidance) to conversing with the public. Central banks rely on such conversations and the creative use of language as a means of exploring the expectational landscape they face, while generating new theoretical and policy insights. At the heart of the experiments is a fundamental premise—an empirical premise—namely, that markets are a function of language. Central bankers create and enter, as it were, a communicative field in which countless protagonists model economic phenomenon for their own purposes, employing their own pragmatic insights and grounded truths. They (and we) are confronted with actors whose futures are enlivened by just about every emotional sensibility, every constellation of thought and belief, reason and unreason, rationality and irrationality, as well as every human proclivity to create truth, untruth, virtue, beauty, and depravity (Bronk 2009; Rudnyckyj 2014). The stories told by these unruly figures can impel or impede the leaps of faith that ratify or foreclose a tractable future (Beckert 2016, 263). The efficacy of monetary policy thus rests on the representational enterprise of these protagonists with whom central bankers must orchestrate prospectively the contingencies of economic stability and growth.

Acknowledgements The research on which this text was based was made possible by my participation in two remarkable scholarly groups: ‘Meridian 180: Transforming the Transpacific Dialogue’, Cornell University Law School, directed by Annelise Riles; and ‘Global Financial Initiative of the Mario Einaudi Center for International Studies’, Cornell University, directed by Hirokazu Miyazaki. It has also benefited from my current participation in the ‘Global Foresight Project’, funded by the Swedish Foundation for Humanities and Social Sciences and directed by Christina Garsten.

Bibliography Abolafia, Mitchel Y. 2010. ‘Narrative Construction as Sensemaking: How a Central Bank Thinks’. Organization Studies 31 (3): pp. 349–67. Akerlof, George A., and Robert J. Shiller. 2009. Animal Spirits: How Human Psychology Drives the Economy and Why It Matters. Princeton, NJ: Princeton University Press.

190

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

A Tractable Future Bank of England. 2013. ‘Monetary Policy Trade-offs and Forward Guidance’. August. http://www.bankofengland.co.uk/publications/Documents/inflationreport/2013/ ir13augforwardguidance.pdf. Bank of England. 2014. ‘News Release—Bank of England Launches Strategic Plan’. 18 March. http://www.bankofengland.co.uk/publications/Pages/news/2014/058.aspx. Beckert, Jens. 2016. Imagined Futures: Fictional Expectations and Capitalist Dynamics. Cambridge, MA: Harvard University Press. Bernanke, Ben. 2015. ‘Inaugurating a New Blog’. 30 March. https://www.brookings. edu/blog/ben-bernanke/2015/03/30/inaugurating-a-new-blog/. Bernanke, Ben, Thomas Luabach, Fredric S. Mishkin, and Adam S. Posen. 1999. Eds. Inflation Targeting: Lessons from the International Experience. Princeton, NJ: Princeton University Press. Blinder, Alan S. 2013. After the Music Stopped: The Financial Crisis, the Responses, and the Work Ahead. New York: Penguin. Blinder, Alan S., and Ricardo Reis. 2005. ‘The Greenspan Standard’. Paper presented at the Federal Reserve Bank of Kansas City Symposium, Jackson Hole, WY, 25–27 August. Borio, Claudio. 2017. ‘Secular Stagnation or Financial Cycle Drag?’. Keynote speech, National Association for Business Economics, 33rd Economic Policy Conference, 5–7 March 2017, Washington, D.C. Bronk, Richard. 2009. The Romantic Economist: Imagination in Economics. Cambridge: Cambridge University Press. Bronk, Richard, and Wade Jacoby. 2016. ‘Uncertainty and the Dangers of Monocultures in Regulation, Analysis, and Practice’. MPIfG Discussion Paper 16/6, Max Planck Institute for the Study of Societies, Cologne. Callon, Michel. 2007. ‘What Does it Mean to Say that Economics is Performative?’. In Do Economists Make Markets? On the Performativity of Economics, edited by Donald MacKenzie, Fabian Muniesa, and Lucia Siu, 311–57. Princeton, NJ: Princeton University Press. Carney, Mark. 2014. ‘One Mission. One Bank. Promoting the Good of the People of the United Kingdom’. Mais Lecture at Cass Business School, City University, London, 18 March. Curé, Benoît. 2013. ‘Revisiting the European Social Contract’. Speech presented at the European Conference at Harvard Europe 2.0—Taking the Next Step, Cambridge, MA, 2 March. Dewey, John. 1927 [1991]. The Public and Its Problems. Athens, OH: Swallow Press. Draghi, Mario. 2012a. ‘Introductory Statement by Mario Draghi, President of the ECB’, Hearing at the Committee on Economic and Monetary Affairs of the European Parliament, Brussels, 9 July. Draghi, Mario. 2012b. ‘Verbatim of the Remarks Made by Mario Draghi’. Speech by Mario Draghi, President of the European Central Bank at the Global Investment Conference in London, 26 July. Ellis, Colin, and Tim Pike. 2005. ‘Introducing the Agents’ Scores.’ Bank of England Quarterly Bulletin Winter 2005: pp. 424–30. Friedman, Benjamin M. 2002. ‘The Use and Meaning of Words in Central Banking: Inflation Targeting, Credibility, and Transparency’. NBER Working Paper, no 8972.

191

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures Fischer, Stanley. 2016. ‘Remarks by Stanley Fischer, Vice Chairman, Board of Governors of the Federal Reserve System’. Presented at ‘A Conference in Honor of Michael Woodford’s Contributions to Economics’ cosponsored by the Federal Reserve Bank of New York, Columbia University Program for Economic Research, and Columbia University Department of Economics, New York, 19 May. Gabor, Daniela, and Bob Jessop. 2015. ‘Mark My Words: Discursive Central Banking in Crisis.’ In Financial Cultures and Crisis Dynamics, edited by Bob Jessop, Brigitte Young, and Christoph Scherrer, 294–315. London: Routledge. Goodhart, Charles. 1975. ‘Monetary Relationships: A View from Threadneedle Street’ and ‘Problems of Monetary Management: The UK Experience’. In Papers in Monetary Economics, Vol, I, Reserve Bank of Australia. Haldane, Andrew G. 2015. ‘Who Owns a Company?’. Speech delivered at the University of Edinburgh Corporate Finance Conference, Edinburgh, 22 May. Holmes, Douglas R. 2014a. Economy of Words: Communicative Imperatives in Central Banks. Chicago, IL: University of Chicago Press. Holmes, Douglas R. 2014b. ‘Communicative Imperatives in Central Banks’. Cornell International Law Journal 47 (1): pp. 15–61. Holmes, Douglas R. 2015. ‘Public Currency: Anthropological Labor in Central Banks’. Journal of Cultural Economy 9 (1): pp. 5–26. Holmes, Douglas R. 2016. ‘Central Bank Capitalism: Visible Hands, Audible Voices’. Anthropology Today 32 (6): pp. 3–7. Knorr Cetina, Karin. 1999. Epistemic Cultures: How the Sciences Make Knowledge. Cambridge, MA: Harvard University Press. Lippmann, Walter. 1927 [2002]. The Phantom Public. New Brunswick, NJ: Transaction. MacKenzie, Donald. 2006. An Engine, Not a Camera: How Financial Models Shape Markets. Cambridge, MA: MIT Press. MacKenzie, Donald, Fabian Muniesa, and Lucia Siu. 2007. ‘Introduction’. In Do Economists Make Markets? On the Performativity of Economics, edited by Donald MacKenzie, Fabian Muniesa, and Lucia Siu, pp. 1–19. Princeton, NJ: Princeton University Press. McCloskey, Deirdre. 1985. The Rhetoric of Economics. Madison, WI: University of Wisconsin Press. Merton, Robert K. 1948. ‘The Self-Fulfilling Prophecy’. Antioch Review 8 (2): pp. 193–210. Miyazaki, Hirokazu. 2013. Arbitraging Japan: Dreams of Capitalism at the End of Finance. Berkeley, CA: University of California Press. Muniesa, Fabian, and Michel Callon. 2007. ‘Economic Experiments and the Construction of Markets’. In Do Economists Make Markets? On the Performativity of Economics, edited by Donald MacKenzie, Fabian Muniesa, and Lucia Siu, pp. 163–89. Princeton, NJ: Princeton University Press. Muth, John. 1961. ‘Rational Expectations and the Theory of Price Movements’. Economentrica 29 (3): pp. 315–35. Nelson, Nicole, Anna Geltzer, and Stephen Hilgartner. 2008. ‘Introduction: The Anticipatory State: Making Policy-Relevant Knowledge about the Future’. Science and Public Policy 35 (8): 546–50.

192

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

A Tractable Future Nelson, Stephen, and Peter J. Katzenstein. 2010. ‘Uncertainty and Risk in the Crisis of 2008’. Paper presented at the Politics in Hard Times Workshop Honoring Peter Gourevitich, University of California at San Diego, 23–24 April. Riles, Annelise. 2000. The Network Inside Out. Ann Arbor, MI: University of Michigan Press. Riles, Annelise. 2011. Collateral Knowledge: Legal Reasoning in the Global Financial Markets. Chicago, IL: University of Chicago Press. Rudnyckyj, Daromir. 2014. ‘Economy in Practice: Finance and the Problem of Market Reason’. American Ethnologist 41 (1): pp. 110–27. Searle, John. 1969. Speech Acts: An Essay in the Philosophy of Language. Cambridge: Cambridge University Press. Smart, Graham. 1999. ‘Storytelling in a Central Bank: The Role of Narrative in the Creation and Use of Specialized Economic Knowledge’. Journal of Business and Technical Communication 13 (3): pp. 249–73. Stark, David. 2009. The Sense of Dissonance: Accounts of Worth in Economic Life. Princeton, NJ: Princeton University Press. Svensson, Lars E.O. 2009. ‘Monetary Policy with a Zero Interest Rate’. Speech presented at the S.N.S. Center for Business and Policy Studies, Stockholm, Sweden, 17 February. http://www.riksbank.se/Pagefolders/39304/090217e.pdf. Sveriges Riksbank. 2015. ‘Minutes of Monetary Policy Meeting 11 February’. http:// www.riksbank.se/Documents/Protokoll/Penningpolitiskt/2015/pro_penningpolitiskt_ 150225_eng.pdf. Woodford, Michael. 2012. ‘Methods of Policy Accommodation at the Interest-Rate Lower Bound’. Paper presented at the Symposium on ‘The Changing Policy Landscape’, Federal Reserve Bank of Kansas City Symposium, Jackson Hole, WY, 31 August–1 September. Yellen, Janet L. 2013. ‘Communication in Monetary Policy’. Speech presented at the Society of American Business Editors and Writers 50th Anniversary Conference, Washington D.C., 4 April.

193

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

9 Central Bank Planning Unconventional Monetary Policy and the Price of Bending the Yield Curve Benjamin Braun

[I]t is absurd to think of a purely ‘objective’ prediction. Anybody who makes a prediction has in fact a ‘programme’ for whose victory he is working, and his prediction is precisely an element contributing to that victory. (Gramsci 1971, 171)

Introduction This volume examines a contradiction at the heart of economic action. While economic action is necessarily future-oriented, the future is characterized by Knightean uncertainty—that is, it cannot be known. In order to avoid paralysis, economic agents imagine futures that enable them to take decisions in the present ‘despite the incalculability of outcomes’ (Beckert 2016, 9). From a macroeconomic perspective, this tension is compounded by the problem of order and stability, which requires that economic actors’ beliefs and actions be coordinated despite uncertainty. In theory, there are two ‘pure’ solutions to this coordination problem. It can be solved: (a) in a centralized fashion by a social planner dedicated to eliminating uncertainty; or (b) by Hayekian speculators whose decentralized beliefs and actions are coordinated via market pricing (Konings 2016, 278). In practice, pragmatic solutions prevail: central institutions ‘reduce’ uncertainty while leaving the price mechanism largely intact. This chapter focuses on a particularly consequential coordinationfacilitating institution, central banking. As a carefully calibrated configuration of technocrats, laws, and technologies, this apparatus is geared towards

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Central Bank Planning

producing a specific ‘genre’ of imagined futures—expectations—and towards getting economic agents to internalize, and coordinate their actions around, those expectations.1 Although the distributional consequences of central banking are beyond the scope of this chapter (see Fontan et al. 2016), they motivate and justify the exploration of the linkages between economic theories, governance practices, and imagined futures (or expectations) in this chapter. Central bank expectation management is consequential because it affects macroeconomic outcomes, often in unintended ways. On one hand, modern monetary governance is built on the premise that central bank guidance on expectations— notwithstanding the occasional blip—enhances economic efficiency and stability. On the other hand, in a world of Knightean uncertainty the issuers of such guidance are prone to punching above their (epistemic) weight. Indeed, it is difficult to distinguish, ex ante, between wise coordination of expectations and the overconfident orchestration of groupthink and herding. There is a long line—from Milton Friedman to John Taylor—of conservative critics of discretionary monetary fine-tuning who have argued that ‘active policy is as likely to amplify as offset the effects of shocks upon the macroeconomy’ (Haldane 1995, 6). If that argument has merit—and recent history suggests it does—what Hayek (1989) dubs the ‘pretence of knowledge’ is not an occasional but rather a widespread feature of ‘governing through expectations’ in conditions of uncertainty (Braun 2015). After a decade of aggressive interventionism by the world’s leading central banks, calls for caution have recently grown louder. One increasingly vocal critic, the Bank for International Settlements (BIS), has urged central banks to show a ‘keener appreciation of the cumulative impact of policies on the stocks of debt, on the allocation of resources and on the room for policy manoeuvre’ (BIS 2016, 8). This chapter contributes to the growing literature on how and why macroeconomic governance oscillates between intended countercyclical stabilization and unintended pro-cyclical amplification of the boom and bust dynamic of the business cycle (Baker and Widmaier 2014; Braun 2014; Bronk and Jacoby 2016; Golub et al. 2015; Widmaier 2016). The chapter advances three main arguments: (i) ‘Keynesian’ fiscal demand management and monetary inflation targeting can be conceptualized as separate ideal-types of macroeconomic state agency: the former operates in a hydraulic manner, while the latter is strategic and performative. Crucially, however, recent quantitative easing programmes mark a return to the hydraulic style of

1 This chapter has benefitted from the comments of Jens Beckert, Richard Bronk, Peter Dietsch, Maximilian Düsterhöft, Dirk Ehnts, Onur Ozgode, Waltraud Schelkle, and Matthew Watson. Any errors are mine.

195

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

macroeconomic governance, albeit via the monetary rather than the fiscal authority, and via the market for financial assets rather than the markets for goods and services. (ii) The chapter proposes a reading of the past three decades of central bank history as a quest to expand the temporal reach of monetary policy into the future. A structural break occurred when central banks shifted from open market operations at the short end of the yield curve to purchases of longer-dated securities.2 In the guise of balance-sheet policy, central banking has—this chapter argues—morphed into a form of central (bank) planning. (iii) While this transition has much to do with the practical challenges of monetary policy implementation and transmission, central bank planning also has an important theoretical lineage in macroeconomics. Building on the literature on the performative effects of economic models and central bank communication in ‘making’ the economy (Christophers 2017; Holmes 2014), but moving towards an understanding more akin to Mitchell (2005) and MacKenzie (2006), this chapter reveals a deeper connection between general-equilibrium theory in macroeconomics and central bank planning.3

Governability Paradigms: Fiscal Demand Management versus Monetary Inflation Targeting Why is it that the economy is governable by means of monetary policy? Contrary to the widespread notion of all-powerful central banks, the effectiveness of monetary policy has always been fragile, both in theory and in practice. While the New Classical view that the economy was essentially ungovernable and that government intervention would, at best, be ineffective was suppressed by the new neoclassical synthesis (Braun 2014, 61), doubts regarding the effectiveness of monetary policy never went away. It has, for instance, long been the ‘prevailing view’ in central banking circles that ‘over the medium term, monetary policy only passively tracks the natural [interest] rate’ (Juselius et al. 2016, 1). More recently, evidence for the ‘globalization of inflation’ has called into question the very notion—foundational for contemporary monetary policy—of inflation as a domestic phenomenon (Auer et al. 2017).

2 The ‘yield curve’ is a visual representation of the relationship between the residual maturity of a debt security and its ‘yield’ or interest rate (see Figure 9.1). 3 For an argument about how the work of MacKenzie (and Michel Callon) can improve the microfoundations of political economy more generally, see Braun (2016a).

196

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Central Bank Planning

In light of these observations, this chapter conceptualizes macroeconomic governability as the unstable effect of historically contingent configurations of institutions, theories, and governance practices (Braun 2014). A governability paradigm is established when a political consensus on macroeconomic policies (Hall 1993) is reinforced by an academic consensus on macroeconomic theories. In the brief history of modern macroeconomics, these two things have come together only twice: in the form of ‘Keynesian’ fiscal demand management, underpinned by the neoclassical synthesis, and in the form of monetary inflation targeting underpinned by the ‘new neoclassical synthesis’ (Braun 2014; Widmaier 2016). These governability paradigms were embedded in two distinct regimes. During the Bretton Woods era of restricted international capital flows and highly regulated financial systems, macroeconomic governance operated through price and interest rate controls and direct fiscal spending. In the context of financial deregulation and financialization, the headquarters of macroeconomic governance shifted from the fiscal authority to the monetary authority. Paving the conceptual way for the remainder of the chapter, this section compares the two paradigms in terms of the ‘four T’s’ of macroeconomic state agency: transmission, temporality, transparency, and theory (see Table 9.1).

Transmission Fiscal demand management influences the economy by increasing or decreasing demand directly through centralized government expenditure. Although ‘animal spirits’ and a ‘multiplier effect’ are important aspects, fiscal demand management operates through interventions that have sizeable first-round effects in the markets for labour, goods, and services (the ‘real economy’). By contrast, inflation-targeting central banks typically rely on open market operations that have only small first-round effects in a small sector of the financial economy, the interbank money market. A gap therefore exists between the operational target of monetary policy—the short-term interbank interest rate—and the long-term interest rates that matter most for the actual targets Table 9.1. The ‘four T’s’ of macroeconomic state agency—two governability paradigms compared ‘Keynesian’ fiscal demand management Transmission

Centralized and direct

Temporality

Present-oriented

Transparency

Opacity

Theory

Hydraulic/optimal control: game against nature

Transition-period monetary policy

Monetary inflation targeting

Decentralized and indirect Future-oriented Secrecy

Transparency Communicative/performative: game with rational agents

197

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

of monetary policy, namely, investment, output, employment, and, ultimately, inflation. To bridge this gap, central bankers rely on the ‘transmission mechanism of monetary policy’. Generally thought of as comprising a bank lending channel, a bank capital channel, a balance-sheet channel (or ‘financial accelerator’), and, more recently, a separate risk-taking channel (Borio and Zhu 2012), the transmission mechanism is indirect, decentralized, and prone to disruption. It also relies heavily on expectations.

Temporality In principle, a Keynesian fiscal stimulus works best in a world in which consumers’ spending decisions are unencumbered by expectations of higher future taxes. By contrast, the future intentionally looms large under inflation targeting, for both theoretical and practical reasons. Theoretically, rational expectations macroeconomics pushed the focal point for macroeconomic coordination into the future by arguing that short-term stabilization measures would be neutralized by rational actors anticipating and discounting the longterm consequences (Lucas and Sargent 1979). The practical reason has to do with the expectational dimension of monetary policy’s transmission mechanism. The impact of a change in the short-term interest rate today depends on how it affects private-sector expectations of inflation and interest rates tomorrow. If market expectations fail to adjust, the interest-rate change will be ineffectual. As the leading monetary theorist put it in the heyday of inflation targeting, ‘[n]ot only do expectations about policy matter, but, at least under current conditions, very little else matters’ (Woodford 2003, 15). This has direct implications for the third ‘T’.

Transparency Anthony Giddens once suggested that Keynesian demand management could perhaps ‘only be effective in circumstances in which the majority of the population, or certain key sets of business actors, do not know what Keynesianism is’ (Giddens 1987, 201). He shared this view with the New Classical economists, who saw non-transparency as a necessary feature of countercyclical policies. According to them, the effectiveness of such policies ‘rests on the inability of private agents to recognize systematic patterns in monetary and fiscal policy’ (Lucas and Sargent 1979, 58)—that is, on their inability to understand the (New Classical) concepts of Ricardian equivalence or the neutrality of money. When monetary policy assumed a larger share of the burden of macroeconomic stabilization policy during the late 1970s and the 1980s—a transitional period between two governability paradigms—it continued to operate on the principles of obscurity and secrecy (Goodfriend 198

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Central Bank Planning

1986). Little changed until the early 1990s (Haldane 2017, 5). Only when inflation targeting gained traction did central banks shift from secrecy to ‘strategic transparency’ (Abolafia and Hatmaker 2013, 541–3; Krippner 2007). The more predictable the central bank, the argument went, the more predictable—and hence governable—the behaviour of market actors. This comparison of the two governability paradigms in terms of transmission, temporality, and transparency, sets the scene for a discussion of the fourth ‘T’—theory—and thus of the deep performativity of central banking.4

Deep Performativity: Uncertainty, Rational Expectations, and Central Bank Planning At the heart of rational expectations macroeconomics lies a paradox. While real-business-cycle and Dynamic Stochastic General Equilibrium (DSGE) models are populated by competing and unflinchingly rational maximizers, the theoretical innovation that brought them about carried the seeds of central (bank) planning. John Muth and Robert Lucas introduced the rational expectations hypothesis (REH) as a modelling assumption, defining rational expectations as ‘essentially the same as the predictions of the relevant economic theory’ (Muth 1961, 316). While not making a normative statement about ‘what firms ought to do’ (Muth 1961, 316), they did make a normative statement about what economists ought to do; namely, assume that the prediction of their model was the point around which firms’ expectations were (normally) distributed. That makes the world of rational expectations a world of Knightean ‘risk’ rather than uncertainty. In real-business-cycle and DSGE models, all agents—or, to be precise, the ‘representative agent’—use the ‘true’ model (that is, the modelling economist’s own model) to predict correctly (the probabilities of) all possible future prices and outcomes (Frydman and Goldberg 2011, 62). The REH has two implications that bear on the theory of central planning.5 First, while different economists may build different models, within each model there is what Thomas Sargent (Evans and Honkapohja 2005, 566) called ‘a communism of models’, whereby ‘[a]ll agents inside the model, the econometrician, and God share the same model’. Second, the assumed omniscience of rational expectations macroeconomics runs counter to the Hayekian case for the market mechanism (Bronk 2013; Hayek 1945). In a world in which 4 For a study of deep performativity in the area of fiscal policy, see Heimberger and Kapeller (2017). 5 A close affinity exists between general equilibrium theory and socialist planning that precedes the advent of REH-based macroeconomics; see Mirowski (2002) and Boldyrev and Kirtchik (2017).

199

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

all agents share the same (correct) economic model, centralized knowledge is just as good as decentralized knowledge. As a result, ‘there would be no need for markets to set prices . . . All economists and each rational market participant would be capable of accomplishing this feat entirely on their own’ (Frydman and Goldberg 2011, 66). This section discusses how these theoretical implications of the REH relate to the practice of central banking.

‘Communism of Models’ and Epistemic Authority In the real world, reducing uncertainty about the ‘correct’ model of the economy is a key element of coordinating private-sector expectations (Nelson and Katzenstein 2014). Central banks have used REH-based models as narrative and coordinative devices for precisely this purpose (Beckert 2016; Holmes 2014). The central banks that pioneered inflation-forecast targeting were the first to incorporate DSGE models—the model-version of the new neoclassical synthesis—into their apparatuses of expectation management. When the central banks of Canada and New Zealand introduced the precursors of today’s DSGE models, these ‘were not mere research projects, but models used for practical policy deliberations under the “forecast targeting” approach to monetary policy’ (Woodford 2009, 276; see also Holmes 2014, 92–7). By the time the financial crisis hit, most leading central banks relied heavily on DSGE models, which largely abstracted from the financial sector, creating a dangerous ‘modelling monoculture’ (Bronk 2011; Bronk and Jacoby 2016). Sargent’s ‘communism of models’ had become a reality. Maintaining this model communism in academic and financial circles requires a type of credibility that differs from the conventional concept of credibility that, in reaction to the REH-inspired time-inconsistency critique, became the holy grail of central banking (Barro and Gordon 1983; Kydland and Prescott 1977). According to this literature, a central bank can acquire credibility on the basis of a historical record of high inflation aversion, because it is ‘bound by a rule or other “commitment technology” ’, or because senior central bankers are employed on an ‘incentive-compatible contract’ (Blinder 2000, 1423). In Blinder’s own definition (2000, 1423), ‘a central bank is credible if people believe it will do what it says’. In practice, this is the credibility of the central bank’s commitment to raise the policy rate when the (expected) inflation rate goes up. In the presence of ‘epistemic uncertainty’ (Nelson and Katzenstein 2014, 363), however, central banks seek not only commitment credibility but also forecasting credibility—or, to use an established term, epistemic authority (Rosenhek 2013). Paraphrasing Blinder’s definition, a central bank enjoys epistemic authority if people believe its forecasts. Market actors may decide to form their own expectations on the basis of a central bank forecast because they believe that it provides the best prediction of 200

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Central Bank Planning

economic fundamentals, and/or that it will serve as a focal point for the coordination of private-sector expectations (Morris and Shin 2002, 1522). In either case, epistemic authority is fragile. In order to bolster it, central banks have invested heavily in the scientization of monetary policy. Increasing the ranks of PhD-level economists both at the staff level and in leadership positions, central banks have effectively transformed themselves into research hubs of unprecedented size and scope (Conti-Brown 2016, 90–3; Marcussen 2009; Mudge and Vauchez 2016). Leadership positions are routinely assigned to academic economists, including Stanley Fischer, Mervyn King, Ben Bernanke, Raghuram Rajan, and Janet Yellen. Regardless of the scientific rigour it may or may not bring to the policy process, scientization has a performative dimension. Models and academic merits are props in a carefully staged performance of competence and knowledge that bolsters the uncertainty-reducing effect of central bank forecasts (Goffman 1959). It is impossible to determine, in real time, where knowledge ends and where what Hayek called ‘pretence of knowledge’ begins (Hayek 1989). Following the financial crisis, some macroeconomists decried a pervasive ‘pretense-of-knowledge syndrome’ in their discipline (Caballero 2010). However, under conditions of uncertainty, the ‘pretention that the fictional depictions [are] indeed true representations of the future’ (Beckert 2013b, 226) is part and parcel of economic decision-making. From this perspective, ‘pretence of knowledge’ is not pathological, as Hayek and Caballero argue, but performs a productive—albeit problematic—function in the communicative apparatus of monetary expectation management (Braun 2015). By targeting long-term interest rates through the twin policies of forward guidance and quantitative easing, central banks have recently extended their reach into much more distant futures. As this chapter will argue in relation to the European Central Bank (ECB), this has made it harder to maintain the pretence of superior central bank knowledge. The attempt to bolster the credibility of its commitment to stabilize inflation at a low but positive rate has undermined the ECB’s claim to epistemic authority.

Non-Market Price Setting: From Hydraulic to Performative Governability, and Back Again The quasi-mechanical connection between government spending and aggregate demand—the economic agency of the Keynesian state—can be conceptualized as hydraulic (Braun 2014, 59; Pahl and Sparsam 2016). Central bank agency under inflation targeting, by contrast, has been described as communicative and performative: central bankers’ utterances ‘are making the economy . . . as a communicative field and as an empirical fact’ (Holmes 2014, 12; see also Krippner 2007). However, these utterances refer to an ‘economy’ that has a prior existence as a theoretical fact. The performative nature of inflation targeting 201

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

comes into full view only if the performative dimension of REH theory-based macroeconomics—namely, its unexpectedly ambiguous relationship with ‘the economy’ and its inbuilt performativity—is explained. Lucas and others argued that Keynesian macroeconomics assumed, wrongly, that modelling the economy and governing the economy were two different things, and that the economy was a mechanical system of aggregates that followed a set of quasi-physical laws (Lucas and Sargent 1979, 12). Interestingly, this ‘optimal control’ mind-set still underpinned monetary policy during the transitional Volcker and (early) Greenspan era: ‘A set of equations described the behavior of the private sector; the job of the central bank was to select the proper settings for its policy instruments to guide the economy along its optimal path’ (Poole and Rasche 2000, 257). Over the long term, however, rational expectations proved to be a game changer. As Kydland and Prescott (1977, 473) put it in their influential article on policy ineffectiveness, ‘economic planning is not a game against nature but, rather, a game against rational economic agents’. New Classical theorists, who dismissed Keynesians for modelling macroeconomic aggregates, aimed at putting macroeconomics on ‘microfoundations’— ‘representative’ households and firms that rationally maximize objective functions and adapt instantly to policy changes. This made the new ‘game against rational economic agents’ trickier than the old ‘game against nature’. The predictability of a mechanical system had been replaced by the strategic calculations of homo economicus, which tended to neutralize or counteract countercyclical policy interventions: ‘The private sector could in principle not be modelled without specifying the monetary policy rule, because the behaviour of optimizing agents could not be predicted without modelling their expectations about monetary policy’ (Poole and Rasche 2000, 257). Thus, by substituting reflexivity for optimal control, Lucas and colleagues unwittingly brought performativity to monetary policy—long before central bankers became expectation managers. Somewhat ironically, rational expectations theorists only saw the downside for governability. It was for monetary policy-makers to discover that the ‘game against rational agents’ could be turned into a ‘game with rational agents’. When central bankers realized that control over the economy depended on their ability to harness market actors’ expectations, they transformed macroeconomic governance from an ‘engineering’ problem into a ‘strategic’ one (Morris and Shin 2008, 88). The notion of a hydraulic system that could be manipulated by skilled engineers was replaced by the performative challenge of making market expectations conform to the rationality standards as defined by those who hoped to govern through these expectations. Here, too, central banks’ responses to the financial crisis marked a turning point. Historically, central banks fixed the price for short-term liquidity in the interbank market. With the shift from conventional interest-rate policy to 202

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Central Bank Planning

unconventional balance-sheet policy, however, central banks expanded their reach to security prices at the long end of the yield curve. Expanding the government’s (consolidated) balance sheet to purchase securities—rather than goods and services, as fiscal policy would—central banks effectively returned to a hydraulic mode of economic governance. Focusing on the euro area, the remainder of this chapter will further examine these reversals: the revival of hydraulic macroeconomic policy and the undermining of epistemic authority.

Forward Guidance and QE in the Euro Area At various points after 2008, the US Federal Reserve, the ECB, the Bank of England, and the Bank of Japan all adopted two types of ‘unconventional’ monetary policies to stabilize financial conditions and stimulate economic activity: forward guidance and large-scale asset purchases, or quantitative easing (QE). These policies are complementary in that both aim explicitly at lowering long-term interest rates when the short-term rate is already at the effective lower bound (Cœuré 2015, 2). However, whereas forward guidance— an advanced version of expectation management—is consistent with conventional ‘interest-rate policy’, asset purchases fall into the economically distinct category of ‘balance-sheet policy’ (Borio and Disyatat 2009, 1).

Forward Guidance: From Transparency to Commitment In light of the long-standing trend for central banks to become more transparent about their actions and intentions, forward guidance embodies continuity rather than change. Smaller central banks, in particular, had long published unconditional forecasts of the future path of their policy rates, thus increasing the reach of their expectation management further into the future (Filardo and Hofmann 2014, 38; Holmes 2014, 77–9). The ECB, too, had been moving towards greater forecast transparency in the early 2000s (Braun 2015, 375–7). However, these ‘Delphic’ forms of forward guidance did not involve a commitment (Campbell et al. 2012). Commitment to particular courses of action thus was the innovative element of the ‘Odyssean’ (ibid.) variant of forward guidance that became prevalent after 2008. The rationale was simple: forward guidance would allow policy-makers to ‘change public expectations of their actions tomorrow in a way that improves macroeconomic performance today’ (Campbell et al. 2012, 3). The ECB announced its own version of forward guidance in June 2013: The Governing Council expects the key ECB interest rates to remain at present or lower levels for an extended period of time. This expectation is based on the

203

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures overall subdued outlook for inflation extending into the medium term, given the broad-based weakness in the real economy and subdued monetary dynamics. (Draghi 2013; emphasis added)

Instead of the quantitative, threshold-based forward guidance ultimately embraced by the Fed (FOMC 2013), the ECB thus chose a cautious, openended variant. While following its peers in making its forward guidance conditional—most notably on the medium-term inflation outlook—the ECB did not commit to maintaining current interest-rate levels until a specific date or macroeconomic outcome was reached. In its Monthly Bulletin, the ECB described its own approach as ‘[q]ualitative forward guidance conditional on a narrative’ (ECB 2014, 68). Notwithstanding the innovative commitment aspects, such forward guidance is—as a purely communicative tool designed to manage expectations— fully consistent with the pre-crisis paradigm of ‘discursive central banking’ (Gabor and Jessop 2015). The same is not true of QE, which brings something other than words to the governability table, namely, the central bank balance sheet.

Quantitative Easing In order to keep the short-term interbank interest rate aligned with its main refinancing rate, the Eurosystem—the ECB and the national central banks— used to provide precisely that amount of reserves that would satisfy the banking system’s liquidity needs, thus keeping supply and demand for reserves in the interbank market in balance. While the Eurosystem conducts collateralized lending operations for this purpose, other central banks, such as the Fed, provide liquidity through open market operations (that is, outright purchases of securities). In purely technical terms, large-scale asset purchases are but an expanded version of the latter, in that the central bank creates new reserves to buy securities in the open market (Cœuré 2015). In economic terms, however, the two are very different. Conventional open market operations are designed to affect the short-term interest rate via the liability side of the central bank’s balance sheet—namely, the amount of reserves provided to the banking system (Friedman 2014, 7). QE open market operations, by contrast, are designed to affect long-term interest rates via the asset side of the central bank’s balance sheet—namely, the amount of securities absorbed from the financial system. Following the example of its peers, the ECB decided to expand its balance sheet by purchasing both government bonds and securities issued by the private sector. In late 2014, a third covered bond purchase programme (CBPP3) and an asset-backed securities purchase programme (ABSPP) marked 204

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Central Bank Planning

the prelude to the public-sector purchase programme (PSPP), which the ECB launched in March 2015. In June 2016, the ECB added the corporate sector purchase programme (CSPP). Together, these various components form the ‘expanded asset purchase programme’ (APP). At year-end 2016, public sector securities accounted for eighty-two per cent of Eurosystem holdings under the APP. The weighted average remaining maturity of the ECB’s holdings of government bonds is currently 8.3 years, showing that the PSPP targets bonds with mid-range maturities. The ECB’s quantitative easing comes with its own form of forward guidance. The ECB has committed to purchasing securities worth EUR 60 billion every month ‘until the end of 2017 and in any case until the Governing Council sees a sustained adjustment in the path of inflation that is consistent with its aim of achieving inflation rates below, but close to, 2% over the medium term.’6 Indeed, it is important to distinguish two channels through which QE affects asset prices and interest rates (Valiante 2017). First, the mere announcement of asset purchases changes expectations. Based on the expectation of greater scarcity—and therefore higher prices—of bonds in the future, demand should be expected to increase already in the present, pushing prices up and yields down. However, this ex ante and performative signalling effect is entirely dependent on the expectation of the ex post effect of actual central bank purchases, which increase asset scarcity in a mechanical, hydraulic manner. In short, the ex ante performative and the ex post hydraulic effect of the quantitative easing programme are mutually reinforcing and together push bond prices up and (long-term) interest rates down. As this chapter will show, the implications for macroeconomic state agency are profound.

Long-Term Interest Rates as Policy Variables: Post-Crisis Central Bank Planning Despite a long-standing trend for the apparatus of monetary expectation management to reach ever further into the future, forward guidance and QE crossed what central bankers had previously considered a red line. The ECB, which used to implement its monetary policy stance by targeting the price for short-term interbank liquidity, was vocal about the risks of non-market price setting further up the yield curve. As recently as 2008, it explained its ‘ “handsoff” approach’, by declaring that ‘developments in longer-term money market interest rates reflect market forces’ and are therefore ‘beyond the ECB’s direct 6 https://www.ecb.europa.eu/mopo/implement/omt/html/index.en.html, accessed on 13 April 2017.

205

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

control’ (ECB 2008, 71, 79). Although they knew it was feasible, central bankers considered efforts to control long-term rates undesirable. The chief architect of the ECB’s monetary policy strategy used a classic Hayekian argument against non-market price setting, arguing that centrally determined long-term interest rates ‘would cease to have their important allocational [sic] function in a market economy by virtue of being relative indicators of scarcity’ (Issing 1992, 293). The consensus under the inflation-targeting paradigm that monetary policy should ‘lead the market’ rather than ‘follow’ it (Blinder 2004, 66–74) was thus limited to the short end of the yield curve. The view was widely shared among monetary policy-makers that minimizing the potentially distorting influence of such leadership required that longer-term interest rates be determined by market forces (Turner 2011, 19).7 After 2008, central banks quietly shelved that view. It became the stated goal of ECB policy to encourage ‘portfolio shifts into longer maturity assets and a compression of long-term yields’ (ECB 2014, 67). Forward guidance and QE thus marked a significant departure from the inflation-targeting paradigm of the pre-crisis period. The boundary between ‘following’ and ‘leading the market’ shifted towards the long end of the yield curve, and central banks now ‘made the long-term interest rate a policy variable’ (Turner 2011, 10). Longterm rates, previously regarded as a barometer for market actors’ expectations of the future, became a lever for central banks to influence that future.

Forward Guidance and the ECB’s Epistemic Authority Critics of forward guidance warn that central banks cannot possibly satisfy market actors’ ‘insatiable’ hunger for information (Issing 2014, 10), and that they would need to escalate their commitments to reassure markets that interest rates would remain unchanged even if economic recovery continued. Indeed, modest signs of recovery in the euro area in early 2014 prompted the ECB to specify that the promise to keep its key interest rates ‘at present or lower levels’ was ‘based on an overall subdued outlook for inflation extending into the medium term, given the broad-based weakness of the economy, the high degree of unutilized capacity, and subdued money and credit creation’ (ECB 2014, 69). This somewhat convoluted statement was intended to reassure market participants that even stronger than anticipated growth would not precipitate a rate rise as long as inflation expectations remained anchored, the output gap positive, and bank lending slow. Central banks felt compelled to escalate their forward guidance for fear that positive economic

7

For an important critique of the ‘barometer’ conception of long-term rates, see Shin (2017).

206

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Central Bank Planning

data would cause market actors to adjust their expectations of the future path of the short-term interest rate upwards. Again, the problem is uncertainty. The goal of forward guidance is to reduce market uncertainty about the future path of interest rates and to ‘talk down’ longer-term rates, thus bending the long end of the yield curve downwards. But longer time horizons imply greater forecast uncertainty. Does the ECB have the epistemic authority needed to steer market expectations of such distant futures? Interestingly, the ECB’s chief economist addressed precisely this concern at the 2014 session of the annual ‘The ECB and its watchers’ conference: Our approach starts from the premise that the central bank doesn’t have superior knowledge about how the world works. Nor are we likely to have better forecasting abilities than the majority of observers. So what we can do is to provide an explicit, well-articulated frame of reference for our actions. . . . In practical terms, this means that communication revolves around providing a narrative about the economy . . . (Praet 2014)

This amounts to a call on market participants to abandon the pretence of superior central bank knowledge that had formed an integral part of the communicative apparatus of expectation management. What is more, Praet was essentially describing forward guidance as an effort to develop persuasive economic narratives. This points to a dilemma that goes beyond epistemic uncertainty over the central bank’s macroeconomic model: namely, the possibility that forecasts—regardless of their accuracy—are being used strategically by central bankers. The following quotes illustrate how observers of central banks have reacted to this possibility. [T]he market knows that central bankers have no superhuman forecasting ability and will tend to view the supposed longer-term forecasts as a version of jawboning, attempts to persuade the market to change its mind for immediate policy purposes. Again there is little empirical evidence that the market responds to such jawboning, and why should it when the central bank is as ignorant of the longerterm future as they are? (Goodhart 2012) [A]s implemented thus far it is not clear why anyone should pay much attention to forward guidance as it is, in our view, mostly ‘cheap talk’. (Buiter 2013, 2) Anyone who awaits central bank predictions of inflation two years ahead in the hope of finding out how prices are likely to change has not been paying attention. When the European Central Bank sets out its predictions for inflation on Thursday, it will be blind luck if the numbers turn out to be right in 2016 . . . Whether it is right is neither here nor there, though. The forecasts matter for how they are used; if the ECB is to take a big step further into experimental monetary policy, it needs to predict inflation will stay well below target. (Mackintosh 2014)

207

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

These statements suggest that forward guidance has caused central bank watchers to adopt a view of central bank forecasts that is more cynical in the sense that it subscribes to Gramsci’s (1971, 171) dictum, cited in the epigraph of this chapter, that there is no such thing as ‘a purely “objective” prediction’. In turn, this highlights a related problem—the time-inconsistency of forward guidance: once the economy improves and inflation rates start rising, central banks with price stability mandates would have a strong incentive to increase interest rates ahead of the schedule they had previously announced (Buiter 2013, 6; Issing 2014, 7; Woodford 2013, 6). The cost of forward guidance is now increasingly clear: central banks have had to mark down their two most jealously guarded assets: their epistemic authority and their commitment credibility. The case of forward guidance holds an important lesson regarding the production of imagined futures and the coordination of expectations under conditions of uncertainty. Under such conditions, ‘it is not accurate predictions of future states of the world that determine decisions, but rather a political game of negotiation and manipulation of the interpretation of a situation’ (Beckert 2013a, 342). From this perspective, the question is not so much why market actors have become more cynical about ‘cheap talk’ by central banks, but why they have become so only now. Until recently, there was little indication that markets considered central bank forecasts as rhetorical devices designed to manipulate their expectations. That this has recently changed suggests that unconventional monetary policies have been testing the limits of the apparatus of expectation management, and thus of performative macroeconomic governance. The potentially resulting loss of governability has been compensated for, however, by the introduction of balance-sheet policy. But while complementary in the goal dimension (lowering long-term interest rates), balance-sheet policy is fundamentally different from forward guidance when viewed as an instrument.

Non-Market Price Setting: The Return of Hydraulic Macroeconomic Governance QE differs significantly from forward guidance in its implications for the nature of economic state agency. Consider the following pleas for euro-area QE, the first by a former ECB Executive Board member, the second by the chief European economist of Goldman Sachs: [I]f central banks really want to change the shape of the yield curve they may have to do more than just talk. . . . In other words, if they want to be effective central banks have to put their money, and balance sheets, where their mouths are. (Bini Smaghi 2013)

208

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Central Bank Planning If you can retain credibility and confidence, that may be all you have to do; markets will do the heavy lifting for you . . . But not all problems can be solved by shifting market expectations and behaviour. Sometimes fundamental changes are needed. (Huw Pill, quoted in Atkins 2014)

These statements chime with the ‘crisis of discursive central banking’ argument (Gabor and Jessop 2015), according to which monetary policy by communication alone has reached the end of the line. Indeed, all leading central banks have concluded that sustained macroeconomic stimulus requires use of the ‘consolidated government sector balance sheet’ (Borio and Disyatat 2009, 2). While it is commonplace to argue that QE has an important fiscal dimension, this chapter highlights another striking parallel between state agency under Keynesian fiscal policy and unconventional monetary policy. Keynesian demand management policy involves the government expanding its balance sheet to purchase goods and services in order to stimulate the (real) economy directly; QE involves the central bank expanding its balance sheet to purchase financial assets in order to stimulate the financial economy directly (with an intended second-round stimulus effect for the real economy). In other words, central banking has acquired what had previously been the exclusive domain of fiscal policy—hydraulic macroeconomic agency. Highlighting the hydraulic transmission mechanism of quantitative easing is not to say that it represents a return to the Keynesian past. On the contrary, QE represents an adaptation of macroeconomic governance to the workings of a financialized economy. Among monetary and financial economists, the idea has recently gained traction that financial market developments have spawned a new transmission channel of monetary policy, the so-called ‘risktaking channel’ (Borio and Zhu 2012). Under this mechanism, changes in interest rates and market expectations about their future path may alter the ‘perceptions of risks and risk tolerance’ of financial firms (ibid. 237). This channel has gained in importance as a result of ‘financial liberalization and innovation’, which have increased the responsiveness of credit creation to swings in risk perception, and therefore the impact of the latter on aggregate demand (ibid. 237). The ECB shares this analysis with regard to the monetary transmission mechanism in the euro area (ECB 2010, 85, 89). Under these conditions, QE is supposed to stimulate aggregate demand through the so-called portfolio rebalancing effect (Draghi 2015). As pointed out by ECB Executive Board member Benoît Cœuré (2015, 2), this is due to a quasi-hydraulic mechanism whereby central bank asset purchases ‘mechanically reduce the supply of securities’. As mentioned earlier in this chapter, the ECB has been buying government bonds with average remaining maturities of eight years. Due to their greater scarcity, the prices of bonds at these and other maturities rise, while yields fall.

209

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures 4

3

Yield in %

2

1

0

–1

3m

6m

1y

2y

3y

5y

7y

10y

15y

20y

30y

Maturity in months and years 4/1/2014

8/1/2014

12/1/2014

12/1/2015

7/1/2016

Figure 9.1 Yield curves for bonds of euro-area governments (all issuers and ratings) Note: Curves represent snapshots taken on the dates indicated in the key to the chart. Source: ECB

It is through this mechanism that QE ‘bends’ the long end of the yield curve. Figure 9.1 shows how the term structure of interest rates has shifted for government bonds in the euro area. While shifts of the yield curve cannot be attributed to monetary policy alone, the disproportionate compression of interest rates at the long end of the yield curve from 2014 onwards suggests that the ECB’s twin policies of forward guidance and QE achieved the desired effect. According to the logic of the portfolio rebalancing channel, this compression sets off a ‘search for yield’ among investors, who move out of low-risk securities into higher risk assets, such as corporate bonds, equities, or loans to firms and households (Cœuré 2015, 2). In addition, the balance sheets of the owners of securities are strengthened as a result of rising financial asset prices.8 In short, central bank security purchases mechanically increase demand— and thus the price—for certain financial assets. While forward guidance has revealed the limitations of performative macroeconomic agency, QE has provided central banks with a hydraulic tool akin to fiscal policy: whereas government spending on goods and services increases firm revenues and household incomes, central bank spending on asset purchases is expected to increase financial firms’ revenues and bolster their balance sheets.

8

210

For a succinct discussion of QE transmission channels, see Haldane et al. (2017, 7–9).

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Central Bank Planning

Conclusion This chapter has argued that the performative nature of central banking has deeper roots in macroeconomic theory than has been acknowledged in the literature. While prominent economists have criticized the rational expectations hypothesis and DSGE models as ‘figments’ and as ‘post-real’ (Frydman and Goldberg 2011; Romer 2017), the performative consequences of these theoretical innovations have been real enough. What options traders did for the efficient market hypothesis, central bankers did for the rational expectations hypothesis. As a result, two key features of the ‘model world’ (Watson 2014)—‘communism of models’ and non-market price setting—gained a foothold in the real world. Focusing primarily on the euro area, the chapter has highlighted the complementary nature of forward guidance and QE, both of which are geared towards bringing longer-term interest rates under central bank control. Deploying these policies to bend the yield curve, the ECB has incurred significant costs. Adding to the finding of a loss of monetary trust among the general public (Braun 2016b), the chapter has highlighted the loss of epistemic authority with financial market actors, who have voiced doubts regarding both the quality and the sincerity of the ECB’s economic forecasts. This weakening of the performative dimension of central bank agency has been compensated, to a certain extent, by QE. From a conceptual perspective, QE constitutes a monetary version—updated for the conditions of a financialized economy— of the hydraulic macroeconomic agency that used to be the hallmark of Keynesian fiscal demand management. Accounting for the costs and benefits of this hydraulic turn in monetary governance remains a major task for students of central banking. One of the key questions concerns the distributional consequences of largescale asset purchase programmes. It remains unclear whether the wealth inequality-increasing consequences of asset price inflation—which are well documented—are compensated by the income inequality-decreasing stimulus effect of asset purchases on GDP growth and employment (ECB 2017, 48–51; Fontan et al. 2016). The other big outstanding question concerns the governability consequences of central (bank) planning. In the 1980s, when central banks still cloaked themselves in obscurity, their justification echoed the Hayekian argument that uncertainty, by virtue of creating expectational diversity, actually has a stabilizing effect. One FOMC member feared that openness about the Fed’s intentions would cause the market to ‘move with a single purpose based on accurate knowledge of the short run objectives of the market’s largest participant, the FOMC.’ (Goodfriend 1986, 77). With forward guidance and QE, bringing about this ‘single purpose’ has become a key plank of monetary 211

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

policy. The price tag of non-market price setting via unconventional monetary policy therefore includes the loss of informational content in financial asset prices. Once a barometer of the decentralized beliefs and actions of myriad market actors, the long-term interest rate has become a policy variable, manipulated by central banks to reduce the ‘perceived downside risk’ for investors (BIS 2013, 1). The potential costs are manifold—increased risktaking, indebtedness, collateral scarcity, and financial instability, to name but a few. It remains to be seen if bending and de-risking the yield curve will make the future less uncertain.

Bibliography Abolafia, Mitchel Y., and Deneen M. Hatmaker. 2013. ‘Fine-Tuning the Signal: Image and Identity at the Federal Reserve’. International Public Management Journal 16 (4): pp. 532–56. Atkins, Ralph. 2014. ‘Draghi Performs Another Euro Conjuring Trick’. Financial Times, 9 May. Accessed 30 January 2017. http://www.ft.com/cms/s/0/450424ec-d77f-11e380e0-00144feabdc0.html?siteedition=uk-axzz3BMMYBWgO. Auer, Raphael, Claudio E. V. Borio, and Andrew J. Filardo. 2017. ‘The Globalisation of Inflation: The Growing Importance of Global Value Chains’. BIS Working Paper 602, Bank for International Settlements, Basel. Baker, Andrew, and Wesley Widmaier. 2014. ‘The Institutionalist Roots of Macroprudential Ideas: Veblen and Galbraith on Regulation, Policy Success and Overconfidence’. New Political Economy 19 (4): 487–506. Barro, Robert J., and David B. Gordon. 1983. ‘Rules, Discretion and Reputation in a Model of Monetary Policy’. Journal of Monetary Economics 12 (1): pp. 101–21. Beckert, Jens. 2013a. ‘Capitalism as a System of Expectations: Toward a Sociological Microfoundation of Political Economy’. Politics & Society 41 (3): pp. 323–50. Beckert, Jens. 2013b. ‘Imagined Futures: Fictional Expectations in the Economy’. Theory & Society 42 (3): pp. 219–40. Beckert, Jens. 2016. Imagined Futures: Fictional Expectations and Capitalist Dynamics. Cambridge, MA: Harvard University Press. Bini Smaghi, Lorenzo. 2013. ‘Why Forward Guidance has Failed to Convince’. Financial Times, 10 September. Accessed 30 January 2017. https://www.ft.com/content/ 74a2ac46-bd75-3117-a411-fc777f5f2d6f. BIS. 2013. ‘Markets Grow Confident on Continued Support’. Bank for International Settlements Quarterly Review (March): pp. 1–13. BIS. 2016. ‘86th Annual Report, 2015/16’. Basel: Bank for International Settlements. Blinder, Alan S. 2000. ‘Central-Bank Credibility: Why Do We Care? How Do We Build It?’. American Economic Review 90 (5): pp. 1421–31. Blinder, Alan S. 2004. The Quiet Revolution: Central Banking Goes Modern. New Haven, CT: Yale University Press.

212

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Central Bank Planning Boldyrev, Ivan, and Olessia Kirtchik. 2017. ‘The Cultures of Mathematical Economics in the Postwar Soviet Union: More than a Method, Less than a Discipline’. Studies in History and Philosophy of Science Part A 63: pp. 1–10. DOI: 10.1016/j.shpsa.2017.03.011. Borio, Claudio, and Piti Disyatat. 2009. ‘Unconventional Monetary Policies: An Appraisal’. BIS Working Paper no 292, Bank for International Settlements, Basel. Borio, Claudio, and Haibin Zhu. 2012. ‘Capital Regulation, Risk-Taking and Monetary Policy: A Missing Link in the Transmission Mechanism?’ Journal of Financial Stability 8 (4): pp. 236–51. Braun, Benjamin. 2014. ‘Why Models Matter: The Making and Unmaking of Governability in Macroeconomic Discourse’. Journal of Critical Globalisation Studies 7: pp. 48–79. Braun, Benjamin. 2015. ‘Governing the Future: The European Central Bank’s Expectation Management during the Great Moderation’. Economy and Society 44 (3): pp. 367–91. Braun, Benjamin. 2016a. ‘From Performativity to Political Economy: Index Investing, ETFs and Asset Manager Capitalism’. New Political Economy 21 (3): pp. 257–73. Braun, Benjamin. 2016b. ‘Speaking to the People? Money, Trust, and Central Bank Legitimacy in the Age of Quantitative Easing’. Review of International Political Economy 23 (6): pp. 1064–92. Bronk, Richard. 2011. ‘Uncertainty, Modelling Monocultures and the Financial Crisis’. Business Economist 42 (2): pp. 5–18. Bronk, Richard. 2013. ‘Hayek on the Wisdom of Prices: A Reassessment’. Erasmus Journal for Philosophy and Economics 6 (1): pp. 82–107. Bronk, Richard, and Wade Jacoby. 2016. ‘Uncertainty and the Dangers of Monocultures in Regulation, Analysis, and Practice’. MPIfG Discussion Paper 16/6. Max Planck Institute for the Study of Societies, Cologne. Buiter, Willem. 2013. ‘Forward Guidance: More than Old Wine in New Bottles and Cheap Talk?’. Citigroup Global Economics View. New York: Citi Research. Caballero, Ricardo J. 2010. ‘Macroeconomics after the Crisis: Time to Deal with the Pretense-of-Knowledge Syndrome’. Journal of Economic Perspectives 24 (4): pp. 85–102. Campbell, Jeffrey R., Charles L. Evans, Jonas D. M. Fisher, Alejandro Justiniano, Charles W. Calomiris, and Michael Woodford. 2012. ‘Macroeconomic Effects of Federal Reserve Forward Guidance’. Brookings Papers on Economic Activity (Spring 2012): pp. 1–80. Christophers, Brett. 2017. ‘The Performativity of the Yield Curve’. Journal of Cultural Economy 10 (1): pp. 63–80. Cœuré, Benoît. 2015. ‘Embarking on Public Sector Asset Purchases’. Speech at the Second International Conference on Sovereign Bond Markets, Frankfurt, 10 March. Conti-Brown, Peter. 2016. The Power and Independence of the Federal Reserve. Princeton, NJ: Princeton University Press. Draghi, Mario. 2013. ‘Introductory Statement to the Press Conference’, 4 July. http:// www.ecb.europa.eu/press/pressconf/2013/html/is130704.en.html. Draghi, Mario. 2015. ‘Introductory Statement to the Press Conference’, 22 January. http://www.ecb.europa.eu/press/pressconf/2015/html/is150122.en.html. ECB 2008. ‘The Analysis of the Euro Money Market from a Monetary Policy Perspective’. Monthly Bulletin (February): pp. 65–79.

213

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures ECB 2010. ‘Monetary Policy Transmission in the Euro Area, a Decade After the Introduction of the Euro’. Monthly Bulletin (May): pp. 85–98. ECB 2014. ‘The ECB’s Forward Guidance’. Monthly Bulletin (April): pp. 65–73. ECB 2017. Annual Report 2016. Frankfurt: European Central Bank. Evans, George W., and Seppo Honkapohja. 2005. ‘An Interview with Thomas J. Sargent’. Macroeconomic Dynamics 9 (4): pp. 561–83. Filardo, Andrew, and Boris Hofmann. 2014. ‘Forward Guidance at the Zero Lower Bound’. BIS Quarterly Review (March): pp. 37–53. FOMC 2013. ‘Federal Reserve issues FOMC statement’, 18 December 2013. http://www. federalreserve.gov/newsevents/press/monetary/20131218a.htm. Fontan, Clément, François Claveau, and Peter Dietsch. 2016. ‘Central Banking and Inequalities: Taking Off the Blinders’. Politics, Philosophy & Economics 15 (4): 319–57. Friedman, Benjamin. 2014. ‘Has the Financial Crisis Permanently Changed the Practice of Monetary Policy? Has It Changed the Theory of Monetary Policy?’. NBER Working Paper no 20128. Frydman, Roman, and Michael D. Goldberg. 2011. Beyond Mechanical Markets: Asset Price Swings, Risk, and the Role of the State. Princeton, NJ: Princeton University Press. Gabor, Daniela, and Bob Jessop. 2015. ‘Mark My Words: Discursive Central Banking in Crisis’. In Financial Cultures and Crisis Dynamics, edited by Bob Jessop, Brigitte Young and Christoph Scherrer, pp. 294–315. London: Routledge. Giddens, Anthony. 1987. ‘Social Theory and Problems of Macroeconomics’. In Social Theory and Modern Sociology, by Anthony Giddens, pp. 183–202. Cambridge: Polity Press. Goffman, Erving. 1959. The Presentation of Self in Everyday Life. Garden City, NY: Double Day. Golub, Stephen, Ayse Kaya, and Michael Reay. 2015. ‘What Were They Thinking? The Federal Reserve in the Run-Up to the 2008 Financial Crisis’. Review of International Political Economy 22 (4): pp. 657–92. Goodfriend, Marvin. 1986. ‘Monetary Mystique: Secrecy and Central Banking’. Journal of Monetary Economics 17 (1): pp. 63–92. Goodhart, Charles. 2012. ‘Longer-Term Forecasts Are a Step Backwards’. Financial Times, 1 February. Accessed 30 January 2017. http://www.ft.com/intl/cms/s/0/ 2034b62a-4c00-11e1-b1b5-00144feabdc0.html. Gramsci, Antonio. 1971. Selections from the Prison Notebooks of Antonio Gramsci. Edited and translated by Quintin Hoare and Geoffrey N. Smith. London: Lawrence & Wishart. Haldane, Andrew G. 1995. ‘Rules, Discretion and the United Kingdom’s New Monetary Framework’. Bank of England Working Paper no 40. Haldane, Andrew G. 2017. ‘A Little More Conversation, A Little Less Action’. Speech at the Macroeconomics and Monetary Policy Conference, Federal Reserve Bank of San Francisco, 31 March. Haldane, Andrew G., Matt Roberts-Sklar, Tomasz Wieladek, and Chris Young. 2017. ‘QE: The Story So Far’. Bank of England Staff Working Paper no 624. Hall, Peter A. 1993. ‘Policy Paradigms, Social Learning, and the State: The Case of Economic Policymaking in Britain’. Comparative Politics 25 (3): pp. 275–96.

214

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Central Bank Planning Hayek, Friedrich A. 1945. ‘The Use of Knowledge in Society’. American Economic Review 35 (4): pp. 519–30. Hayek, Friedrich A. 1989. ‘The Pretence of Knowledge (Nobel Lecture 1974)’. American Economic Review 79 (6): pp. 3–7. Heimberger, Philipp, and Jakob Kapeller. 2017. ‘The Performativity of Potential Output: Pro-Cyclicality and Path Dependency in Coordinating European Fiscal Policies’. Review of International Political Economy 24 (5): pp. 904–28. Holmes, Douglas R. 2014. Economy of Words: Communicative Imperatives in Central Banks. Chicago. IL: University of Chicago Press. Issing, Otmar. 1992. ‘Theoretical and Empirical Foundations of the Deutsche Bundesbank’s Monetary Targeting’. Intereconomics 27 (6): pp. 289–300. Issing, Otmar. 2014. ‘Forward Guidance: A New Challenge for Central Banks’. SAFE White Paper Series no 16, Goethe University Frankfurt, Research Center SAFE – Sustainable Architecture for Finance in Europe. Juselius, Mikael, Claudio Borio, Piti Disyatat, and Mathias Drehmann. 2016. ‘Monetary Policy, the Financial Cycle and Ultralow Interest Rates’. Bank of Finland Research Discussion Paper 24–2016. Konings, Martijn. 2016. ‘Governing the System: Risk, Finance, and Neoliberal Reason’. European Journal of International Relations 22 (2): pp. 268–88. Krippner, Greta R. 2007. ‘The Making of US Monetary Policy: Central Bank Transparency and the Neoliberal Dilemma’. Theory and Society 36 (6): 477–513. Kydland, Finn E., and Edward C. Prescott. 1977. ‘Rules Rather than Discretion: The Inconsistency of Optimal Plans’. Journal of Political Economy 85 (3): pp. 473–91. Lucas, Robert E., and Thomas J. Sargent. 1979. ‘After Keynesian Macroeconomics’. Quarterly Review 3 (2): pp. 49–72. https://www.bostonfed.org/-/media/Documents/ conference/19/conf19d.pdf. MacKenzie, Donald. 2006. An Engine, Not a Camera: How Financial Models Shape Markets. Cambridge, MA: MIT Press. Mackintosh, James. 2014. ‘E.C.B. Left to Ponder Onset of Stealth Tightening’. Financial Times, 4 March. Accessed 30 January 2017. http://www.ft.com/cms/s/0/ca022222a3c5-11e3-aa85-00144feab7de.html?siteedition=uk-axzz39vUAgl99. Marcussen, Martin. 2009. ‘Scientization of Central Banking: The Politics of A-Politicization’. In Central Banks in the Age of the Euro: Europeanization, Convergence, and Power, edited by Kenneth Dyson and Martin Marcussen, pp. 373–90. Oxford: Oxford University Press. Mirowski, Philip. 2002. Machine Dreams: Economics Becomes a Cyborg Science. Cambridge: Cambridge University Press. Mitchell, Timothy. 2005. ‘The Work of Economics: How a Discipline Makes Its World’. European Journal of Sociology 46 (2): pp. 297–20. Morris, Stephen, and Hyun Song Shin. 2002. ‘Social Value of Public Information’. American Economic Review 92 (5): pp. 1521–34. Morris, Stephen, and Hyun Song Shin. 2008. ‘Coordinating Expectations in Monetary Policy’. In Central Banks as Economic Institutions, edited by Jean-Philippe Touffut, pp. 88–104. Cheltenham, UK: Edward Elgar.

215

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures Mudge, Stephanie L., and Antoine Vauchez. 2016. ‘Fielding Supranationalism: The European Central Bank as a Field Effect’. Sociological Review Monographs 64 (2): 146–69. Muth, John F. 1961. ‘Rational Expectations and the Theory of Price Movements’. Econometrica 29 (3): pp. 315–35. Nelson, Stephen C., and Peter J. Katzenstein. 2014. ‘Uncertainty, Risk, and the Financial Crisis of 2008’. International Organization 68 (2): pp. 361–92. Pahl, Hanno, and Jan Sparsam. 2016. ‘The IS-LMization of the General Theory and the Construction of Hydraulic Governability in Postwar Keynesian Macroeconomics’. In Enacting Dismal Science, edited by Ivan Boldyrev and Ekaterina Svetlova, 151–81. New York: Palgrave Macmillan. Poole, William, and Robert H. Rasche. 2000. ‘Perfecting the Market’s Knowledge of Monetary Policy’. Journal of Financial Services Research 18 (2–3): 255–98. Praet, Peter. 2014. ‘Current Issues and Challenges for Central Bank Communication’. Speech at the ECB and its Watchers XV, Frankfurt, 12 March. Romer, Paul. 2017. ‘The Trouble with Macroeconomics’. The American Economist, forthcoming. Rosenhek, Zeev. 2013. ‘Diagnosing and Explaining the Global Financial Crisis: Central Banks, Epistemic Authority, and Sense Making’. International Journal of Politics, Culture, and Society 26 (3): pp. 255–72. Shin, Hyun Song. 2017. ‘How Much Should We Read Into Shifts in Long-Dated Yields?’. Speech at the US Monetary Policy Forum, New York City, 3 March. Turner, Philip. 2011. ‘Is the Long-Term Interest Rate a Policy Victim, a Policy Variable or a Policy Lodestar?’ BIS Working Paper no 367, Bank for International Settlements, Basel. Valiante, Diego. 2017. ‘The “Visible Hand” of the ECB’s first Quantitative Easing’. International Economics and Economic Policy 14 (4): pp. 601–24. Watson, Matthew. 2014. Uneconomic Economics and the Crisis of the Model World. London: Palgrave Macmillan. Widmaier, Wesley W. 2016. Economic Ideas in Political Time. Cambridge: Cambridge University Press. Woodford, Michael. 2003. Interest and Prices: Foundations of a Theory of Monetary Policy. Princeton, NJ: Princeton University Press. Woodford, Michael. 2009. ‘Convergence in Macroeconomics: Elements of the New Synthesis’. American Economic Journal: Macroeconomics 1 (1): 267–79. Woodford, Michael. 2013. ‘Forward Guidance by Inflation-Targeting Central Banks’. CEPR Discussion Paper no 9722, Centre for Economic Policy Research, London.

216

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Section IV Constructing Futures in Finance

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

10 Predicted Uncertainty Volatility Calculus and the Indeterminacy of the Future Elena Esposito

The Uncertainty of the Open Future The specificity of finance is not that it faces the uncertainty of the future, but that it makes uncertainty a resource. In modern society this uncertainty is a chronic condition that few of us would willingly give up. The uncertainty of the future is the flip side of a future that is open (Koselleck 1979): it is the basic condition that allows for self-determination and freedom, dissolves the puzzles of individual free will, and radically changes the relationship with the past. None of us would be comfortable with the idea of a ‘closed’ future, in which our actions can only confirm (willingly or not: see Oedipus) an already decided order and produce a predetermined future. Our future, the future of modern society, is open in the sense that it is not already decided—in contrast to the closed notion of time assumed in earlier societies (Koyré 1957). In ancient and pre-modern societies, temporal uncertainty was assumed to concern only human beings, who have a limited perspective and not enough information to know what will happen tomorrow. Time (tempus) was the restricted temporal dimension of human beings. A higher, all-seeing entity could know the course of things and referred to a different temporal dimension: eternity. From its perspective both the past and the future were given and knowable, in an ‘aeternitas’ that was indifferent to the passing of days (Luhmann 1991, 42). Even if human beings could never overcome uncertainty (it would have been foolhardy to put oneself at the same level as the deity), it was not considered radical because it relied on the assumption of

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

an ultimate order of things that existed and was certain—although inaccessible to us. The idea of an open future questions this order and inevitably introduces a radical uncertainty affecting every observer, God included. Our open future is undetermined and cannot be otherwise. No one, not even God, can know the future now, because it does not yet exist and will be produced also as a result of the present and future contingent actions and decisions of ourselves and others (Luhmann 1980; Shackle 1972). The human capacity for freedom and invention is the flip side of the indeterminacy of the future. This holds for many areas of society, from private life to public regulations, and from organizations to families and individual careers. It is not a specific prerogative of economy and finance. But financial markets, in particular, have to face the puzzle (and the opportunity) of building expectations about the uncertain future; that is, they have the task of ‘exploiting the freedom of unknowledge’ (Shackle 1979, 92). In finance, you often have to take decisions today that should bring benefits tomorrow, without exhaustive information about future conditions and without anyone else having it. Here uncertainty is not a diffused existential state, but the very condition of the decision. Not surprisingly in this context, the discussion of uncertainty and its management has been particularly intense and problematic at least since Knight (1921 [1964]; see, for example, Bronk and Jacoby 2016; Demange and Laroque 2006; Goodhart 1989; Rizzo 1979; Schmidt 1996; and Shackle 1955 [1990]). In finance you must decide, and the lack of knowledge produces practical dilemmas. How can you deal with uncertainty in a rational way, maintaining a form of control that guarantees that the decision about the unpredictable future is not taken arbitrarily? This chapter discusses the way in which structured finance manages and controls the openness of the future as a source of profit. To do this, a specific form of fiction is needed, based on the construction of a present image of the future and its risks. This is expressed by the evaluation of implied volatility, which is the main focus of this chapter. The problem with the approach taken by structured finance, which ultimately led to the financial crisis of 2007–8 and ensuing problems, is not that it is based on a fiction (which is inevitable), but that it does not take sufficient account of its consequences—that is, of the circular way in which the fictional future affects the (not-yet-existing) future reality. The argument presented here relies on the dual nature of the future: the intersection and combination between the present future and the future present. The chapter claims that financial models that promise in the present to neutralize risk tend instead, in times of turbulence, to increase the unpredictability of the future present. The conclusion proposed is that we should consider the dependence of the future on the present in 220

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Volatility Calculus and Indeterminacy

terms of discontinuity rather than continuity: what we do today affects future reality, but usually not as we expect.

Radical Uncertainty and Measurable Uncertainty or ‘Risk’ Finance’s relationship with uncertainty remains shaped by the enlightening and powerful distinction between risk and uncertainty proposed by Frank Knight in 1921. Referring to the tools of probability calculus, Knight observes that, in some cases, even if a given outcome is not certain, the numerical probability of its occurrence is known. In these cases of ‘measurable uncertainty’ it is possible rationally to programme one’s activities with tools such as insurance or statistical procedures that actually neutralize uncertainty. Calculable uncertainty, which Knight calls ‘risk’, ‘is not in effect an uncertainty at all’ if properly managed (Knight 1921 [1964], 20). Risk can be calculated using data on past regularities that are assumed to persist. There is, however, another kind of uncertainty that cannot be calculated or controlled, however much information you obtain, because it is related to an insuperable imperfection of knowledge about the future (Knight 1921 [1964], 197ff, 233ff). This uncertainty cannot be eliminated by using probability calculations or other statistical procedures because it depends, for example, on the uniqueness of the contingent particular. Knight believes that this uncertainty is the very basis of entrepreneurs’ ability to obtain profits and is the true engine of the economy’s dynamics. As he puts it: ‘If all changes were to take place in accordance with invariable and universally known laws, they could be foreseen for an indefinite period in advance of their occurrence . . . and profit (or loss) would not arise’ (Knight 1921 [1964], 198). In other words, the entrepreneur is not rewarded for taking risks that can be perfectly calculated according to known regularities but, instead, ‘for an ability to exploit uncertainty’ (Stark 2009, 15). Although mainstream economics is still oriented to computable risks, Knight’s distinction is an unavoidable reference in all discussions of uncertainty, eventually becoming a kind of dogma (Luhmann 1991, 9). Anyone who discusses uncertainty and tries to develop Knight’s approach further is exposed to the accusation of misunderstandings or mistakes. Almost a century has gone by since Knight’s reflections, however, and finance has changed enormously. In particular, since the 1970s structured finance has spread, using advanced financial instruments (derivatives, securitizations, CDOs, and others) to manage uncertainty and risk. The goal, in many cases, is to package and repackage risks in order to create a range of asset classes that allow for a rational calculation of investment risk and a viable market for these products. The promise of these tools, in Knight’s terms, is to transpose more 221

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

and more uncertainty into risks—that is, into forms that, given the computational capacity of computers, can be estimated with the available formalized models (CAPM, VaR, and others) and hence allow for the creation of standardized and marketable products. To what extent is this promise realistic? Is Knight’s distinction between risk and uncertainty still adequate to describe the ongoing processes? Is calculated risk on financial markets actually not uncertainty at all or is it a fiction?

Present Future and Future Present The promise of structured finance was to have found the tools to handle the unpredictability of the future reliably, and this gave impetus to the enormous expansion of finance from the 1970s onwards, in what amounted to a systematic use of the future in the present (Esposito 2011). The models it uses and the instruments on which it relies (primarily derivatives) seem to be able to ‘commodify’ uncertainty, turning it into a good that can be traded on markets, buying, selling, and reselling it to make a profit (Arnoldi 2004, 23–6; Bryan and Rafferty 2007, 136). In order to deal with uncertainty as a commodity, however, it must be priced, estimating the ‘vagaries’ of the future even in the awareness that it cannot be exactly predicted. Derivatives are a fundamental tool because they refer to this problem directly. In futures, for instance, the uncertainty of the future is expressed by the fact that the futures price1 of an asset does not coincide with the expected value of the future spot price. In other words, the futures price in three months is not the same as the expected price of the asset three months later.2 The present evaluation of the future is different from the predicted future evaluation. The opportunity to make a profit relies on the difference between these two prices, and on exploiting the difference between what can be expected today and what will happen in the future—the exploitation of uncertainty. In order for this uncertainty3 to be bought and sold—that is, in order to trade with derivatives—its price must have an objectivity, albeit a provisional one. The measure used to determine the amount to pay must stand as a fact, one that properly, or at least adequately, represents the object at stake. Only in this way does hedging with derivatives, for example, make sense, claiming to reflect the movements of the underlying market. This is why, according to 1

The price at which parties to a futures contract agree to transact upon the settlement date. There are technical words to describe this circumstance. The situation in which the futures price is below the expected future spot price is called ‘normal backwardation’; the one in which it is higher is called ‘contango’. 3 Expressed in the form of volatility: see the section ‘The Evaluation of Future Uncertainty’. 2

222

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Volatility Calculus and Indeterminacy

many observers, modern finance was born in 1973 (for instance, MacKenzie 2006, chapter 5; Mandelbrot and Hudson 2004, 161; Millmann 1995, 47), when a way was devised to give uncertainty a price that seemed to be objective and independent from the idiosyncratic estimates of individuals. Uncertainty was allegedly transformed into calculable risk. This date corresponds to the spread of the Black–Scholes4 formula to price options in financial markets. The scheme is ingenious and presupposes a precise conceptualization (and neutralization) of uncertainty. The difficulty of pricing options5 rests on a traditional lack of clarity on the difference between two dimensions of the future: the present future and the future present (Luhmann 1976, 140; 1991, chapter 2). This chapter will argue that the interplay of these two dimensions helps us to evaluate the claim of structured finance and derivatives trading to exploit uncertainty by turning it into a measurable commodity. The present future is the future as we can expect it today, on the basis of currently available information and statistical models: the future from the view of the present. It can be observed, specified, and controlled, but it is always a fiction, because it refers to a reality that does not exist yet. The future present, on the other hand, will be real but at a later time, as a result also of today’s actions, choices, and innovations. Nobody can know it with certainty now. One would think that, in order to fix the price of derivatives, one ought to know (or at least estimate) what price the underlying asset will have at maturity. One should know the future present. On this level, there is obviously no possible objectivity. Black–Scholes’ solution marked a turning point because it avoids the problem by moving to the time perspective of the present future, which has a kind of objectivity, based on what one can presently know of the asset at stake. It is not an objective assessment of an unknowable future reality, but the evaluation of a present given: the current image of the future and the information on which it is based. A way was devised to give uncertainty a price that seems objective and independent of the idiosyncratic assessments of any individual, allowing commodification and exchange on markets.

The Evaluation of Future Uncertainty The keystone of the Black–Scholes formula is the calculation of volatility of securities and of markets—that is, their variability over time and tendency to react to circumstances. This variability can be calculated, generating a value 4 Or ‘Black–Scholes–Merton’, as MacKenzie (2006) calls it in order to acknowledge the role of Robert C. Merton. However, in the debate on the pricing of derivatives, one speaks mostly of the Black–Scholes formula, and we will conform here to the prevailing habit. See also Black and Scholes (1973). 5 A difficulty that tormented all attempts to develop a formula about it, from Vinzenz Bronzin to Paul Samuelson; see also Hafner and Zimmerman (2009).

223

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

that stands for the risk faced by actors. Taking into account volatility, time (the longer the time, the higher the risk), and the present value of an asset (which includes an estimate of its riskiness), one can give the option a price that allegedly corresponds to an objective estimate, a price that all operators can agree upon and take as a basis for their calculations. Volatility expresses the intensity of the variation of the price of the underlying asset, the speed of the market. It is a value that corresponds to the amount of the variation, not to its direction. It does not tell us whether an asset is going good or bad, only the degree to which it fluctuates. If one knows that an asset has low volatility, for example, one can expect that its value will not decrease sharply, nor will its price increase significantly. If volatility increases, the risk to which operators are exposed also increases, given that unpredictability rises. Those who buy or sell options deal with volatility, with an estimate of operators’ variability of loss or gain. The trend of volatility is often more important than the price of the underlying asset. Those who trade with options are successful if they are able to guess how much operators expect prices to change (by guessing the management of risk), not by guessing how prices themselves will change. If, for example, one buys a call expecting an increase, a situation can arise in which, while the market rises, volatility decreases, causing the option to lose value (risk seems to decrease). When selling only volatility, earnings can occur even when the underlying market does not drop or rise (Lee 2015). In general, if volatility increases, the value of options rises (risk increases). If it decreases, they become cheaper. The problem is that future volatility remains unknown, and therefore, unaccounted for. While one can know the past, operators are interested in future uncertainty, which cannot be known. Past tendencies are informative, and this information is actively used by the markets, but it indicates only what observers expect for the future, not what the future will be. Predictions based on this information affect real movements, although nobody knows how. Financial practice distinguishes the following three kinds of volatility, corresponding to three distinct observation perspectives. (i) Historical volatility is relative to the past and is measured by the deviation of the values of an asset from the average. This form of volatility is a direct measurement of the price movements of the underlying asset in a given time period. It is high if the asset was turbulent, low if it remained quiet. Like all data based on the past, it is a kind of reference which appeals to certainty, but says little about the future trends of the asset at stake in a restless market. There is nothing that prevents a stable asset from beginning to oscillate suddenly. Historical volatility can, however, affect the other kinds of volatility. 224

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Volatility Calculus and Indeterminacy

(ii) Advanced volatility expresses the subjective expectations of an operator and, therefore, how one expects the asset to behave. Like all subjective perspectives, it is always uncertain and opaque. One cannot know with certainty what people think, or whether they are right. (iii) Implied volatility is strange and self-referential. It has become the key concept for the traffic of derivative markets as a whole. It does not depend on the price variations of the underlying asset (as historical volatility does), but on the prices of the derivative itself. It does not measure how turbulent the movements of the underlying asset are, but how variable the expectations of the operators about these movements are. Different derivative contracts on the same underlying asset (for instance, a stock) will normally have different implied volatility, depending on the duration of the contract (an option expiring in six months will have a different implied volatility than an option expiring in one month for the same underlying asset). Implied volatility expresses an estimate of the operators’ perception of future market movements (rather than an estimate of the movements themselves, which are of course unpredictable). It expresses what it seems reasonable to expect on the basis of past movements and of the available information. One often hears that implied volatility indicates the ‘sentiment’ of the market and of investors at a given time. This makes sense only if we consider that this ‘sentiment’ can be very different not only from what will actually turn out to be true (as measured by future historical volatility), but also from what the individual subjects believe and expect—that is, from advanced volatility. The distinction between advanced volatility and implied volatility indicates that the latter does not claim to reflect what the operators really think, but what it is generally thought that they think. In other words, it indicates the prevailing opinion on the prevailing opinion.6 In this sense, everyone thinks and expects what she or he thinks, and refers to implied volatility to make profits by exploiting the difference between his or her opinion and the perceived prevailing opinion, or at least by trying to do so. Implied volatility can be understood to measure this perception of the prevailing opinion, the projection of the future from the considered present: the present future on the basis of data and information available at the time. Implied volatility, then, is not the ‘real’ volatility. It does not reflect the uncertainty of the market, but only that uncertainty that the market expects that the market expects. Like Knight’s risk, it is a calculable measure, which has its own objectivity to which one can refer. Implied volatility is actually the 6

In the sense of the ‘Keynesian beauty contest’: Keynes (1936 [1973], 156). See also Stark (2013).

225

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

reference for the models of risk assessment, for the models that price options (starting from the famous Black–Scholes formula that seemed to give objectivity to the indeterminacy of the future and of expectations). The great advantage of the Black–Scholes formula is that it found a way to estimate implied volatility, which is as circular as the notion itself and may in fact work just because of that. Implied volatility is calculated by applying the Black–Scholes model (that gives options a price) backwards.7 Once the price of an option is known, one inserts it into the formula, retrieving a value for volatility that can be used in future calculations. One thereby builds the future by projecting forward a calculation of implied volatility derived from the past. Even knowing that the future is uncertain, one expects its uncertainty to be predictable and calculable from the past. Uncertainty should thereby become transformed into calculable risk.

Performativity and Counter-Performativity If one accepts the risk assessment procedure proposed by the Black–Scholes formula and by the corresponding models (as was the case from the mid-1970s until at least the second half of the 1980s), one can assume that others will also use it to build their strategies. There is an apparently objective basis upon which one can deal with uncertainty, despite the lack of objective knowledge about uncertain futures. The derivatives market became extremely technical and formalized, with the calculation of implied volatility serving as a cornerstone and, in some instances, almost as a substitute for reality. Not only are complex strategies of hedging and speculation realized through such means, but also those of ‘volatility trading’, with options and swaps on volatility, and reflective strategies based on expectations of the evolution of volatility, ultimately yielding concrete gains and losses. Starting from the calculation of volatility, other measures can then be calculated that allow consideration of the sensitivity of options with regard to other factors that affect their prices. An example of such measures are the famous ‘greeks’ (so called because they are conventionally indicated with letters from the Greek alphabet), such as the variation in the price of the option relative to the variation in the price of the underlying (δ—delta), the decrease in value of the option as the deadline approaches (θ—theta), the relationship between the price of the option and the interest rate (ρ—rho), the variation in the value of the option relative to the variation of the volatility of the 7

Beunza and Stark (2012, 401ff) discuss the calculation of implied volatility as a case of the more general practice of backing out, which allows market actors to combine the rigidity of models with the observations of other observers.

226

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Volatility Calculus and Indeterminacy

underlying (ν —customarily standing for ‘vega’), the variation in the option when the underlying has an infinitesimal change, or its elasticity (ω—omega). A complex network of calculations of calculations is thus produced that allows for more and more detailed guidelines in the management of uncertainty, guidelines that become more influential the more they are used to direct operations. The component of performativity of financial markets, rightly underlined and analysed by many authors since Callon (1998)—see, for example, MacKenzie (2006, 2009); MacKenzie, Muniesa, and Siu (2007); and Callon, Millo, and Muniesa (2007)—now comes to light. MacKenzie, in particular, has reconstructed the history of the success of the Black–Scholes formula from this perspective. Today, after the financial crisis and the subsequent turbulences showed the inaccuracies and problems of the models of structured finance, it is easy to say that the formula ‘is simply wrong’.8 More interesting, however, is to ask how it could have been ‘right’ for so many years, effectively guiding the expectations and behaviour of operators. Even if the Black–Scholes formula is wrong, it is not wrong in a simple way. To study its evolution can be very informative with regard to the workings of derivative markets. MacKenzie’s argument is that financial theory adequately described reality, not because it was realistic in an absolute sense, but because reality itself changed as a result of the theory—that is, because of performativity, which explains how a ‘wrong’ theory could work so well and appear so plausible for many years. The problem, however, is that at a certain point performativity turned into counter-performativity (MacKenzie 2006, 184f, 259f; 2007, 75f): the theory still produced reality, but a reality different from the one it predicted. Why? How can we explain this reversal in the effect of reality production by theory (Esposito 2013)?

The Production of Unpredictability The weakness of the models of structured finance goes back to the same distinction that constitutes their efficacy: the distinction between the present future and the future present. This distinction does not reflect two alternative ways of evaluating the future, one of which will prove ‘right’ and the other ‘wrong’. They are ways to deal with two different futures, or rather two different dimensions of the future, the present future and the future present, whose interplay produces the uncertainty that finance must manage. This duplication is the condition of the open future—of the fact that today the 8

So Mandelbrot and Hudson (2004, 259 it. trans).

227

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

future does not exist yet and will come about in a way that also depends on our present actions, including the fictions and expectations built in the present (to anticipate it). The present future is the future as we can see it in the present. As Knight observed, it is inevitable for actors to react ‘to the “image” of a future state of affairs’ (Knight 1921 [1964], 201). This image is inevitably a fiction in the sense that it envisages a reality that does not (yet) exist,9 but at the same time it is not a pure fantasy because it has some basis in actual expectations or beliefs and cannot be changed at will. The same happens in novels with the characters and the events of narrative fiction. Everyone knows that Robinson Crusoe is a sailor and a castaway, even if they know that he never existed and that everyone else knows it. Nevertheless, when it comes to Crusoe and his adventures, it would simply not be true to say that he is a woman or that he died in the shipwreck, and one can assume that others share the same reference. Fiction has its objectivity10 and produces real consequences: for example, the experiences made while reading novels affect behaviour and expectations in real life. Similarly, financial traffic also produces its own kind of fiction, which is used to manage and exploit the uncertainty of the future. The present future is a fiction because it refers to a reality that does not exist yet, but it is not an arbitrary fantasy, since it uses probabilistic calculations and controlled models that help allow the actors to predict the course of events to the extent it is characterized by Knightean ‘risk’. Like all fictions, financial models about the future are extremely controlled constructions— much more so, indeed, than reality—but they are not accurate representations of a future reality. They can refer only to present images and to the currently available data derived from the past. These constructions, then, are neither simply false nor wrong. Nor are they true. Their significance is rather that they can be extremely useful for directing decisions and actions. Their fictional scenarios are controlled and nonarbitrary, and have concrete and very real consequences, because they guide and orient our decisions and our behaviour. What will be real, however, is the future present, a present that does not exist yet but will come about later, as a consequence in large part of our present decisions and actions. This makes the future inevitably unpredictable. No one can locate himself or herself in the future present, ‘ahead’ of today and its constraints, because the future reality does yet not exist and cannot be known in advance. It is the future in whose past there is today’s present, with our present future and all our attempts to

On the function of fiction in the economy, see Bronk (2009) and Beckert (2016). The objectivity of fiction relates to its narrative structure: an alternative reality with its own rules and references, which, however, does not exist (Esposito 2007; Henrich and Iser 1983). 9

10

228

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Volatility Calculus and Indeterminacy

anticipate it. Indeed, the ‘true uncertainty’ of the future is partly a product of our (innovative) present attempts to control it in the form of calculable risk. Implied volatility is a fiction in this sense. Taken as the measure of risk, it does not deal with actual future contingencies, or even with future uncertainty (nothing guarantees that a turbulent asset will go on being turbulent in the future), but concerns the expectations of observers on uncertainty—what everyone expects others to expect.11 Implied volatility measures the level of restlessness of a financial activity as one can imagine it on the basis of its past restlessness. Implied volatility concerns the present future. This is the measure that guides options pricing models. It is at the level of this fiction that one can find an order, but the measure can only appear objective because the expectations of the observers are confused with the actual future development of the world: fiction is confused with future reality. The shared fiction of finance is controlled and not subjective. It produces reality, but not necessarily the one we expect. What will happen in the future (the future present) is affected by what we do and expect today (the present future), but not necessarily because it confirms our expectations. It can deviate from the expected course because many observers are motivated by their expectations to try to prevent the expected outturn. For example, regulators may react to the spread of speculation with new regulations designed to limit it, but these new regulations may then force new innovations in structured finance that were previously unimagined. The future present, in this as in many other cases in social life, may become more unpredictable the more we try to control it. The certainty provided by the calculation of implied volatility is hard-wired into the Black–Scholes model, but—rather than reflecting what will actually happen—reflects what those using the model expect others to expect. Here the model can be highly effective in practice, as the performativity effects accompanying its diffusion show. If everyone uses the Black–Scholes model, everyone expects that a given volatility trend is expected, and volatility tends to follow that trend. But there is no guarantee that common expectations are correct, if and when the circle of performativity is broken—that is, in the cases in which, for various reasons, what operators actually expect (advanced volatility) separates from implied volatility. This typically happens in cases of panic related to stock-market crashes, in coincidence with unforeseeable events, and, in general, in all cases in which confidence in the shared fiction is shaken. Operators begin to expect unforeseeable events and, hence, an irregular and disorderly development of real uncertainty. The available models are not equipped for this kind of situation. 11 This gives rise to ‘reflexive modelling’ that combines the information derived from the past with the cognitive interdependence of traders (Beunza and Stark 2012, 404ff).

229

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

The Unpredictability of Risk In times of turbulence a different quality of risk is discovered, a ‘market internal risk’ (MacKenzie 2006, 183ff) that does not behave according to financial models and expresses the so-called (and much debated) volatility ‘smile’ or ‘skew’ in its formalization.12 What does this mean? The calculations of volatility predict a given trend of risk, which should remain substantially stable. The Black–Scholes model assumes a volatility with a regular evolution, one that allows for the calculation of implied volatility, which will be used in the forecasts for different financial activities. Volatility should increase when time distance increases, but should be the same for all strikes (that is, for the different expected prices) at a certain expiry. When the volatility smile emerges, however, the graph shows that more risky and presumably more unlikely options13 tend to have higher implied volatility than less risky options.14 Since the price of options rises when volatility increases, this means that the riskiest options are proportionately more expensive than less risky ones. This clearly contradicts any assumption that risk is under control. People are ready to pay more for risky (out-of-the-money) options as way of ‘insuring’ against extreme (and uncertain) events. Apparently, the prices reflect a situation in which unpredicted movements are feared, a situation in which it is considered wise to insure against unlikely events. A risky option is therefore paradoxically often more likely to be profitable to the seller of options in the first instance. By taking on more genuine Knightean uncertainty, the option seller can earn higher profits, at least until things go wrong. This configuration appeared after the crisis of 1987, when on 19 and 20 October the markets were submerged by a wave of sales and the Dow Jones index fell by more than twenty per cent—an event whose probability, according to the calculations used by financial theory, was less than 1 in 1050. According to the calculations, it was a practically impossible event, but it nevertheless occurred15 and produced a kind of shock, highlighting the vulnerability of markets and the fact that the best hedging techniques cannot guarantee against losses. The fit between theory and markets was lost (and will never be recovered). One began to observe that the variability of markets does not follow a normal statistical distribution (as in the movement of particles in a fluid—the Brownian motion that served as a model for options pricing 12 The volatility smile is a pattern sloping upward at either end that can emerge in the graph of implied volatility instead of the expected flat surface. 13 Options that are distant from the level of the index (out-of-the-money). 14 Options that are closer to the level of the index (at-the-money). 15 See also Mandelbrot and Hudson (2004, 6ff it. trans.); MacKenzie (2006, 184ff). A classic ‘black swan’ in the sense of Taleb (2007).

230

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Volatility Calculus and Indeterminacy

techniques), but shows the complex trends of chaotic systems, with the possibility of much wider and much more frequent fluctuations. It became clear that markets can show ‘wild variations’. The assessment of volatility that detects the level of irregularities of markets does not behave in a balanced and predictable way, but itself tends to be volatile and erratic. Deviant movements are not neutralized by an order inherent in the market, but can instead feed on themselves and increase enormously, and occur much more often than the insignificant frequencies previously predicted. The risk calculation systems used in financial markets do not take into account the inherent danger of markets reacting to themselves and their models; that is, these financial models underestimate the possibility of wild fluctuations, thereby becoming an additional factor of (intrinsic) uncertainty. Since then, markets have absorbed this experience and have absorbed the expectation of catastrophic events. Reacting to the forecasts of models, they may exhibit phenomena of counter-performativity, such as those that the ‘volatility smile’ detects. When sudden price changes occur, people tend to think that the hedging models do not work. They therefore abandon these models, further strengthening the original movements. Faced with a fall in prices, many agents panic and begin to sell, even if the hedging calculations indicate that they should wait in order to achieve the designed hedging. This waterfall of sales, in turn, produces new sales, leading to a further fall in prices. Markets react to themselves and to their calculation of risk, and this distorts the outcome and falsifies predictions made without considering this reflexivity. In this frame, the very success and spread of portfolio insurance systems becomes an additional factor of uncertainty. One even talks of ‘model risk’ (Esposito, 2011, 189; Rebonato, 2001), in order to indicate this particular kind of instability. In general, the problem with models for risk formalization is that they assume that risk behaves randomly, as in the notion of random walk that should, paradoxically, make risk controllable.16 What is treated as ‘risk’, however, turns out to be uncertainty because it does not move randomly. Uncertainty presents a series of correlations and stickiness, reacting to risk prediction and producing discontinuities and changes.17 The movements of risk remain uncertain; and it is therefore particularly risky to think that they can be controlled.

16 In the sense of the Random Walk Hypothesis (RWH), applied in this case to the movements of risk: see also Malkiel (1999) and Lo and MacKinlay (1999). 17 Mandelbrot and Hudson (2004), chapter XI, speak of two forms of wild variability: discontinuity (which depends on the features of the event) and pseudo-cycles (which depend on the order of occurrence of the events).

231

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

Conclusion Uncertainty moves in a non-random way because of forms of dependency, where the past influences the future, but not necessarily in the sense of continuity: the future present depends on the present future, but can deviate from it. What happened yesterday affects what will happen tomorrow, but not always because things will go as expected. In fact, as in episodes of counterperformativity, it can happen that one behaves in the opposite way because of past experience. The past teaches us, but we cannot predict what lessons will be taken from it. The trends of volatility show this. A wide variation makes a further variation more likely (thereby increasing the frequency of extreme events), but we do not know in which direction. A growth in volatility can be followed by a further growth, but volatility can also abruptly decrease.18 In finance, the presence of (second-order) correlations increases uncertainty rather than reducing it, and the memory of the past seems to introduce a further unpredictability, because it makes the structure of expectations even more complex and interconnected. All of these risk ‘anomalies’ are based on reflexive configurations not considered by current models. We need a theory that takes account of the reality of fiction and of the circular intertwining between the present future and the future present, and of fiction in its effects on its field of application.19 With the label ‘risk society’ (Beck 1986; Luhmann 1991), the sociological reflection on risk has already broadened the scope of economic reflection on uncertainty, including Knight’s argument. In this approach the concept of risk is brought back to the current social meaning of potential future damage, from the perspective of an actor who must decide in the present under conditions of uncertainty. The radicality of Knight’s ‘true uncertainty’ is maintained and reinforced because the notion of controlled risk becomes empty.20 Given the dependence of the future on the present, the actor knows that no calculation of probabilities or of utility functions can eliminate or neutralize the possibility of repenting in the future the present decision. It may well be that in the future other opportunities or other criteria emerge as a consequence of what you decide to do, so that it would have been better to choose otherwise. Today you can know neither what will happen tomorrow, nor the criteria by which you will judge the future. Even if the calculation were initially correct, you may perhaps discover in the meantime that the predictability of profit has led others to take the same decision (making it less profitable), or has given rise

The technical expression is ‘autoregressive conditional heteroscedasticity’. Systems theory speaks here of ‘autology’: see also Luhmann (1997, 16ff). 20 In a terminological difficulty creating a lot of confusion, what sociological theory calls ‘risk’ is much closer to the Knightean notion of uncertainty than to an alleged measurable risk. 18 19

232

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Volatility Calculus and Indeterminacy

to hostile speculation. Alternatively, you may discover that your investment success has changed the image of your company for customers, who may criticize its speculative attitude and distance themselves, thereby reducing your profits. The dependence of the future on present decisions highlighted by the events of structured finance show that the calculus of measurable risk, even when it is initially correct, is not enough to ensure security—that is, the lack of future damage—nor to ensure that operators already know the future in the present when they must decide. Measurable risk neutralizing uncertainty and providing a condition of security from any possible damage may actually be a fallacy—one that suggests the continued relevance of the sociological concept of reflective risk incorporating the broad idea of uncertainty. The calculability of risk is itself risky and one is actually always exposed to possible damage, especially when one thinks one is safe. On this view, the ‘volatility smile’ should be seen not as an anomaly to be corrected, but as evidence of how markets learn from experience and from past risk, and of how one can try to deal with these. Market dynamics reproduce uncertainty against the attempts to control it, thus reproducing the very resource that finance has always exploited. Uncertainty cannot be eliminated and the claim to do so can, paradoxically, generate new risks (for oneself and others). All one can do is try to apply a form of reflexive rationality that includes the volatility smile and its consequences for markets (Luhmann 1984, 640ff). According to this rationality, paper markets are not unreal, and their unpredictable operations are (often) not irrational at all. The task of sociological observation should be to highlight the reflexive aspects of risk and to analyse the corresponding form of rationality.

Bibliography Arnoldi, Jakob. 2004. ‘Derivatives: Virtual Values and Real Risks’. Theory, Culture & Society 21 (6): pp. 23–42. Beck, Ulrich. 1986. Risikogesellschaft: Auf dem Weg in eine andere Moderne. Frankfurt am Main: Suhrkamp. Beckert, Jens. 2016. Imagined Futures: Fictional Expectations and Capitalist Dynamics. Cambridge, MA: Harvard University Press. Beunza, Daniel, and David Stark. 2012. ‘From Dissonance to Resonance: Cognitive Interdependence in Quantitative Finance’. Economy and Society 41 (3): pp. 383–417. Black, Fischer, and Myron Scholes. 1973. ‘The Pricing of Options and Corporate Liabilities’. Journal of Political Economy 81 (3), pp. 637–54. Bronk, Richard. 2009. The Romantic Economist: Imagination in Economics. Cambridge: Cambridge University Press.

233

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures Bronk, Richard, and Wade Jacoby. 2016. ‘Uncertainty and the Dangers of Monocultures in Regulation, Analysis, and Practice’. MPIfG Discussion Paper, 16/6, Max Planck Institute for the Study of Societies, Cologne. Bryan, Dick, and Michael Rafferty. 2007. ‘Financial Derivatives and the Theory of Money’. Economy and Society 36 (1): pp. 134–58. Callon, Michel. 1998. ‘Introduction: The Embeddedness of Economic Markets in Economics’. In The Laws of the Market, edited by Michel Callon, pp. 1–57. Oxford: Blackwell. Callon, Michel, Yuval Millo, and Fabian Muniesa. 2007. Market Devices. Oxford: Blackwell. Demange, Gabrielle, and Guy Laroque. 2006. Finance and the Economics of Uncertainty. Oxford: Wiley-Blackwell. Esposito, Elena. 2007. Die Fiktion der wahrscheinlichen Realität. Frankfurt am Main: Suhrkamp. Esposito, Elena. 2011. ‘The Structures of Uncertainty: Performativity and Unpredictability in Economic Operations’. Economy and Society 42 (1): pp. 102–29. Esposito, Elena. 2013. ‘Economic Circularities and Second-Order Observation: The Reality of Ratings’. Sociologica 2. DOI: 10.2383/74851. Goodhart, Charles A.E. 1989. Money, Information and Uncertainty. 2nd edn. London: Macmillan. Hafner, Wolfgang, and Heinz Zimmermann. 2009. Vinzenz Bronzin’s Option Pricing Models: Exposition and Appraisal. Berlin: Springer. Henrich, Dieter, and Wolfgang Iser. 1983. Funktionen des Fiktiven. München: Fink. Keynes, John Maynard. 1936 [1973]. The General Theory of Employment, Interest and Money. London: Macmillan. Knight, Frank H. 1921 [1964]. Risk, Uncertainty and Profit. New York: Sentry. Koselleck, Reinhart. 1979. Vergangene Zukunft: Zur Semantik geschichtlicher Zeiten. Frankfurt am Main: Suhrkamp. Koyré, Alexandre. 1957. From the Closed World to the Infinite Universe. Baltimore, MD: Johns Hopkins Press. Lee, Benjamin. 2015. Risking Together (I) and (II). New York: Manuscript. Lo, Andrew W., and Craig MacKinlay. 1999. A Non-Random Walk down Wall Street. Princeton, NJ: Princeton University Press. Luhmann, Niklas. 1976. ‘The Future Cannot Begin: Temporal Structures in Modern Society’. Social Research 43 (1): pp. 130–52. Luhmann, Niklas. 1980. ‘Temporalisierung von Komplexität: Zur Semantik neuzeitlicher Zeitbegriffe’. In Gesellschaftsstruktur und Semantik. Studien zur Wissensoziologie der modernen Gesellschaft, Vol. 1, by Niklas Luhmann, pp. 235–300. Frankfurt am Main: Suhrkamp. Luhmann, Niklas. 1984. Soziale Systeme: Grundriß einer allgemeinen Theorie. Frankfurt am Main: Suhrkamp. Luhmann, Niklas. 1991. Soziologie des Risikos. Berlin: de Gruyter. Luhmann, Niklas. 1997. Die Gesellschaft der Gesellschaft. Frankfurt am Main: Suhrkamp. MacKenzie, Donald. 2006. An Engine, Not a Camera: How Financial Models Shape Markets. Cambridge, MA: MIT Press.

234

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Volatility Calculus and Indeterminacy MacKenzie, Donald. 2007. ‘Is Economics Performative? Option Theory and the Construction of Derivatives Markets’. In Do Economists Make Markets? On the Performativity of Economics, edited by Donald MacKenzie, Fabian Muniesa, and Lucia Siu, pp. 54–86. Princeton, NJ: Princeton University Press. MacKenzie, Donald. 2009. Material Markets: How Economic Agents Are Constructed. Oxford: Oxford University Press. MacKenzie, Donald, Fabian Muniesa, and Lucia Siu. 2007. Eds. Do Economists Make Markets? On the Performativity of Economics. Princeton, NJ: Princeton University Press. Malkiel, Burton G. 1999. A Random Walk down Wall Street. New York: Norton. Mandelbrot, Benoît, and Richard L. Hudson. 2004. The (Mis)behavior of Markets: A Fractal View of Risk, Ruin, and Reward. New York: Basic Books. Millman, Gregory J. 1995. The Vandals’ Crown: How Rebel Currency Traders Overthrew the World’s Central Banks. New York: Free Press. Rebonato, Riccardo. 2001. ‘Managing Model Risk’. In Mastering Risk: Vol. 2: Applications, edited by Carol Alexander, pp. 82–115. London: Financial Times/Prentice Hall. Rizzo, Mario J. 1979. Time, Uncertainty, and Disequilibrium: Exploration of Austrian Themes. Lexington, MA: D.C. Heath and Company. Schmidt, Christian. 1996. Uncertainty in Economic Thought. Cheltenham, UK: Edward Elgar. Shackle, George L.S. 1955 [1990]. Uncertainty in Economics and Other Reflections. Cambridge: Cambridge University Press. Shackle, George L.S. 1972. Epistemics and Economics: A Critique of Economic Doctrines. Cambridge: Cambridge University Press. Shackle, George L.S. 1979. Imagination and the Nature of Choice. Edinburgh: Edinburgh University Press. Shackle, George L.S. 1990. Time, Expectations and Uncertainty in Economics. Cheltenham, UK: Edward Elgar. Stark, David. 2009. The Sense of Dissonance: Accounts of Worth in Economic Life. Princeton, NJ: Princeton University Press. Stark, David. 2013. ‘Observing Finance as a Network of Observations’. Sociologica 2. doi: 10.2383/74854. Taleb, Nassim Nicholas. 2007. The Black Swan: The Impact of the Highly Improbable. New York: Random House.

235

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

11 Uncertain Meanings of Risk Calculative Practices and Risk Conceptions in Credit Rating Agencies Natalia Besedovsky

Introduction: Close Encounters with Two Different Conceptions of Risk ‘I’m not here to predict a crisis. If I could predict a crisis, I wouldn’t work at Moody’s, I would own Moody’s!’ (ES, sovereign analyst, Moody’s, April 2010)

ES laughs and tells his joke as if he had just come up with it. He is a middle-aged, large, dark-haired sovereign credit rating analyst at Moody’s, a credit rating agency (CRA) that rates the creditworthiness of more than 130 countries, 11 000 companies, 21 000 public finance issuers, and 76 000 structured finance obligations.1 It seems clear he wants to present himself as selfreflective about his (and his company’s) difficult role in the most severe financial crisis since the 1930s, which shook the financial markets to their foundations and made whole economies falter. But at the same time, he wants me to believe that he is comfortable and confident he is doing the right thing. His confidence is reflected in the setting of the interview: one of Moody’s many conference rooms on the 26th floor of 7, World Trade Center, at 250 Greenwich Street in Manhattan. This address is at the heart of Wall Street, with an impressive view over Manhattan, and directly next to the ghostly construction area of Ground Zero, where the new World Trade Center was about to be built (this was in 2010). 1

https://www.moodys.com/Pages/atc002.aspx (as of 15 July 2014).

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Meanings of Risk

With his joke about the limited predictive power of analysts, he explains his fundamental epistemological assumption concerning what a credit rating means: it is not—and cannot be—predictive, but represents only an opinion on the current state of a country, company, or security. Throughout the interview, ES does an impressive job of finding an elegant balance between acknowledging the criticisms directed towards credit rating agencies and trying not to be too defensive, and between underlining his humility as an analyst and insisting on the importance of ratings. On a different day, in Midtown, NYC, I am talking to a former Moody’s structured finance analyst, BC, who left Moody’s in the early 2000s to create her own company, rating structured securities with a secret database and algorithm. BC is a slender, understatedly but elegantly dressed woman, who laughs a lot. She, too, is very self-confident and relaxed, and instantly takes over the interview, asking about the sound quality of my recording device. She has a charming way of making me feel comfortable, with compliments for being a sociologist who wants to understand financial markets. In the course of the interview, she would name Levi-Strauss, quote Marx, and talk about the democratic potential of structured finance credit. Again, her personality and persuasive abilities are impressive. But, as a sceptical researcher, I pose a similar question to the one I asked ES: How much does BC believe in rating analysts’ ability to predict the future with precision? Her answer stands in stark contrast to ES’s, quoted at the beginning of this section: NB: And the idea is that this precision can be achieved? BC: It can be achieved. Of course it can be achieved. I mean. We can put men on the moon. There is nothing we can’t do. Both analysts’ demeanour is pretty similar: cultured, worldly, well-spoken, refusing to embody the stereotype of the greedy Wall Street trader. Despite their similarities in habitus, however, ES and BC belong to two very different worlds within the credit rating agency cosmos. ES can be characterized as a traditional rating analyst, while BC is a so-called ‘quant’. The ‘traditional’ rating culture can be described as bookish, nerdy, perhaps even ‘dusty’. It is sometimes compared to a Jesuit order: studious, self-referential, and not fully integrated or interested in Wall Street circles. Most of the analysts that belong to this culture are economists without a specific specialization in finance. There are even a few social scientists or analysts with a major in humanities. The quants, on the other hand, typically have a very different educational background: they are mathematicians, engineers, physicists, or computer scientists, or have an advanced degree in finance. They are ‘geeks’, with a strong inclination towards a ‘scientific’ self-representation. While the traditional rating analysts describe themselves as experts in a holistic sense—that 237

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

is, knowing about the specificities of financial markets and of the industries or countries they rate—the quants are experts in number crunching. Most importantly, however, they differ fundamentally in their conception of what risk is, how to measure it and therefore, what a rating represents. This stands in contradiction to what ratings claim to do and represent—what they claim as their unique selling point or their function in financial markets— namely, to provide an assessment of credit risk that renders investments that differ across industries and types of credit securities comparable through representation on a single, simple, and linear scale. This chapter takes a practice-theory perspective and combines it with insights from social studies of finance and risk management literature to show how calculative practices of risk assessment are inherently intertwined with the knowledge they produce. Through the decisions on methods, variables, weights, and, more broadly, calculative practices, risk is created in practice (Kalthoff 2005). It ceases to be an abstract idea and becomes specific and concrete. In other words, to understand the de facto meaning of an epistemic concept in a specific field, it is necessary to look at the calculative practices that define it. To this end, the chapter studies the calculative practices of rating analysts as social phenomena that shape their conceptions of risk. Based on the analysis of the rating practices of two different groups of analysts within their respective agencies2—the sovereign (country) rating and structured finance rating groups—this chapter identifies two different conceptions of risk that co-exist within CRAs: the diagnostic conception of sovereign rating and the technical conception of structured finance rating. Due to their different methods and epistemological assumptions, they differ crucially in their attitudes towards Knightean uncertainty. The aim of this chapter is not to judge or criticize the different methodologies or risk conceptions of rating analysts.3 Instead, the intention is to provide a detailed insight into these conceptions and highlight some crucial implications for a sociology of credit risk. The chapter argues that the risk conceptions are inherent in the methodological approaches used and are a precondition of the way CRAs deal with rating the creditworthiness of countries or structured securities, respectively. This is particularly consequential in the case of structured finance rating, because it is this specific ‘epistemic culture’ (Knorr Cetina 1999, 2007) that enables the creation of structured finance securities in the first place. Any critique of ratings or rating methods therefore needs to take 2 Interviews were conducted by the author with analysts and former analysts from the two largest CRAs, namely Moody’s and Standard & Poor’s. 3 For different assessments and explanations concerning their role in financial markets in general and in the financial crisis in particular, see, for example, Abdelal (2007), Besedovsky (2012), Carruthers (2013), MacKenzie (2011), Partnoy (1999, 2006), Rom (2009), Sinclair (2005), and White (2010).

238

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Meanings of Risk

into account the rating analysts’ underlying conceptions of risk, their acknowledgement (or neglect) of radical uncertainty, and their resulting claims about their ability to predict the future.

Risk in Sociological Literature There has been a special interest in the concept of risk in sociology since the 1980s, spurred by Ulrich Beck’s book Risk Society (Beck 1986). In addition to Beck, this interest can be seen in the works of Niklas Luhmann, Anthony Giddens, or scholars in the Foucauldian tradition.4 These authors have in common that they see risk as a central (sometimes even defining) concept of contemporary societies—what they call ‘modernity’. They contrast this ‘modern’ idea of risk with other, traditional notions that characterize pre-modern times, and usually try to define risk by juxtaposing it to a possible opposite, be it safety, danger, ‘fortuna’, or uncertainty (Bonß 2013; Castel 1991; Giddens 1990; Luhmann 2005). The broader definitions and dichotomies that appear in the sociological literature on risk lie outside the scope of this chapter. Of particular interest, however, is the question of calculability. Beck, Giddens, and Luhmann all agree that the concept of risk involves some kind of (belief in) calculability. Part of what makes risk a ‘modern’ concept is its ‘secular’ view of causality. At the same time, the three authors are—to different degrees—sceptical of the actual calculability of risks, and especially of the calculative capabilities of the so-called risk experts. Luhmann, for example, sees rational calculation of risk as a necessity for action, but stresses the limits of this calculability. One reason for this is that there is always insufficient knowledge about the future, not only in terms of (exogenous) dangers, but also concerning risks that are endogenous because they are consequences of the person’s decisions: ‘It is a question of decisions that serve to bind time, although we cannot gain sufficient knowledge of the future; indeed, not even of the future we generate by means of our own decisions’ (Luhmann 2005, 12–13). In addition, calculating risk is risky in itself, and the risk of calculation can only be seen by a secondorder observer of the person calculating the risk. Beck, Giddens, and Luhmann also have in common that they define risk as something negative that should be avoided, or, if this is not possible, reduced. By contrast, as argued here, some financial market participants’ conception of risk implies that it should be seen as a neutral, sometimes even positive, phenomenon.

4 For detailed overviews of the sociological literature on risk, see Lupton (2013), Power (2014), and Zinn (2008).

239

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

Despite the importance of risk in the economic sphere, the literature on risk alluded to so far in this chapter has had little impact on much recent economic sociology, which takes a different starting point for the discussion of risks, borrowing from Frank Knight’s (1921 [1964]) distinction between risk and uncertainty. Knight’s distinction is not primarily normative or attributive, but epistemological. He distinguishes between risk, which refers to instances in the future to which we can assign probabilities, and uncertainty, which refers to instances that are not calculable and therefore unforeseeable. Knight acknowledges that the difference between statistics and other non-statistical estimates or judgements is not always clear and that even statistical calculations face several potential problems. He describes, for instance, the problem of assuming homogeneity of cases in statistical analysis, and the limitations of comparability—a phenomenon now known as commensuration (Espeland and Stevens 1998). The uniqueness of every business decision, on the one hand, and the need to assume homogenous cases in statistical analysis, on the other hand, lead to practical problems for entrepreneurs using statistics: they have to make a judgement about whether they are in a situation that is sufficiently similar to those used in the statistical models that they should let the statistics guide their decisions, or whether they are in a different situation, where it does not make sense to do so (Knight 1921 [1964], 228f). Knight is generally sceptical of overreliance on statistics, and also makes a point of explaining that only in situations of uncertainty can entrepreneurs actually make profits. Drawing on Skidelsky (2009), Bronk (2011, 9) further distinguishes between epistemological uncertainty (‘where relevant probabilities are unknown’, for instance due to the complexity of the situation) and ontological uncertainty (‘where they are logically unknowable’). While innovations in risk-assessment practices can to some degree decrease epistemological uncertainty and turn it into calculable risk, ontological uncertainty means that it is impossible to know ‘even the categories and possible nature of what has yet to be created or yet to evolve’ (ibid.). Drawing on Knight’s distinction, Beckert (1996, 2016) develops a critique of neoclassical economics, based on what he calls its conflation of uncertainty and risk. He argues that economic actors are often confronted with situations where they cannot calculate risks, because they do not know the corresponding probabilities. Instead, they have to deal with instances of uncertainty that cannot be tackled with the tools and models of economics (Beckert 1996). To reduce this uncertainty, Beckert argues, actors need to rely on social constructions such as tradition, norms, institutions, social networks, organizational structures, or narratives when making decisions. These social mechanisms stand at the core of sociological analysis, and sociology is therefore, according to Beckert, well equipped to analyse them. Beckert therefore suggests setting the problem of uncertainty ‘as the vantage-point for economic sociology’ (Beckert 1996, 817). 240

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Meanings of Risk

While much work has been done in economic sociology in an effort to understand the institutions, networks, and cultural artefacts that help in dealing with uncertainty (Beckert and Aspers 2011; Fligstein and Dauter 2007; Lounsbury and Hirsch 2010; Smelser and Swedberg 2005), the concept of risk itself has often been underappreciated. However, in financial markets, financial actors in everyday practice usually (and, as shown here, increasingly) use the idea of risk (and not so much the concept of uncertainty) to describe possible futures. They are mainly not concerned with the social mechanisms that reduce uncertainty. Instead, their main goal is to increase the calculability of future events. Risk (and not uncertainty) has become a ‘powerful organizing category’ (Power 2014) of financial markets. Economics as a discipline—its overall logic and ideology, as well as its models—plays an important performative (Callon 1998) role in shaping the conceptions, models, and epistemological assumptions of financial market participants. The social sciences therefore have to go beyond criticizing economics and financial markets for overestimating the extent of the calculability of future events. They must also accept the influence of financial models and engage empirically with risk as one of the most central categories of knowledge in financial markets—as defined by financial economics and financial market actors. The role of financial sociology, in particular, is to study the practitioners’ concepts of risk and the practices of calculation that they use to ‘transform’ (perceived) uncertainty into (perceived) risk. In other words, for a deeper understanding of financial markets, we need to take the notion of risk seriously and study it as a central category of knowledge and practice in financial markets. From this perspective, the category of risk is not reserved to what Frank Knight defined as risk—that is, the measurement of ‘ergodic’ regularities of behaviour. If the operation of risk assessments in financial markets is studied as a social practice, it shows instead that many of these actors treat all situations as calculable forms of risk and thus leave no room for the category of uncertainty in Knight’s sense.

The Co-Constitution of Calculative Practices and Conceptions of Risk The representation and calculation of risk, and thus the meaning of risk itself, is the result of contingent practices.5 As Power (2014) concisely puts it, risk has a ‘thoroughly hybrid and multi-logic nature’. Literature on risk management has argued that among risk managers there are fundamental differences 5

For a full discussion of practice theory, see Reckwitz (2002) and Knorr Cetina et al. (2001).

241

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

concerning what is conceived as risk, how to measure it, and the degree of trust in quantitative calculative models of risk assessment (Mikes 2009, 2011; Power 2007)—with changes over time and in different organizational settings (Poon 2009; Power 2014; Scott and Perry 2012). The meaning of risk is not fixed. The different conceptions of risk contain the imprints of practices, norms, ideas, and institutions. Michael Power calls this an apparatus of risk (Power 2012), a system of thought embedded in a specific culture. Beyond the discursive level, the practices of risk assessment and evaluation are equally important for an understanding of the meaning of risk. Knorr Cetina describes these practices of knowledge production as ‘epistemic cultures’. First used primarily for differentiating scientific fields (Knorr Cetina 1999), she expands this concept to include expert groups in finance (Knorr Cetina 2007), defining epistemic cultures as ‘cultures of creating and warranting knowledge’, comprising ‘sets of practices, arrangements and mechanisms bound together by necessity, affinity and historical coincidence that, in a given area of professional expertise, make up how we know what we know’ (Knorr Cetina 2007, 363). Knowledge and practice, in this understanding, are ‘reciprocally constitutive, so that it does not make sense to talk about either knowledge or practice without each other’ (Orlikowski 2002, 250). They cannot be seen as separate, but are mutually constituting and co-evolving. Abstract ideas only become ‘real’ (as in ‘realized’) in practice. Calculative practices create the objects they measure and are at the same time the concrete manifestations of the abstract concepts they represent. Herbert Kalthoff uses the concept of ‘revealing’: ‘as a form of “revealing”, calculation is a central part of the ordering system, and it constitutes the objects it calculates in the sense that it fixes their existence: objects are revealed as objects that are calculable’ (Kalthoff 2005, 73). Calculations highlight the calculated aspects of a phenomenon and turn these aspects into its defining characteristics. For instance, saying that a country is rated AAA does more than present this country as creditworthy compared with, say, BB-rated countries; it also marks it as a specific kind of market participant in the sovereign bond market. The most important implication of this perspective is that those who perform the calculative practices are the ones who decide upon the de facto meaning of what they are calculating, measuring, or evaluating. This does not imply that those who measure a concept, even when they have a monopoly in measuring it, have the absolute and definitive power to define the concept. As the evaluation or assessment devices are used by other actors, they can be appropriated and used for different purposes, which then might change the meaning in those contexts (because using them is also a practice). But the actual definitional power of what the devices really mean lies in the selection of the criteria used to produce the knowledge claims these devices represent. To understand the de facto meaning of the 242

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Meanings of Risk

abstract idea, one thus has to understand the calculative practices that ‘realize’ these ideas in practice.

Risk as a Social Category—Defined by Those Who Measure It What does risk mean for financial actors? What are the practices of assessing risk? And what are the (normative and practical) implications and consequences of their ideas of risk? To better understand the different conceptions of risk prevailing in financial markets, this chapter examines, in particular, the credit risk rating practices of two major CRAs, Moody’s and Standard & Poor’s. The analysis presented relies on thirty hours of interviews with CRA analysts and other experts in credit rating. To complement the interviews and triangulate the research, a variety of documents from an extended array of methodological publications of CRAs, as well as textbooks on structured finance ratings have been included in the analysis.6 The remainder of the chapter demonstrates that methodological differences can be traced back to (and help reinforce) fundamentally different concepts of credit risk and the corresponding normative implications that co-exist in rating agencies.7 Ideally, credit ratings should decontextualize from the region, the issuer, the industry, the kind of product, and even from time. In other words, through the rating within a single rating scale, a Greek sovereign bond and a mortgagebacked security are made comparable with each other. Through a rating, even the most complicated structured finance products can be ‘domesticated’ (Carruthers 2010, 164). In practice, however, the comparability of ratings is a huge problem for rating agencies. As one analyst states: ES: One of the most difficult things we have is, we try to say ‘a triple A’s a triple A’s a triple A’. Meaning: If I give you a triple A in the US or for sovereign, then it should be the same risk as a triple A corporate, as a triple A anything else. It’s very difficult. These difficulties stem, on the one hand, from the general problem of rendering commensurable (Espeland and Stevens 1998) such disparate phenomena as mortgage-backed securities, bonds, companies, banks, or countries. But there 6 A more detailed methodological discussion is provided in the author’s dissertation thesis (Besedovsky 2015). 7 The two conceptions of risk described here are derived from interviews with rating analysts and CRA publications and do not represent the conceptions of investors, policy-makers, or any other actors that use the ratings. The intent is not to find a general conception of risk or give a new definition, but to highlight the specificities of credit risk rating practice and their implications.

243

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

is also a second problem for rating agencies: the ways of thinking—the ontological and epistemological assumptions—and the rating practices differ substantially between analysts of different sectors. For example, while sovereign (country) and corporate ratings tend to use a traditional approach, structured finance rating analysts operate with stochastic methods from financial mathematics. These fundamental differences of methods and meanings of ratings correspond to two fundamentally different conceptions of risk: the diagnostic and the technical conception of risk.

The Diagnostic Risk Conception of Sovereign Rating Practices Sovereign rating practices can be described as holistic. Next to fundamental analyses of basic economic data, the everyday practices of sovereign analysts, who are usually assigned about seven countries, consist of a variety of tasks: following political and economic discourse in news outlets, establishing and maintaining contacts with informants in the countries concerned, and visiting the countries at least once a year for several days, where they talk with different actors in the government (mostly in ministries of finance), with the opposition, as well as local academics and other experts. It could be considered similar to what Knorr Cetina (2011) describes as financial ‘miniethnographies’. For a rating decision, these insights are combined with quantitative, but mostly descriptive, analyses of macroeconomic data, democracy indices, and other indicators. It is therefore a case-centred approach that takes into account the heterogeneity of cases. Its main goal is to try to understand and make explicit the causal relations between aspects of the country and its creditworthiness. One of their main challenges is to guarantee that the ratings of different regions are comparable to each other, for which they have a separate cross-region expert group.8 Due to their holistic approach and their relational thinking, for sovereign analysts, a rating is a relative judgement and an ordinal ranking of how safe and stable countries are with regard to paying their debts. Although risk is an assessment relevant to possible events in the future, the assessment of countries’ creditworthiness is seen primarily as a current opinion on their status in the present. In fact, there is no particular emphasis on probabilistic thinking in sovereign rating and so the ratings are not supposed to represent precise probabilities of default. Indeed, some interviewees suggested that their ratings do not represent probabilities at all. Alluding to so-called default studies, where actual defaults in each rating category are calculated, some analysts

8

244

For a detailed account of the rating methods, see Besedovsky (2018).

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Meanings of Risk

argued that it was only possible to know ex post—and only on an aggregate level—whether the ratings were reasonably accurate.9 A rating of creditworthiness is generally considered a measure of stability, safety, or ‘health’. This is expressed in statements that the most important overall factor for a sovereign rating is the country’s shock absorbing capacity (Moody’s Investor Service 2008; interview data). The underlying idea is that stability and strength increase resilience with regard to economic shocks, or other shocks affecting the economy, such as natural disasters or political crises. This is also reflected in the use of the word ‘vulnerability’, which sovereign analysts often used in the interviews. A highly rated country is considered robust, ‘healthy’, and safe. A low-rated country is considered weak, ‘unhealthy’, and unsafe. This conception is close to what Castel (1991) calls ‘dangerousness’. In this conception, risk clearly has a normative connotation: it is something negative—a problem, weakness, or a danger. This corresponds to the use of risk in non-financial contexts, where risk is usually used to refer to hazard, danger, or harm (Lupton 2013). Second, credit risk is attached to a specific entity. Risk represents almost a ‘character trait’ of the country, similar to what Lupton, referring to individuals, describes as ‘an immanent quality of the subject’ (Lupton 2013, 124). High levels of risk in sovereign rating are therefore closely related to the idea of deviance (Castel 1991), as something that is not normal, should not occur, and is in consequence associated with blame. The negative connotation of risk can be seen in the comments that are published together with rating changes, where they often hint at possible ways for countries to get a better rating. For sovereigns, a rating represents the creditworthiness of a country in an almost literal sense. The basic question is: ‘Is the country worthy of being trusted with credit?’ The rating then represents the ‘school grade’ of the character trait ‘creditworthiness’. With regard to the epistemological dimension, the diagnostic conception assumes that risk is something that cannot be fully calculated, but rather comes with some inherent level of uncertainty. This is because every country is, to a certain extent, regarded as singular and unique. The epistemological belief that sovereign credit risk is not fully calculable encourages a holistic, expert-driven, and experiential approach—and vice versa.

The Technical Risk Conception of Structured Finance Rating Practices Structured finance products such as mortgage-backed securities (MBSs) or collateralized debt obligations (CDOs) can be described as among the most 9

Additionally, because there are relatively few sovereign defaults, even these statistics are treated with a certain caution. One analyst explained that even in the lowest non-default category of ratings, probably about seventy per cent of countries have never actually defaulted.

245

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

influential innovations of financial markets in the twentieth century. CRAs played a constitutive role in creating the market for structured products from the outset by developing standards of credit analysis, and thereby constituting the pricing norms of structured securities. The basic principle of a structured security is to pool a large number of assets (for example, mortgage loans from an originating bank) into a portfolio. This pooled security then gets tranched, that is, structured according to the priority of claims of principal and interest against the pool. Accordingly, different ratings are given to different tranches of the portfolio. The junior/mezzanine and equity tranches serve as a protection for the senior tranches, because senior tranches are prioritized claims, which usually get a triple-A rating.10 The rating agencies’ job is to calculate the amount of protection needed for the senior tranche. In this way, CRAs determine the structure of the security concerned. They not only perform calculations, but are in essence the architects of the security.11 Due to this particular structure, the structured finance ratings logic differs from the sovereign logic in several ways. Structured finance rating analysts use quantitative historical data from the structured security’s underlying assets (for example, mortgages, car loans, corporate bonds) and create mathematical models that produce probabilistic statistics. In contrast to sovereign rating analysts, who focus on understanding causal relationships between different economic and political aspects of countries and their creditworthiness, structured finance analysis is much more inductive, often based on data mining techniques, and concerned with populations and correlations. It also tends to treat the underlying assets as homogenous (for example, all mortgages have a FICO-score and are therefore comparable to each other), abstracting from individual differences. Structured finance ratings, as opposed to sovereign ratings, therefore have an (implicit) default probability and are understood as precise cardinal measures of credit risk. Moreover, they not only represent the probability of default, but first and foremost the ‘loss severity’ (S&P), or ‘loss given default’ (Moody’s). The concept is defined as ‘the product of the likelihood of observing a default and the severity of the loss conditional on default’ (Coval et al. 2009, 8). This means that they take into consideration the severity of the losses once a default occurs for every tranche. The shift from defining structured finance ratings purely as default probabilities to including loss severity in the calculations can be traced back to innovations in structured ratings at Moody’s during the 1990s (Raynes and Rutledge 2003).

10

For a detailed and comprehensive description of structured securities, see MacKenzie (2011). Based on this principle, there have subsequently been a multitude of variations and innovative alterations in structured finance with increasing complexity, such as multiple securitizations or synthetic financial products. 11

246

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Meanings of Risk

This focus on loss given default is perhaps the most important innovation of structured finance rating and one of the major differences compared with sovereign ratings.12 Crucially, it implies that a rating can be understood as a measure of risk-adjusted yield. In sovereign and corporate ratings, the aspect of ‘relative quality’ and default frequency remains the main focus, while— since technically, structured finance products do not default—a structured finance rating is a representation of the estimated rate of return. In structured finance, the meaning of ratings has therefore changed from ‘relative quality (quality or safety ratings) and payment certainty (default frequency) to riskadjusted yield’ (Raynes and Rutledge 2003, 29–31). This is reflected in the rationale of the investor: ES: Yes, they would call it the equity tranche or the mezzanine tranche, and you know of course, there’s, the returns would be higher here. But also they might say ‘hey, five percent loss, fine, you know, we’ll charge an extra amount of money, we expect to make money somewhere else. We can deal with it.’ You know. [ . . . ] It’s—even if you lose by—it’s all mathematical. It will be ‘how much money can I expect to make?’ (emphasis added). This risk-adjusted yield definition of rating suggests that it is neither necessary nor possible to minimize risk by being prudent. The only concern is with reducing the variance of risk, not the risk itself (Wigan 2009), which is done through increasing the quantity of the underlying assets, calculating the cushion you need, and diversification and hedging. The rating therefore is a calculation of the ‘right’ amount of cushion for an investor to buy a security. The concept of risk is just a variable that has to be calculated in order to find the right compensation for an investment. The contrast with the sovereign rating perspective is made evident by the lack of interest of the structured finance interviewees in discussing any ‘solutions’ to risky securities. While sovereign analysts often explained what a country would have to change to become more creditworthy, in structured finance the idea is simply to compensate risk with higher yield. BC: So, that is what I mean, that structured finance is the, because it’s based upon aggregates, and . . . because it’s statistical, we can measure actual performance against benchmarks . . . And once we do that we can determine what is the fair rate of return for the company versus the fair rate of return for the lender to the company. That’s it. In other words, in structured finance ‘credit ratings can intuitively be thought of as a measure of a security’s expected cash flow’ (Coval et al. 2009, 8). In 12 For a more detailed analysis and a historical description of this innovation, see Besedovsky (2018).

247

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

simple terms, the main question is: ‘How much money will I get back for the tranche I bought, given the average and distribution of default risk and loss severity in my portfolio and the structure of the security?’ Structured finance ratings are therefore not an assessment of creditworthiness in the diagnostic sense of sovereign rating. This difference is crucial, because it leads to a fundamentally different conception of risk that differs in several dimensions from the risk concept of sovereign rating. First, the negative connotation of risk is mostly abandoned. Instead, risk is seen as a necessary precondition and defining characteristic of financial markets because it represents an opportunity to make a profit. Secondly, in terms of attribution, risk is not attached to specific entities, but is an abstract variable of the likelihood of future events. This concept of risk ‘dissolve[s] the notion of a subject or a concrete individual, and put[s] in its place a set of factors, the factors of risk’ (Castel 1991, 281). It therefore exempts the individual from responsibility for the consequences of decisions based on risk calculation. Since they rely on probabilistic thinking, they ‘appear to be able to guarantee that even if things do go wrong, one can have acted correctly. . . . They immunize decision making against failure’ (Luhmann 2005, 13). Finally, the structured finance conception of risk also has the implication that nothing is risk free. There is no conception of ‘the normal’ and ‘the (risky) deviant’—risk is just about probabilities.13 The technical conception of risk is based on the assumption that it can be calculated. This perspective assumes that actors (or organizations) are able to calculate and manage uncertainties in a ‘rationalized’ (Power 2007, 23) way. Ideas might differ concerning how well the current tools, models, and datasets calculate risks, but the idea is that in theory a transformation from uncertainty into risk is possible. In other words, a fundamental assumption of structured finance rating models is that there is no ontological uncertainty (that is, the sort of indeterminacy that can never be amenable to calculation). Indeed, the choice of stochastic models as methods of producing ratings in structured finance leaves no other option but to exclude radical uncertainty from the risk assessment. In one respect, though, certain traces of Knight’s distinction between risk and uncertainty persist in structured finance given its antipathy to the incalculable. The ‘danger’ to be avoided is the presence of any factor that cannot be calculated; and, consequently, structured finance analysts try to grasp this epistemic aspect of Knightian uncertainty and find a way to transform it into a variable in the model.

13

248

For a similar argument in the case of fraud risk, see Power (2012).

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Meanings of Risk

The broad assumption that there is no risk-free endeavour implies that, instead of trying to avoid risks, they can, and should, be actively managed. Risk should be managed and exploited as an opportunity for profit. If you have the right calculative tools to measure risk, then there are financial strategies that allow you to maximize your profit and avoid negative returns (for example, through hedging). Risk therefore is only a problem if it is badly managed. In consequence, and contrary to the diagnostic logic, a distinction between a safe and a risky investment does not make sense, because any potential loss is already compensated by the calculation of the riskadjusted yield. Table 11.1 summarizes in stark terms the nature of the contrast between the diagnostic and technical conceptions of risk manifested in the divergent credit rating practices of the sovereign (country) and structured finance analysts discussed in this chapter. However, the contrast should not be overstated because as far back as the 1980s, and particularly since the 1990s, CRAs have faced increasing pressures to make all their ratings cardinal, more precise, and more ‘predictive’. The need for precision is particularly important in structured finance markets. Some of these markets rely completely on the belief that ratings are precise predictions of defaults and the corresponding losses— for example, the market for CDOs. The credit ratings of the underlying assets are used as main input for structuring the CDO. In this case, ratings are necessarily assumed to have cardinal meaning and represent the statistical probability of default and loss given default. But even corporate and sovereign analysts are increasingly pressured (and willing) to include financial market information—for example, so-called ‘market sentiment’—in their rating practices, especially after criticism for their role in the European sovereign debt Table 11.1. Sovereign vs structured finance rating conceptions of risk Dimensions

Sovereign (diagnostic)

Structured finance (technical)

Meaning of rating

Safe vs unsafe entity/investment

Right amount of compensation for an investment: riskadjusted yield

Nature of risk

Negative, weakness, hazard

Neutral category in principle; implicit positive connotation: opportunity to make profit

Attribution

Risk as a character trait of an entity

Probability, everything has risks: attribution only for purposes of accountability and transparency

Epistemological: relation to Knight’s uncertainty

Risk not completely knowable, uncertainty is part of a rating

Calculable, transformation from uncertainty into risk is possible

Epistemological: adequate methodology

Holistic perspective, experience, expertise, logical inference

Historical data of underlying assets, probabilistic inference

Normative implications

Reduce, prudence

Manage, hedge, exploit

249

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

crisis.14 This fundamental shift towards the technical conception is also observable in other contexts (see, for instance, Poon 2009; Wigan 2009).

The Technical Risk Conception and the Financial Crisis The logic of the technical approach to risk is made clear by reactions to the crisis observed within CRAs. Two examples from the interviews show how radical the technical approach is with regard to Knightean uncertainty. The first is related to the political discourse on CRAs after the financial crisis of 2007/2008. In the aftermath of the crisis, several investigations by regulators have taken place in the EU and the United States. The US Congress has conducted several congressional hearings explicitly concerned with credit rating agencies. The stark contrast between the two conceptions of risk outlined in this chapter is evident in these hearings. For instance, in his introduction to a Congressional Hearing in 2011, Senator Carl Levin stated: For a hundred years, Main Street investors trusted U.S. credit rating agencies to guide them toward safe investments. Even sophisticated investors, like pension funds, municipalities, insurance companies, and university endowments, have relied on credit ratings to protect them from Wall Street excesses and distinguish between safe and risky investments. But now that trust has been broken. (US Senate 2011, 289; emphasis added)

When the structured finance analyst quoted in the introduction was confronted with this distinction that regulators expected to be made, her answer was straightforward: BC: Well, it’s a very interesting problem because the financial regulator is also conservative, so they also want to—their thinking is much more normative. They want to reward good companies and punish bad companies. . . . [W]hat they [the regulators] mean by good and bad is certain versus uncertain, but risk-adjusted returns factor that level of uncertainty into the result. NB: OK, so actually, if the ratings were a perfectly right [measurement of credit risk] then there would be no difference between safe and unsafe, because we would have the perfect compensation for—? BC: That’s exactly right. You see, that’s brilliant, that is exactly right. The market is supposed to neutralize the risk. She explains here how even uncertainty (when there is not enough information, for instance, with a new class of financial products) can be modelled as 14

250

For a detailed analysis of this phenomenon, see Besedovsky (2018).

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Meanings of Risk

one dimension of risk and ‘factored’ into the risk calculation by increasing the risk-adjusted returns. In other words: even ‘not knowing about something’ can—in the epistemic culture of structured finance—be translated into a precise number and therefore transformed into risk. Her criticism of the regulators’ normative perspective and her belief in calculability, prediction, and the neutralization of risk are a poignant representation of the technical approach. The second example relates to the rating analysts’ explanations for the financial crisis. Carruthers (2013) uses the Knightean distinction for his assessment of CRAs’ role in the crisis. He argues that while CRAs promised to eliminate uncertainty via ratings, ‘various problematic uncertainties lurked beneath the surface’ (Carruthers 2013, 542). For Carruthers, ‘even when true uncertainty is made to look like risk, it nevertheless remains uncertainty’ (Carruthers 2013, 543). His analysis suggests that the financial crisis ‘uncovered’ CRAs’ overconfidence in their ability to transform uncertainty into risk and was thus a return to uncertainty. Just as the two kinds of rating analysts analysed in this chapter differ in their conception of risk, they have different things to say about the CRAs’ role in the financial crisis. Most sovereign rating analysts in the interviews would agree with Carruthers, as they explained the CRAs’ role in the crisis as having put too much trust in quantitative models, taking up the perspective of quantitative scepticism (Mikes 2011). In contrast, some of the opinions of the interviewed structured finance analysts can be described as quantitative enthusiasm (Mikes 2011). Their explanation of the CRAs’ role in the crisis addresses neither the structured finance methods per se nor any kind of overconfidence in the ability to transform uncertainty into risk. Instead, they argue that the specific models were wrong, that certain variables or correlations were forgotten or neglected, and that they were not quantitative enough. TS, for instance, a former structured finance analyst with an astrophysics background who had left one of the big CRAs prior to the crisis, criticized the agencies for having ‘50 different ways’ to rate deals. He argued that this was a sign that the rating process was not a fully serious endeavour. For him, the failure of CRAs could be explained by their lack of knowledge of what he called—in an allusion to physics—the fundamental laws of structured finance. These two examples show how unapologetically structured finance rating analysts defend the idea that the risk of structured securities can be perfectly calculated through the models of structured finance. During the financial crisis, academics, policy-makers, and the media heavily criticized CRAs and blamed them for playing a crucial role in the development of the crisis. In the first years after the crisis, they focused on structured finance ratings, and not on sovereign rating methods. There seemed to be a period of time when there were calls to go back to the diagnostic approach of sovereign 251

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

rating. Just a couple of years later, however, the criticism shifted from the CRAs’ role in the mortgage crisis (and the structured finance methods) towards the ‘bias’ of the expert-driven sovereign rating methodology in the euro crisis. It seems that the trend has again turned in favour of greater standardization, quantification, striving for mathematical precision, and general orientation towards risk-adjusted yield measurements, not least due to reforms in the regulations of CRAs in the United States and the EU.15

Conclusion In his article ‘Capitalist Dynamics’, Beckert (2014) states that ‘capitalism’s extraordinary power—and at the same time its Achilles heel—is its ability to motivate actors to take risks despite the uncertainty of achieving desired outcomes and the likeliness of disappointment’ (Beckert 2014, 10). This chapter gives an example from finance of how this is done, and analyses a central mechanism that creates this ability: calculative practices of risk assessment. Financial markets are epistemic systems and ‘knowledge’ plays a central role in their functioning (Knorr Cetina 2007). But more importantly, calculative practices shape ideas, and the calculative devices are concrete manifestation of these ideas. In the decisions on which aspects of a phenomenon are measured, which ones are left out, and what is compared, and in other specific calculative manipulations, the concrete object of credit ‘risk’ is realized in practice. The calculative practices of rating production and the specific conceptions of risk are co-constitutive and co-evolve. In the case of CRAs’ risk assessments, the category ‘risk’ does not have a uniform meaning; rather, different conceptions of risk co-exist and compete within one organization. The de facto meanings of risk differ depending on the calculative practices and the epistemic culture in which they are embedded. Hence, contrary to what the one-dimensional rating scale suggests, the intended meaning of a rating depends on the way it is produced. Despite the CRAs’ main selling point—comparability—there are fundamental differences in what a rating represents. The two epistemic cultures identified within CRAs correspond to two entirely different conceptions of risk: the diagnostic conception of sovereign rating and the technical conception of structured finance rating. For sovereign analysts, uncertainty is part of risk, and therefore assessments such as ratings are not precise predictions of risk but relative ordinal rankings. Structured finance analysts, by contrast, assume that any kind of uncertainty can be factored into probabilistic risk models. 15

252

For a critical assessment of this trend, see Bronk and Jacoby (2016).

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Meanings of Risk

The structured finance practices described here highlight a very specific way in which financial market actors deal with Knight’s distinction between risk and uncertainty: they make it disappear. While radical uncertainty may be an important factor in structured finance markets, it has no place in the practices of structured finance rating practitioners. Indeed, they rely on the assumption that ontological indeterminacy does not exist. Structured finance depends on this premise and could not operate without it. If rating analysts did not believe (or did not act ‘as if’ they believe) in the absence of uncertainty, they could not perform their functions. Structured securities such as MBSs and CDOs, and especially their more sophisticated variations (such as synthetic CDOs), can only be realized by models that assume that perfect credit risk calculation is possible. This perspective is central to their business model. CRAs calculate risks for other financial actors, and they are one of the many financial service providers that, while calculating risks, make profits from fees rather than by actually taking the risk of investing in the rated products themselves. In fact, even the potential reputational risks CRAs might face for misevaluating a large number of securities as a result of the epistemic assumptions they make does not seem to affect them too much, as shown in their relatively quick recovery from their losses during the financial crisis.16 The conclusion that is possible based on the analysis of this chapter is quite different from a critique of rating failure or mistakes. The idea that any kind of uncertainty can, with the right models, be transformed into calculable risk is a necessity and a precondition for the logic of structured finance. As long as financial markets work based on this logic, ‘knowledge’ about structured finance products will be based on this assumption. Indeed, it could be argued that the belief in precise calculability of risk is one of the most powerful and influential imaginaries of financial markets. There are indications that this might be a long-lasting and general trend. There is a need, however, for further investigations into how these conceptions of risk are represented among those financial actors who use credit ratings—for instance to structure their portfolios, or, within regulations using ratings to regulate banks and other financial entities. The theoretical argument developed in this chapter should hopefully encourage further studies to pay more attention to the co-constitution and co-evolution of calculative practices and meanings. Calculative practices—‘little’, mundane, and seemingly purely technical things (MacKenzie 2009)—can shape ideas and should be regarded as central mechanisms that potentially contribute to shaping the current economic system. 16 For a historical account of how little their business success was tied to evidence for their ratings’ accuracy, see Carruthers (2013).

253

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

Bibliography Abdelal, Rawi. 2007. ‘A Common Language of Risk’. In Capital Rules, by Rawi Abdelal, pp. 163–281. Cambridge, MA: Harvard University Press. Beck, Ulrich. 1986. Risikogesellschaft: Auf dem Weg in eine andere Moderne. Frankfurt am Main: Suhrkamp. Beckert, Jens. 1996. ‘What Is Sociological about Economic Sociology? Uncertainty and the Embeddedness of Economic Action’. Theory and Society 25 (6): pp. 803–40. Beckert, Jens. 2014. ‘Capitalist Dynamics: Fictional Expectations and the Openness of the Future’. MPIFG Discussion Paper 14/7, Max Planck Institute for the Study of Societies, Cologne. Beckert, Jens. 2016. Imagined Futures: Fictional Expectations and Capitalist Dynamics. Cambridge, MA: Harvard University Press. Beckert, Jens, and Patrick Aspers. 2011. Eds. The Worth of Goods: Valuation and Pricing in the Economy. New York: Oxford University Press. Besedovsky, Natalia. 2012. ‘Politischer Ritterschlag für Ratingagenturen: Regulatorisches Outsourcing und der Beitrag von Gesetzgebern zur Macht der Ratingagenturen’. In Entfesselte Finanzmärkte: Soziologische Analysen des modernen Kapitalismus, edited by Klaus Kraemer and Sebastian Nessel, pp. 225–42. Frankfurt am Main: Campus. Besedovsky, Natalia. 2015. ‘Calculating Financialization: Credit Rating Agencies, Conceptions of Risk, and Calculative Practices as Drivers of Social Change’. PhD Dissertation, Humboldt University of Berlin. Besedovsky, Natalia. 2018. ‘Financialization as Calculative Practice: The Rise of Structured Finance and the Cultural and Calculative Transformation of Credit Rating Agencies’. Socio-Economic Review 16 (1): pp. 61–84. Bonß, Wolfgang. 2013. ‘Risk. Dealing with Uncertainty in Modern Times’. Social Change Review 11 (1): pp. 5–36. Bronk, Richard. 2011. ‘Uncertainty, Modelling Monocultures and the Financial Crisis’. Business Economist 42 (2): pp. 5–18. Bronk, Richard, and Wade Jacoby. 2016. ‘Uncertainty and the Dangers of Monocultures in Regulation, Analysis, and Practice’. MPIfG Discussion Paper 16/6, Max Planck Institute for the Study of Societies, Cologne. Callon, Michel. Ed. 1998. The Laws of the Markets. Oxford: Blackwell. Carruthers, Bruce G. 2010. ‘Knowledge and Liquidity: Institutional and Cognitive Foundations of the Subprime Crisis’. In Markets on Trial: The Economic Sociology of the US Financial Crisis, edited by Michael Lounsbury and Paul Morris Hirsch, pp. 157–82. Bingley, UK: Emerald Group Publishing. Carruthers, Bruce G. 2013. ‘From Uncertainty toward Risk: The Case of Credit Ratings’. Socio-Economic Review 11 (3): pp. 525–51. Castel, Robert. 1991. ‘From Dangerousness to Risk’. In The Foucault Effect. Studies in Governmentality: With Two Lectures by and an Interview with Michel Foucault, edited by Graham Burchell, Colin Gordon, and Peter Miller, pp. 281–98. Chicago, IL: University of Chicago Press.

254

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Meanings of Risk Coval, Joshua, Jakub Jurek, and Erik Stafford. 2009. ‘The Economics of Structured Finance’. Journal of Economic Perspectives 23 (1): pp. 3–25. Espeland, Wendy Nelson, and Mitchell L. Stevens. 1998. ‘Commensuration as a Social Process’. Annual Review of Sociology 24: pp. 313–43. Fligstein, Neil, and Luke Dauter. 2007. ‘The Sociology of Markets’. Annual Review of Sociology 33: pp. 105–28. Giddens, Anthony. 1990. The Consequences of Modernity. Stanford, CA: Stanford University Press. Kalthoff, Herbert. 2005. ‘Practices of Calculation: Economic Representations and Risk Management’. Theory, Culture & Society 22 (2): pp. 69–97. Knight, Frank H. 1921 [1964]: Risk, Uncertainty and Profit. New York: Sentry Press. Knorr Cetina, Karin. 1999. Epistemic Cultures: How the Sciences Make Knowledge. Cambridge, MA: Harvard University Press. Knorr Cetina, Karin. 2007. ‘Culture in Global Knowledge Societies: Knowledge Cultures and Epistemic Cultures’. Interdisciplinary Science Reviews 32 (4): pp. 361–75. Knorr Cetina, Karin. 2011. ‘Financial Analysis: Epistemic Profile of an Evaluative Science’. In: Social Knowledge in the Making, edited by Charles Camic, Neil Gross, and Michèle Lamont, pp. 405–42. Chicago, IL: University of Chicago Press. Knorr Cetina, Karin, Theodore R. Schatzki, and Eike von Savigny. 2001. The Practice Turn in Contemporary Theory. London: Routledge. Lounsbury, Michael, and Paul Morris Hirsch. 2010. Markets on Trial: The Economic Sociology of the US Financial Crisis. Bingley, UK: Emerald Group Publishing. Luhmann, Niklas. 2005. Risk: A Sociological Theory. New Brunswick, NJ: Aldine Transaction. Lupton, Deborah. 2013. Risk. 2nd edn. London: Routledge. MacKenzie, Donald. 2009. Material Markets: How Economic Agents are Constructed. Oxford: Oxford University Press. MacKenzie, Donald. 2011. ‘The Credit Crisis as a Problem in the Sociology of Knowledge’. American Journal of Sociology 116 (6): pp. 1778–841. Mikes, Anette. 2009. ‘Risk Management and Calculative Cultures’. Management Accounting Research 20 (1): pp. 18–40. Mikes, Anette. 2011. ‘From Counting Risk to Making Risk Count: Boundary-Work in Risk Management’. Accounting, Organizations and Society 36 (4): pp. 226–45. Moody’s Investor Service. 2008. ‘Sovereign Bond Ratings’. September 2008. Orlikowski, Wanda J. 2002. ‘Knowing in Practice: Enacting a Collective Capability in Distributed Organizing’. Organization Science 13 (3): pp. 249–73. Partnoy, Frank. 1999. ‘The Siskel and Ebert of Financial Markets?: Two Thumbs Down for the Credit Rating Agencies’. Washington University Law Quarterly 77 (3): pp. 619–712. Partnoy, Frank. 2006. ‘How and Why Credit Rating Agencies Are Not Like Other Gatekeepers’. In Financial Gatekeepers: Can they Protect Investors?, edited by Yasuyuki Fuchita and Robert E. Litan, pp. 59–102. Tokyo: Nomura Institute of Capital and Washington, D.C.: Brookings Institution Press. Poon, Martha. 2009. ‘From New Deal Institutions to Capital Markets: Commercial Consumer Risk Scores and the Making of Subprime Mortgage Finance’. Accounting, Organizations and Society 34 (5): pp. 654–74.

255

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures Power, Michael. 2007. Organized Uncertainty. Oxford: Oxford University Press. Power, Michael. 2012. ‘The Apparatus of Fraud Risk’. Accounting, Organizations and Society 38 (6–7): pp. 525–43. Power, Michael. 2014. ‘Risk, Social Theories, and Organizations’. In Oxford Handbook of Sociology, Social Theory and Organization Studies: Contemporary Currents, edited by Paul S. Adler, Paul Du Gay, Glenn Morgan, and Mike Reed, pp. 370–92. Oxford: Oxford University Press. Raynes, Sylvain, and Ann Rutledge. 2003. The Analysis of Structured Securities: Precise Risk Measurement and Capital Allocation. Oxford: Oxford University Press. Reckwitz, Andreas. 2002. ‘Toward a Theory of Social Practices: A Development in Culturalist Theorizing’. European Journal of Social Theory 5 (2): pp. 243–63. Rom, Mark Carl. 2009. ‘The Credit Rating Agencies and the Subprime Mess: Greedy, Ignorant, and Stressed?’ Public Administration Review 69 (4): pp. 640–50. Scott, Susan, and Nicholas Perry. 2012. ‘The Enactment of Risk Categories: The Role of Information Systems in Organizing and Re-Organizing Risk Management Practices in the Energy Industry’. Information Systems Frontiers 14 (2): pp. 125–41. Sinclair, Timothy J. 2005. The New Masters of Capital: American Bond Rating Agencies and the Politics of Creditworthiness. Ithaca, NY: Cornell University Press. Skidelsky, Robert. 2009. Keynes: The Return of the Master. London: Allen Lane. Smelser, Neil J., and Richard Swedberg. 2005. The Handbook of Economic Sociology. 2nd edn. Princeton, NJ: Princeton University Press. US Senate. 2011. ‘Wall Street and the Financial Crisis: Anatomy of a Financial Collapse’. Hearing before the Permanent Subcommittee on Investigations of the Committee on Homeland Security and Governmental Affairs United States Senate, 112th Congress. White, Lawrence J. 2010. ‘Markets: The Credit Rating Agencies’. Journal of Economic Perspectives 24 (2): pp. 211–26. Wigan, Duncan. 2009. ‘Financialisation and Derivatives: Constructing an Artifice of Indifference’. Competition & Change 13 (2): pp. 157–72. Zinn, Jens O. 2008. Social Theories of Risk and Uncertainty: An Introduction. Malden, MA: Blackwell.

256

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Section V Managing Expectations in Innovative Businesses

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

12 Processing the Future Venture Project Evaluation at American Research and Development Corporation (1946–73) Martin Giraudeau

Introduction The founding of American Research and Development Corporation (ARD) in 1946 in Boston is generally considered to be the starting point of the venture capital industry as we know it today.1 The company’s business model was novel at the time: the publicly owned investment fund would focus exclusively on funding innovative new ventures, so as to ensure both the payment of dividends to its shareholders and large capital gains (Hsu and Kenney 2005). The project seemed excessively ambitious: by dealing with the least established kinds of firms—for which no prior records were available on which to base one’s judgement—and, among these, with firms that were proposing to bring to market new types of products, it would be confronted with the highest possible levels of uncertainty in trying to assess investment proposals. However, the undertaking did eventually succeed financially, generating returns of over 100 per cent per year, on average, from its founding to the company’s sale to Textron in 1972. For these reasons, the ARD case is an historically important and particularly telling example of attempts by business actors to manage uncertainty; and

1 The influence of the company on the later development of the venture capital industry came from the teaching of its managing director, Georges F. Doriot, at the Harvard School of Business Administration; the foundation of competing venture capital companies by former ARD employees, especially in what was soon to be known as Silicon Valley; and the opening up by ARD of subsidiary companies in Canada and Europe.

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

this chapter proposes to explore in detail the methods in place at ARD for appraising new venture proposals.2 These methods were designed and implemented throughout the period by ARD’s managing director, Georges F. Doriot, who had been a professor at the Harvard School of Business Administration since the mid-1920s where he remained until 1966 in parallel with his work at ARD and other directorial positions. Doriot was a pragmatist, influenced by Alfred North Whitehead’s process philosophy. He saw knowledge as always incomplete and consequently advocated the need for constant observation and learning from events. Highly critical of the reliance on fixed theories and routinized rules of thumb in business, he put in place at ARD management procedures that were consistent with these principles and centred around a long and thorough ‘investigation’ of the projects. His advice was to gather all available data on each venture proposal and, at the same time, to ‘live’ with the entrepreneurs and their projects, so as to develop ‘feelings’ about them that would complement available ‘hard’ information. The case examined in this chapter emphasizes the importance of the organized ‘work’ of entrepreneurs and venture capitalists (Giraudeau 2007) in dealing with uncertain futures, and shows that this work is primarily a work of knowledge—although of knowledge understood as fundamentally incomplete. Doriot’s focus at ARD, and in business in general, was on knowing the future to the greatest possible extent by relying on hard forms of knowledge, and on softer ones when the former were not available. Uncertain futures had somehow to be grasped, and this could be done only through a long and painstaking study effort, strictly organized around a series of knowledge technologies and techniques. The future had to be processed. Studying Doriot’s and ARD’s methods of project appraisal in this way adds to the existing literature on the case, which has to date focused exclusively on two other dimensions: the person of Doriot, portrayed in hagiographic terms (Ante 2008; Gupta 2004); and the institutional form of ARD as an original venture capital company coping with the legal frameworks available in its time for investment funds (Etzkowitz 2002; Hsu and Kenney 2005). These limitations are in fact typical of the broader literature on the history of venture capital (Coopey 1994; Fohlin 2013, 2016; Gompers and Lerner 2001; Lerner 2002, 2009; Reiner 1989), which this chapter therefore also aims to complement.

2 The study of the case is based primarily on the Doriot American Research and Development Papers at Baker Library, Harvard Business School (referred to in this chapter as ‘BL’), as well as on the Doriot collection of the French Library in Boston, which is on permanent loan to Baker Library (the materials from the French Library collection are referred to in this chapter as ‘FL’ and under the French Library’s cataloguing system). I am grateful to the Baker Library staff, and especially to Laura Linard, for giving me access to these materials, as well as helpful guidance.

260

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Processing the Future

When project appraisal practices in the venture capital industry have actually been considered, the focus has not been on the use of structured organizational processes or knowledge technologies and techniques, but rather on the role of interpersonal and unmediated interactions, be it through live ‘demos’ of technologies (Rosental 2007, 2013) or business project presentations and, even more so, informal interactions (Shapin 2008, 209–68). These studies have, of course, acknowledged the existence of visual PowerPoint presentations and the circulation of written documents, but formal devices and structured organizational procedures were not made to count for much in the exchanges between entrepreneurs and investors, compared with the speech acts, bodily performance, and interpersonal networks of the participants. ‘Without the advantages of familiarity’, Steven Shapin wrote, ‘the satisfaction of formal criteria means almost nothing’ (Shapin 2008, 288). In the case of business ventures, and especially of the most uncertain ones, financial credit would be an ‘almost’ exclusively intersubjective and primarily informal issue. In their writings, the editors of this volume have reached similarly sceptical conclusions about the possibility of calculated knowledge of uncertain futures, while highlighting the role of imagination. Richard Bronk (2009) placed particular emphasis on the role of specific personal talents in helping decision-makers cope ‘successfully’ with uncertain futures: intuition, vision, creativity, and the ability to spot emerging patterns. So, for example, The creative entrepreneur is usually someone who combines hard-headed calculation with sometimes disturbing and outlandish visions of possible breakthroughs and pitfalls; she is also endowed with an intuitive sense of where opportunities might lie; and she is often gripped by a consuming passion to win recognition for herself and her company in the battle to succeed. (Bronk 2009, 238)

The reliance on ‘hard’ knowledge and formal procedures remains at best a side feature of entrepreneurial behaviour, at the service of ‘the most important aspect of our humanity’, namely its creativity (ibid., 255). The imagination of this somewhat heroic entrepreneur is acknowledged to be a ‘reasoning imagination’ that enables her or him to ‘stress-test . . . visions of the future with a rational analysis of their feasibility in the light of . . . past experience’ and the implications of stable constraints (ibid., 221). But, if imagination is seen as working ‘hand in hand’ with rational analysis (ibid., 304), the emphasis is firmly on the need to ‘imagine’ what the future may hold in conditions where it cannot be known because it has ‘yet to be created’ (218) by the entrepreneur and others with whom he or she interacts. Jens Beckert (2016, 217–68) explained more fully the use of instruments and formal procedures by economic actors but saw them as ‘instruments of 261

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

imagination’ used to cope with uncertain futures. Beckert argues that these analytical procedures and calculative and scenario devices should not be seen as instruments of knowledge. Instead, the plans and expectations of entrepreneurs and venture capitalists, like those of other economic actors operating in uncertain situations, are in essence ‘fictional’—that is, deprived of a direct referential, knowledge link to a future reality as yet non-existent. These ‘fictional expectations’ and associated ‘scientific props’ can at best aim for plausibility and credibility, never correctness. The Doriot case study qualifies this literature in an interesting way and downplays the element of creativity and imagination in entrepreneurship. It suggests that, if entrepreneurs and venture capitalists do rely on imagination, it is in the form of organized and instrument-based procedures and as a default complement to more reliable knowledge. The aim of many of the analytical and decision procedures employed is explicitly to establish as much knowledge as possible about the emerging future. The remainder of this chapter demonstrates this, first, by presenting Georges F. Doriot’s general process approach to management as it appears primarily in his teaching and speeches, and then by exploring the practical implementation of these general principles at ARD.

The ‘Sense’ of the Future Georges F. Doriot’s academic and business career took place in a transitional period: it started between the two World Wars, when large corporations, gaining in size and influence, were subjected to harsh criticism, and it finished in the 1970s, when entrepreneurship became an institutionalized field of practice. In this period, when ‘small business’ became an object of intellectual and political interest, Doriot was among those who attempted to develop a new approach to business, which rejected management systems because of the illusion of control they generated. Influenced in part by the pragmatic philosophy of the period, he acknowledged, instead, the impossibility of full control over business situations, due to their excessive complexity and variability, related to the importance of the ‘human factor’ in their constitution. However, if complete control over a necessarily uncertain future was unachievable, it nevertheless remained a desirable horizon for him: management should tend towards it as much as possible, through the thorough study of every specific business situation. Doriot held that sufficient knowledge was not readily available for confronting most business situations. In the management of large corporations operating in a stable environment, one could rely on ready-made information and management methods, but this was a truly exceptional case. Most of the time, 262

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Processing the Future

uncertainty was too high for one to count on pre-set knowledge for solving the problems at hand, and this was especially the case in small, new, and innovative companies. Such enterprises were, in Doriot’s eyes, like ‘rubber balls’ bouncing around various surfaces, or weak ‘balloons’ pushed around by uncontrollable winds.3 Consequently, pre-set knowledge had to be considered with the utmost caution—especially that resulting from accounting measurements and calculations. It was not possible to rely, or at least rely exclusively, on any such standard ‘yardsticks’ to answer the questions raised in the management of most businesses, and especially of small innovative new ventures. The skills required from people working in and with these types of firms, but also from the top managers of most other companies, therefore had to be of a different kind. These difficulties were particularly acute in relation to valuation issues. A value could, of course, be assigned even to new ventures. Doriot had learned early on, as a student at Harvard in the early 1920s, how to value assets, including at ‘fair value’; and, every year, ARD issued fair value accounts of the companies it invested in, its ‘affiliates’. But Doriot was extremely careful, in ARD’s annual reports, at board of directors meetings, and at shareholder assemblies, always to highlight that the stated fair values could not be safely relied on: they amounted, in his view, to little more than guesswork about future revenues and costs, and could grossly over- or under-state the actual value of ARD’s affiliates, as well as, therefore, that of ARD itself. This wariness that Doriot applied to accounting metrics in fact went much deeper. As managing director of ARD, he rejected the reliance on any kind of set management system. Rejecting first of all the Taylor system, he also dismissed the alternative systems that were later proposed, which he considered to be similarly misleading managerial ‘fashions’. Doriot was even critical of unreflective use of the conclusions of the ‘human relations’ school that arose nearby at Harvard around Elton Mayo, and of all kinds of competing approaches to management that developed over the course of his career, especially if they stemmed from academic disciplines such as ‘sociology’ and, later, ‘psychology’.4 In short, Doriot was opposed to the notion of ‘administration’ altogether— regretting, for instance, that the term figured in the name of the Harvard school he taught at. But if Doriot rejected knowledge systems, and their application in business under the rubric of ‘administration’, he did like another term, ‘operations’. He considered that an approach focused on operations would have

FL: SP-Do. 988 Dor, G.F. Doriot speech on ‘Modern Recovery Experiments’, 1935. See, for example, FL: SP-Do. 988 Dor, G.F. Doriot speech on ‘Modern Recovery Experiments’, 1935. 3 4

263

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

some purchase on the constant shifts and uncertainties that business actors were confronted with, especially in smaller, newer, and/or more innovative entities. If ‘operations’ meant something to Doriot, it was because it evoked in him a sense of self-generative ‘movement’. The rubber ball, or the balloon, was to have its own force. It had, in one word, to be animated with a form of ‘life’, a term that recurs constantly in Doriot’s description of his activities as a venture capitalist. The approach he came up with and promoted relentlessly— in his teaching, at ARD, and in the numerous speeches he gave to business audiences—was summarized as ‘giving life to combinations of men and ideas’.5 In this same spirit, Doriot often referred to ARD’s affiliates as his and his colleagues’ ‘children’, and explained that his task was to help men and companies to ‘grow’, not in a financial way but in a broad sense, to the maximum of their potential. The ‘movement’ he focused on in business was not the predictable movement of mechanical systems, but the more flexible and continuous movement of living organisms, and especially of ‘human life’. Despite Doriot’s understanding of the indeterminacy and ‘vitality’ of the business world, he never simply renounced control, nor the aim of attaining the knowledge that allowed it. Rather, he saw control as having to be regained in each singular situation, through other means than pre-set recipes. The first task of the wise entrepreneur was to reduce the number of forces constraining his company, for instance by avoiding subjecting it to the desiderata of the banker. Some freedom of operation had to be regained. But, above all, the movement of business could also be controlled to a certain extent. Economic ‘life’ could be managed. The way to give ‘life’ to ‘combinations of men and ideas’ was, for Doriot, to identify and ‘nurture’ them. ‘Ideas’ were defined broadly as any kind of innovation, be they technological or commercial, productive and so on, very much along the lines of Doriot’s Harvard colleague Schumpeter’s definition of innovation (Peneder and Resch 2015). Identifying any such innovation was the task of entrepreneurs and venture capitalists, and to a lesser extent of all business ‘leaders’. The ability to do so was a crucial skill required from them. What mattered even more to Doriot was the ‘men’: he repeatedly stated that one should invest in a ‘grade-A man with a grade-B idea’ rather than in a ‘grade-B man with a grade-A idea’.6 ‘Men’ were the main source of life, of movement in the ‘combinations of men and ideas’, and choosing the right ‘men’ was therefore a crucial way of controlling this movement. 5 Women were primarily considered by Doriot as spouses—his own, those of his students, and those of the entrepreneurs applying for ARD funding. 6 Cited, for example, in Gene Bylinsky, ‘How Do You Turn a Scientific Idea into a Moneymaking Business? That’s the Profitable Mission of General Doriot’s Dream Factory’, Fortune, August 1967, pp. 103–36.

264

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Processing the Future

Their technological, managerial, or financial skills did matter, but what counted above all were their moral traits. Doriot was eager to work with men who were ‘hard-working, helpful, decent’, and who showed ‘politeness, deftness, interest in people, dependability’.7 More crucially, he stressed the need to identify ‘superior men’, who could lead a new venture to success, thanks to moral traits such as courage, resourcefulness, curiosity, willingness to learn, ambition, or leadership. These traits constituted a specific ‘drive’ that allowed ‘creativity’, or the ability for these exceptional men to rely on what Doriot called their ‘creative glands’ and ‘imagination’.8 The ‘movement’ of business operations came from this ‘creativity’, which could be found in a rare class of exceptional individuals. The difficulty, of course, was to identify the ‘right’ men and ideas: ideas had to be innovative, and men creative; and imagination was therefore at the heart of Doriot’s approach to business. But it was a very specific form of imagination, which tended towards knowledge. What was needed of entrepreneurs and venture capitalists, Doriot explained, was an ‘ability to sense—to foresee—to smell’.9 Decision-making in the uncertain world of entrepreneurship had to rest on what he called a ‘sense’ of the future. In the absence of the possibility of full knowledge, it was a tentative quasi-knowledge that had to be relied on. This approach can be better understood by looking at its origins. It should first be noted that Doriot was the son of a French car engineer. He had been tempted by engineering himself, initially planning to register at MIT rather than at Harvard upon his arrival in the United States in 1921. He also worked alongside engineers for the US Army, before and during the Second World War. These biographical elements are important because Doriot seemed to share an engineering sense of technical, as well as managerial, ingenuity, focused on finding specific solutions to localized issues more than on producing general laws. He had noticed this difference in his approach to business while working in a major New York investment bank after his business administration studies: there, he had proved keener than his colleagues on developing thorough knowledge of the manufacturing processes of the bank’s clients, as well as of their markets, instead of relying exclusively on their accounting data. Beyond these biographical experiences, what must be emphasized particularly strongly are the intellectual influences Doriot was subject to as a student and then as a professor at Harvard. The Harvard School of Business Administration was, from 1919 to 1942, directed by Dean Wallace B. Donham, who

7 8 9

FL: SP-Do 985 Dor, G.F. Doriot speech on ‘the role of the entrepreneur’ in Montreal, 1973. FL: SP-Do 985 Dor, G.F. Doriot speech on ‘the role of the entrepreneur’ in Vancouver, 1974. FL: SP-Do 984 Dor, G.F. Doriot speech to DEC financial officers in Hyannis, 2 October 1973.

265

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

impressed on the school, its students, and faculty, a specific view of business and business education (Khurana 2007, 126–30 and 174–84). Donham did not reject the term ‘administration’ itself, but he gave it his own definition: administration involves consideration and weighing of both current facts and past experience in efforts to apply foresight where foresight is possible . . . The skillful administrator should in most cases foresee in general terms the possibilities which may occur, and this foresight should be specific enough to guard against the worst effects of the specific happening. (Donham 1936, 405)

To develop such foresight among American business administrators, Donham promoted the professionalization of management, and this professionalization rested not on the learning of pre-set theories and recipes, which could at best prove to be useful guidelines in given situations, but rather on the acquisition of analytic skills, through the cautious study of specific business ‘cases’—with the help of analytic cues borrowed from a variety of disciplines, ranging from engineering to economics and sociology. It was Donham himself who had brought Doriot back to Harvard, as Associate Dean, after his stint in the New York investment bank. The two men worked closely together and Doriot, for instance, embraced the case method, to the point that he exported the method to France, by initiating the creation of the Centre de Préparation aux Affaires at the Paris Chamber of Commerce, and then INSEAD. Like Donham, Doriot believed in the merits of nonstandardized approaches to specific business situations, and of broad analysis at the crossroads of numerous disciplines, rather than within the closed bounds of unique theories. Donham’s views, and with his Doriot’s, were informed by the process philosophy of Alfred North Whitehead, who joined Harvard in 1924. Donham invited Whitehead to give speeches at the business school and referred to him regularly; and Donham found in his work a definition of knowledge well suited for his purposes. For Whitehead, knowledge was never complete and he therefore placed the emphasis on learning as a continuous process, and one that articulated discipline with imagination. At this intersection of discipline and imagination was ‘foresight’, which Whitehead analysed in a lecture given at the Harvard School of Business Administration that was then published in a 1931 book by Donham. In his eyes, the role of business education was to ‘prepare the young to face novel conditions’ by acquiring ‘a deeper knowledge of the varieties of human nature’ (cited in Hendley 2000, 185). This deeper knowledge would trigger their imagination and lead to the development of foresight when they were confronted with concrete and often surprising business situations. Whitehead’s Science and the Modern World (1925 [2011]), in which he exposed his general theory of knowledge, was one of the few readings assigned to his students by Doriot, who 266

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Processing the Future

could not but be sensitive to such a definition of knowledge as a process intrinsically imbued with imagination and allowing foresight. For Donham and Whitehead, the figure on whom the modern professional business person had to be modelled was the medical clinician. As Donham phrased it, ‘the experienced clinician utilizes the reports of the bacteriologist, the chemist, and the radiologist, who well know that they are themselves incompetent to prescribe treatment. As one of my colleagues has pointed out, the clinical physician has always dealt with cases’ (Donham 1936, 409). The colleague in question was unmistakably the influential Harvard physiologist Lawrence J. Henderson, who had extended his reach to the study of sociology and business administration. Henderson, who also referred to Whitehead on education, and whom Doriot invited to lecture on his course,10 advocated the reliance on Hippocratic virtues in business: therapeutic wisdom, based on the experience of numerous cases, was to be preferred over false rationalizations (Cross and Albury 1987, 181; Henderson 1927). The physiological simile, however, went further: for Donham, Whitehead, Henderson, and Doriot, rapid technological progress was threatening the equilibrium of the social system, which they saw as an organism whose internal balance was disturbed by chance events in its environment. Referring to the notion of ‘milieu intérieur’ promoted by Claude Bernard (also an assigned reading on Doriot’s course), they sought ways of restoring social homeostasis;11 and the promotion of a Hippocratic approach to business was intended to permit this. The carefully controlled form of ‘life’ that Doriot aimed at giving to business found its roots in this organicist understanding of society. If Doriot borrowed Schumpeter’s open definition of innovation and also seemed to share many of his views on the personality traits required from entrepreneurs, he was not inclined to follow him in the idea of ‘creative destruction’. Doriot’s target was rather the facilitation of the seamless acceptance of new ‘ideas’ within markets and society, through a broad and cautious study of every situation at hand. He was in that sense quite close to Frank Knight’s perspective on entrepreneurship. In Knight’s view, the entrepreneur was the person who was able and willing confidently to assign subjective probabilities to the possible outcomes of singular situations. This ‘subjective’ exercise was, however, far from being mere guesswork. The entrepreneur aimed at progressively ‘securing better knowledge of and control over the future’, by relying, on the one hand, on the hard data that was becoming

10 BL: Lawrence Joseph Henderson Papers, Lecture to Georges F. Doriot’s Class, December 1, 1936. 11 The notion of homeostasis was itself coined by another Harvard physiologist of the time, Walter B. Cannon (see Cross and Albury 1987).

267

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

increasingly available thanks to the development of statistical offices, trade journals, and so on, and, on the other hand, on an understanding, based on past experience, of the given situation (Knight 1921 [1965], part III, chapter 8, §42).12 The output of agencies specialized in knowledge, in particular, ‘increases the value of the intuitive “judgments” on the basis of which [the business manager’s] decisions are finally made after all’ (Knight 1921 [1965], III, 8, 43). Furthermore, Knight considered that there was an uneven distribution, across the population, of the ability to deal with uncertainty: a specific form of intelligence and character was required that not everyone has, which led to what he called (drawing, like Doriot et al., on organicism) a social ‘cephalization’, whereby some particularly gifted individuals took over the central control function of the social body (ibid., Part III, Chapter 9, §8). As for most of the other influences on Doriot discussed here, the emphasis on a specific form of knowledge able to adapt to specific circumstances led to a strong form of social elitism. It is clear then that Doriot was part of a broader intellectual movement that acknowledged—particularly after the 1929 crisis—the omnipresence of uncertainty in the economy, and more broadly in society, but did not abandon the ideal of thorough knowledge. Instead, the members of this movement developed a theory of knowledge, finding in Whitehead their leading figure for that purpose. If ‘mechanical objectivity’ (Daston and Galison 2007) could not be achieved in the social sciences, and especially in the new interdisciplinary field of business administration, this did not mean that the social system could not be understood or acted upon rationally. The uncertainty of the world could be grasped reasonably well, for all practical purposes, thanks to a process and diagnostic form of knowledge, understood not as a stock but as a flux and grounded in organicist metaphors. This form of knowledge involved some subjectivity, that of the Hippocratic physician, but it was nevertheless reliant on prior experience and learning, and a thorough analysis of past and present situations. If imagination was involved in the process, it was an imagination triggered and simultaneously disciplined by the mixture of prior knowledge and present observation. It is precisely this approach to knowledge that was relied on in the practice of entrepreneurship and venture capital at ARD. 12 It is interesting to note that Knight referred to the development of proto-venture capitalism, and explained how the economic actors involved dealt with uncertainty by pooling a large number of ventures together to get the benefits of portfolio diversification: ‘A considerable and increasing number of individual promoters and corporations give their exclusive attention to the launching of new enterprises, withdrawing entirely as soon as the prospects of the business become fairly determinate. The gain from arrangements of this sort arises largely from the consolidation of uncertainties, their conversion by grouping into measured risks which are for the group of cases not uncertainties at all’ (Knight 1921 [1965], part III, chapter 8, § 38).

268

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Processing the Future

The Project Appraisal Process at ARD Project appraisal at ARD was a long, in fact never-ending process, which continued throughout the nurturing of its affiliates by the investment company. The early stages of this appraisal, for new proposals submitted to the company, is of particular interest here. It was structured along an extended period of months of what was named the ‘investigation’, which preceded any investment decision. If the role of ‘feelings’ was put forward in this process, the ‘sense’ of the future that was expected to be gained through investigation was far from unstructured. It was produced through strict procedures and the use of rigorous knowledge technologies and techniques. Doriot was averse to writing—at least in print form. It should be noticed, for instance, that he only published a handful of brief articles in his entire career as a professor and well-known business figure (Doriot 1930, 1932, 1944). Even when giving speeches, he regularly joked at the end of his talks that his audience should pay no attention to what he had said. He wanted to avoid being considered the father of a new system. This rejection of a certain use of formality was also present at ARD. When evaluating a new venture project, employees were told: ‘you may do so on the basis of the feeling of wanting to go ahead but without concrete justification. It is this kind of feeling that produces good managerial decisions.’13 Relatedly, the organization of ARD was highly personalized; decisions were in the hands, or minds rather, of individuals involved in durable face-to-face relationships with entrepreneurs. It was a small company of fewer than ten employees, who were invited to ‘live with the principals’ of the projects over weeks and months, so as to develop the required feelings for the ‘men’ and their ‘ideas’. However, despite this emphasis on feelings and face-to-face interaction, Doriot did in fact have a number of clearly articulated and strongly enforced administrative procedures in place at ARD. They were designed in response to information-sharing needs, among the parties within ARD, and towards outside parties to which the company was accountable. Administrative formalities were used as a means to mediate between different categories of individuals involved in the firm. If the investigation process was personalized, it was nevertheless collective. Doriot asked to be directly informed by his employees of any doubts or issues they had with a project. Feelings were not incommunicable but had to be shared and discussed, be it orally or in writing, among the employees of the venture capital firm, so that a collective sense of the future could be produced. However ‘soft’ the approach, it was nevertheless framed and controlled by specific techniques. 13

FL: SP-Do 990 Dor, G.F. Doriot document on ‘analysing a company’, undated.

269

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

If Doriot promoted the reliance on feelings or a ‘sense’ of the future, it was only as a means of extending knowledge in directions where it was otherwise unavailable, and even these feelings had to be grounded. Doriot was in fact wary of certain types of feelings. Indeed, one of the challenges of Doriot’s approach was to control the influence of emotions in the construction of ‘feelings’ for the appraised projects. Doriot regularly warned his employees, as well as ‘sources’—the people, such as investors and board members, who would recommend interesting projects to them—that ‘love’ for given men and/or ideas should be dreaded as a potentially misguiding emotion.14 ‘Living with the principals’ could lead to an emotional attachment to them or their project that would make one overlook deficiencies—for instance, by disregarding important financial information or the lack of necessary skills or traits in the entrepreneurs. Dangerous attachments could also develop for specific types of projects or industries deemed exceptionally promising. Doriot was wary of ‘fashionable’ imaginaries of the future, starting with technological imaginaries (Jasanoff and Kim 2015), for instance regarding the field of electronics or nuclear energy. ‘People love to finance mysteries’, he complained.15 The right kinds of feelings were for Doriot the ones that were grounded in thorough analysis, and a broad understanding of business, society, and human nature. To avoid misplaced feelings developing at ARD, he imposed a strict requirement for due process, developing a number of guidelines for the appraisal of venture proposals. First, there were a number of intellectual guidelines designed to help employees crystallize what their sense of a given project was. They included a number of ‘curves’ or ‘chains’ of ‘evolution’, which described the different paths that a human individual or a company could go through in their ‘lives’.16 ARD employees were invited to reflect on what the latent curve of a given individual or company was (rising? flat? falling?), and at what place on that curve the person or firm was currently (at the time of evaluation)—all with the purpose of answering the fundamental question: ‘can individual and organization grow?’17 Another such guideline was an invitation to develop a new kind of balance sheet for the projects under consideration: the traditional accounting assets were to be thought of as liabilities, and the liabilities as assets. Such a technique helped his staff formalize, and thereby visualize, the predicament of the venture proposals being examined. It gave structure to their sense of what was

14 15 16 17

270

BL: Georges F. Doriot ARD Papers, ARD Annual Report, 1946. FL: SP-Do. 1007 Dor, G.F. Doriot document on ‘project study’, 1963. FL: SP-Do. 1007 Dor, G.F. Doriot document on ‘curves of evolution’, 1963. FL: SP-Do. 1007 Dor, G.F. Doriot document on ‘project study’, 1963.

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Processing the Future

going on, furnishing their analysis with a set of conceptual grids that enabled a shared interpretation of emerging reality. The most important of Doriot’s guidelines for the appraisal of new projects were long and numerous lists, which could be found everywhere in Doriot’s teaching notes and memos to his ARD employees. They included especially lists of personality traits, to help assess the ‘men’. Other lists were developed regarding ‘ideas’ and the ‘combinations of men and ideas’. Lists were such a crucial type of practical vehicle for Doriot’s conceptions of project appraisal because they had the advantage of not being systematic: they simply included a variety of points that Doriot believed should be considered when evaluating the promise of a given project. If ‘feelings’ could not be grounded in systematic approaches, they nevertheless had to be based on information, the scattered and incomplete but reliable ‘factual knowledge’ that he said should be combined with ‘emotional drive’.18 With his lists, Doriot attempted to promote what he called ‘reasoned thinking, and proper understanding of the aims, possibilities, and limitations of the undertaking in which one is considering a financial commitment’.19 ‘It should be understood and appreciated’, he added elsewhere, ‘that a careful, intelligent, constructive, well conducted investigation can lead to a venture with far better hopes of success’.20 There could be no feeling for, nor successful building of, companies without informed thinking; and this informed thinking could be much more formal than one would have imagined based on Doriot’s emphasis on feelings. Even accounting data could, to a certain extent, prove useful: ‘Figures and statistics, carefully developed, can form a foundation upon which you will build feelings.’21 Beyond these intellectual guidelines, and in spite of his general rejection of the term ‘administration’, Doriot also developed some administrative guidelines, and even a set of administrative rules framing the ‘investigation’ process. A strict schedule was defined for the appraisal of projects: a series of meetings were to take place in the right order before decisions could be made. Internal ‘project meetings’ were to take place frequently to review the advances in the process. When sufficient understanding of a project had been achieved, it would be presented in a meeting of the board of directors, which took place on a termly basis. The board of directors was expected to voice opinions and concerns, and to make a decision on whether the project could be of interest to the company. Ultimately, the final decision to continue investigation, or to invest in a project, was to be made by an executive

18 19 20 21

FL: SP-Do 990 Dor, G.F. Doriot document on ‘analysing a company’, undated. BL: Georges F. Doriot ARD Papers, ARD Annual Report, 1946. BL: Georges F. Doriot ARD Papers, ARD Annual Report, 1970. FL: SP-Do 990 Dor, G.F. Doriot document on ‘analysing a company’, undated.

271

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

committee meeting, which took place shortly after the board meeting. Feelings had to lead to decisions, and the decision-making process followed a due administrative order. This schedule was connected with specific formalities. The proposal of projects could take different forms: a source could mention in a letter to Doriot or another ARD employee the case of an entrepreneur with an ‘idea’, for instance an academic or engineer who had obtained a patent for a new technology; entrepreneurs could also get in touch with ARD directly. Yet in all of these cases, documents had to be provided. These were not specified at first, but it seems that in the first decade of the company entrepreneurs sent mostly technical documents, describing their innovative products. ARD would then ask for complementary information, as in the famous 1957 case of Digital Equipment Corporation, ARD’s greatest financial success: the two entrepreneurs submitted a brief document presenting the computer they intended to commercialize and, following a meeting at ARD headquarters in Boston, the venture capital company asked them for further documents, starting with the resumes of the entrepreneurs.22 Feelings for company prospects had then to be grounded in paperwork, as well as in analytic and administrative procedures, and all the more so as time went by. Little by little, as the number of proposals increased with the rise of the world of entrepreneurship and its venture capital industry, education programmes, incubators, and so on, the guidance became more specific (Giraudeau 2011, 2012). In 1960, for instance, the ARD annual report stated: Prior to undertaking these personal meetings, it is helpful if ARD can learn as much about a project as possible. Therefore, proposals should preferably be accompanied by information about the products, people, facilities, financial records and projections, so that a clear understanding of the past history, present status and future prospects of the project can be gained.23

Business plans would not appear as a named category and in their standardized form until the early 1970s, but all the components were already listed in these requests for information on submitted projects. Formalities continued after the submission of initial proposals and accompanying paperwork. The entire purpose of the investigative process may have been to build up informed feelings within ARD employees, but the decision schedule in place required the preparation of further documents: the received

22 Computer History Museum, item # 102664470, Kenneth E. Olsen and Harlan E. Anderson, letter to William H. Congleton of American Research and Development Corporation, 22 April 1957. 23 BL: Georges F. Doriot ARD Papers, ARD Annual Report, 1960.

272

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Processing the Future

projects were to be recorded within the ‘project book’, which Doriot himself reviewed regularly; and reports on each project had to be prepared prior to the ‘project meetings’, building up towards the practical end of the investigative process, which was to produce a ‘formal plan’ that could be considered in the general and executive board meetings. ARD would thus, technically, produce an in-house business plan. Further and even more detailed due diligence formalities applied at later stages. Altogether, the investigation and nurturing of new ventures involved a dense trail of paperwork, which was closely monitored by Doriot himself. It is therefore clear that Doriot’s emphasis on the ‘life’ of business and the development of feelings did not entail, in his practice at ARD, a full rejection of administrative formalities. In spite of his critique of administration, he put in place sophisticated decision-making procedures within the firm. There was no contradiction, in fact: the procedures established by Doriot were particular ones, distinct from what he called ‘administration’. They were designed to make it possible for ARD staff to ‘live with the principals’, while at the same time remaining thorough and objective in their processing of the available information, as well as accountable internally to Doriot and externally to the firm’s shareholders. Doriot summarized his warnings to his staff at ARD like this: ‘The analysis cannot be an autopsy but it must be a living idea’,24 by which he meant primarily that it was an ongoing process rather than a terminal one. In terms of procedure, it was especially crucial for him to keep things open, and thereby alive. Decisions should always come late, at the end of a long period of investigation, rather than being made fast, either under the influence of misguided ‘love’ for a project or, just as dangerously, under the illusory impression of already having sufficient ‘knowledge’ of a project’s future. A project could always be turned down if further, worrying information appeared—even after a company had become an affiliate; and ARD did divest from some companies under Doriot’s watch. A project could always be reopened for investigation, and Doriot regularly went through the project book when new information emerged about a given project. The aim was always, for Doriot, to ‘learn’, and thereby increase his knowledge, and improve the procedures he had put in place. Doriot’s approach to entrepreneurship and venture capital did not deal with uncertainty by ensconcing it in systematic forecasts and plans, but nor did it rely on the personal intuition and creativity of imaginative individuals. Rather, it was structured around a thorough process of investigation, which aimed at maximizing the amount of knowledge available on every venture. This

24

FL: SP-Do 990 Dor, G.F. Doriot document on ‘analysing a company’, undated.

273

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

knowledge, even if based on feelings, was produced through specific techniques and technologies, thanks to which it could be formalized to an extent at least sufficient to make discussion, and consequent decisions, possible. While the determinants of business success are always complex, Doriot’s approach appears to have had considerable impact from an economic perspective—whether financially for the company’s shareholders, or in terms of the development of technologies and companies, or as a model organization in the development of the venture capital industry. To some extent at least, the economic success of ARD can perhaps be attributed to Doriot’s talent not so much as a critic of administration, but as an implementer of the precepts of the pragmatic movement in business administration. Under the influence of this movement, he was an ingenious administrator, relying on specific formalities that allowed his staff to implement an original approach to the appraisal of new venture projects.

Conclusion The ARD case can be seen as a typical case of ‘when formality works’. Arthur Stinchcombe, indeed, showed that administrative procedures and tools do sometimes prove helpful, even in uncertain settings: budgets, for instance, play a number of beneficial roles in the coordination of actors, even when the numbers they put forward are not actually reliable; they provide the necessary structure and hence discipline to the actors’ personal visions and interactions (Stinchcombe 2001). The administrative formalities in place at ARD would exemplify such a role for formalities in business practices. Tellingly, Stinchcombe relied, like Doriot and many of his contemporaries, on an organic metaphor to describe such a phenomenon. He presented formalities as the ‘backbone’ of decision-making, which would support the soft tissue of the organism, and thus allow its movement (ibid., 132). But what is particularly interesting to notice is that, for Stinchcombe as for Doriot, recognizing the limitations of mechanical approaches to business was not equated with the discarding of knowledge, nor with its mere replacement with imagination and emotions. If imagination and emotions were present in the observed settings, they were triggered by the available knowledge, and their potential insights would lead to the production of new knowledge. As Bruno Latour put it in relation to scientific work, ‘most strokes of genius, most flashes of intuition . . . can be explained by the proximity, on the tables of laboratories, of recombined traces’ (Latour 1985). For that reason, imagination and emotions were under constant control at ARD. Although considered useful, they were seen by the actors as a threat, to the 274

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Processing the Future

same extent as systematic approaches were. Imaginative foresight was under strict knowledge oversight. This emphasis on the part played by knowledge and its formalities in the practice of entrepreneurship and venture capital shows that these activities cannot be characterized as exclusively or even primarily unmediated interpersonal exchanges. Further, without denying a certain role for a specific form of imagination, it limits the degree to which the production and use of the future in uncertain settings can be considered ‘romantic’ (Bronk 2009) or ‘enchanted’ (Beckert 2016). The imagination of the actors observed in the Doriot case could not simply substitute for deficient knowledge in situations of uncertainty. Nor was it an individual and unmediated faculty. Rather, imagination was indistinguishable from the learning process itself, engendered by knowledge as much as stimulating it, and engrained in technologies and procedures. For this reason, imagination in this case cannot exactly be called ‘fictional’, or only in a restrictive sense. Whereas a novel gains its credibility mostly from its internal consistency and the inclusion of ‘little true facts’ that make the story look ‘as if ’ it were real (Barthes 1982), the formalized expectations of Doriot and his employees were knowledge artefacts, involved in a deeply referential relationship to the world around them in the form of present pointers to the future. This was all the more true as these documents were process artefacts themselves: they were drafts undergoing constant revision that stuck as closely as possible to observed developments in the situation (Giraudeau 2008). To a large extent, the process was so extended temporally that ARD seemed to be waiting for the future to be realized before a conclusion, or a decision, was pronounced about it. Pointing out the limited romanticism and enchantment of the capitalist practices examined here does not, however, bring us back to Weber’s iron cage. It does consist in observing that economic actors rely on calculative devices and administrative procedures. But this emphasis on knowledge practices and devices does not point to a predetermined ‘rationality’ of the actors. It simply invites us to consider another regime of action that is neither rational nor charismatic nor traditional, but instead reliant on a continuous, and continuously incomplete, effort to attain knowledge in its dealings with the world, present and future.

Acknowledgements I am most grateful to the two editors of this volume for their close reading and sharp discussion of earlier versions of this chapter; to Lukas Rieppel for eye-opening suggestions on the intellectual genealogy of Doriot’s approach to business; and to Zsuzsanna

275

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures Vargha as well as the participants in the ‘History and Philosophy of Political Economy’ and ‘Accounting, Organizations and Institutions’ seminars at the London School of Economics and Political Science for their insightful comments.

Bibliography Ante, Spencer E. 2008. Creative Capital: Georges Doriot and the Birth of Venture Capital. Cambridge, MA: Harvard Business Press. Barthes, Roland. 1982. ‘L’effet de réel’. In Littérature et réalité, edited by Gérard Genette and Tzvetan Todorov, pp. 81–90. Paris: Le Seuil. Beckert, Jens. 2016. Imagined Futures: Fictional Expectations and Capitalist Dynamics. Cambridge, MA: Harvard University Press. Bronk, Richard. 2009. The Romantic Economist: Imagination in Economics. Cambridge: Cambridge University Press. Coopey, Richard. 1994. ‘The First Venture Capitalist: Financing Development in Britain after 1945, the Case of ICFC/3i’. Business and Economic History 23 (1): pp. 262–71. Cross, Stephen J., and William R. Albury. 1987. ‘Walter B. Cannon, L. J. Henderson, and the Organic Analogy’. Osiris 3 (1): pp. 165–92. Daston, Lorraine J., and Peter Galison. 2007. Objectivity. New York: Zone. Donham, Wallace B. 1936. ‘The Theory and Practice of Administration’. Harvard Business Review 14 (4): pp. 405–13. Doriot, Georges F. 1930. ‘The Meaning of Rationalization in Europe’. The Annals of the American Academy of Political and Social Science 149 (1): pp. 61–6. Doriot, Georges F. 1932. ‘An Engineering Phase of World Peace’. Advocate of Peace through Justice 94 (1): pp. 38–9. Doriot, Georges F. 1944. ‘Environmental Protection’. Proceedings of the American Philosophical Society 88 (3): pp. 196–203. Etzkowitz, Henry. 2002. MIT and the Rise of Entrepreneurial Science. London: Routledge. Fohlin, Caroline. 2013. ‘Creating Modern Venture Capital: Institutional Design and Performance in the Early Years’. Available at SSRN: http://dx.doi.org/10.2139/ssrn. 2197840. Fohlin, Caroline. 2016. ‘The Venture Capital Divide: Germany and the United States in the Post-War Era’. Available at SSRN: http://dx.doi.org/10.2139/ssrn.2849237. Giraudeau, Martin. 2007. ‘Le travail entrepreneurial, ou l’entrepreneur schumpetérien performé’. Sociologie du Travail 49 (3): pp. 330–50. Giraudeau, Martin. 2008. ‘The Drafts of Strategy: Opening Up Plans and Their Uses’. Long Range Planning 41 (3): pp. 291–308. Giraudeau, Martin. 2011. ‘Imagining the (Future) Business: How to Make Firms with Plans?’. In Imagining Organizations: Performative Imagery in Business and Beyond, edited by François-Régis Puyou, Paolo Quattrone, Chris McLean, and Nigel Thrift, pp. 213–29. London: Routledge. Giraudeau, Martin. 2012. ‘Remembering the Future: Entrepreneurship Guidebooks in the US, from Meditation to Method (1945–1975)’. Foucault Studies 13: pp. 40–66.

276

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Processing the Future Gompers, Paul A., and Josh Lerner. 2001. ‘The Venture Capital Revolution’. Journal of Economic Perspectives 15 (2): pp. 145–68. Gupta, Udayan. 2004. The First Venture Capitalist: Georges Doriot on Leadership, Capital, and Business Organization. San Francisco, CA: Gondolier. Henderson, Lawrence J. 1927. ‘Business Education as Envisaged by the Scientist’. Harvard Business Review 5: pp. 420–3. Hendley, Brian. 2000. ‘Whitehead and Business Education: A Second Look’. Interchange 31 (2–3): pp. 179–95. Hsu, David H., and Martin Kenney. 2005. ‘Organizing Venture Capital: The Rise and Demise of American Research and Development Corporation, 1946–1973’. Industrial and Corporate Change 14 (4): pp. 579–616. Jasanoff, Sheila, and Sang-Hyun Kim. 2015. Dreamscapes of Modernity: Sociotechnical Imaginaries and the Fabrication of Power. Chicago, IL: University of Chicago Press. Khurana, Rakesh. 2007. From Higher Aims to Hired Hands: The Social Transformation of American Business Schools and the Unfulfilled Promise of Management as a Profession. Princeton, NJ: Princeton University Press. Knight, Frank H. 1921 [1965]. Risk, Uncertainty and Profit. New York: Harper & Row. Latour, Bruno. 1985. ‘Les “vues” de l’esprit: une introduction à l’anthropologie des sciences et des techniques’. Culture Technique 14: pp. 5–29. Lerner, Josh. 2002. ‘When Bureaucrats Meet Entrepreneurs: The Design of Effective “Public Venture Capital” Programmes’. The Economic Journal 112 (477): pp. F73–F84. Lerner, Josh. 2009. Boulevard of Broken Dreams: Why Public Efforts to Boost Entrepreneurship and Venture Capital Have Failed—and What to Do about It. Princeton, NJ: Princeton University Press. Peneder, Michael, and Andreas Resch. 2015. ‘Schumpeter and Venture Finance: Radical Theorist, Broke Investor, and Enigmatic Teacher’. Industrial and Corporate Change 24 (6): pp. 1315–52. Reiner, Martha L. 1989. The Transformation of Venture Capital: A History of Venture Capital Organizations in the United States, PhD Dissertation, University of California, Berkeley. Rosental, Claude. 2007. Les capitalistes de la science: enquête sur les démonstrateurs de la Silicon Valley et de la NASA. Paris: CNRS Éditions. Rosental, Claude. 2013. ‘Toward a Sociology of Public Demonstrations’. Sociological Theory 31 (4): pp. 343–65. Shapin, Steven. 2008. The Scientific Life: A Moral History of a Late Modern Vocation. Chicago, IL: University of Chicago Press. Stinchcombe, Arthur L. 2001. When Formality Works: Authority and Abstraction in Law and Organizations. Chicago, IL: University of Chicago Press. Whitehead, Alfred N. 1925 [2011]. Science and the Modern World. Cambridge: Cambridge University Press.

277

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

13 Discounting and the Making of the Future On Uncertainty in Forest Management and Drug Development Liliana Doganova

The Problem of Uncertainty How do economic actors make decisions about investments whose returns are in the future and uncertain? This chapter focuses on the main tools used to assess investments: discounted cash flow (DCF) and net present value (NPV). Their rationale is as follows: to make a decision about an investment project, one has to outline the cash flows that the project will generate in the future, discount them by applying a certain discount rate, and then add them up. The sum of the discounted cash flows indicates the net present value of the project; if this value is positive, the project is worth investing in. Discounting future flows is the basic principle of valuation in corporate finance and corporate practice. It carries a very particular theory of value, which rests on two central assumptions. First, value stems from the future, rather than from the present or the past. Second, the future is to be discounted, because it is distant and any choice of action involves opportunity costs and uncertain outcomes. As finance textbooks explain, a euro today could be saved at a certain rate of interest, and hence be worth more than one euro in, say, two years’ time. Moreover, the same euro could be invested in a different project, which might provide higher returns. In this theory of value, time is seen as having a cost, and investment is seen as a sacrifice, because it implies missing the opportunity to do something else. The aim of this chapter is to position discounting and its theory of value as an object of sociological analysis by questioning how, and with what effects,

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Discounting and the Making of the Future

economic actors discount the future in order to make statements about the value of things and determine relevant courses of action. Examining the future in order to decide about the present is not straightforward. First, as historians and sociologists have shown, this is not a natural characteristic of economic actors, but rather a competence that they have gradually, and sometimes painfully, acquired with the development of capitalism. In his study of the industrial revolution in Great Britain, Pollard (1965) described all the efforts that employers had to make in order to get workers—described by contemporaries as ‘improvident’, having ‘no care for the morrow’ (p. 196)—to enlarge their time horizons beyond the present day. Bourdieu (1963) and Beckert (2016) have emphasized that such a change in attitude to the future is a key element in the transition of a traditional society to capitalism. Likewise, Muniesa (2016) has analysed the role of pedagogical material at Harvard Business School in setting the ‘habit’ of capitalization, that is, the idea that the value of things resides in their capacity to generate future streams of revenues for an investor (Birch 2016; Doganova and Muniesa 2015; Muniesa et al. 2016). Moreover, as many scholars in the field of economics have noted, an orientation towards the future entails uncertainty, and uncertainty impedes the very possibility of calculation. Knight (1921) famously distinguished between situations characterized by ‘risk’—in which ‘calculation a priori’ or ‘statistics of past experience’ render uncertainty a matter of measurable probabilities—and situations characterized by genuine uncertainty, the best example of which ‘is in connection with the exercise of judgment or the formation of . . . opinions as to the future course of events’ (p. 233). Keynes (1936) coined the term ‘animal spirits’ to describe the behaviour of homo economicus in situations of uncertainty, for example when faced with ‘estimating the yield ten years hence of a railway, a copper mine, a textile factory, the goodwill of a patent medicine’ (p. 149). Focusing on the economics of innovation, Freeman (1986) described the assessment of R&D projects as ‘a process of political advocacy and clash of interest groups rather than sober assessment of measurable probabilities’ (p. 151); and he compared the use of valuation techniques such as DCF to ‘tribal dances’, which certainly ‘play a very important part in mobilizing, energizing and organizing’ (p. 167), but have little to do with the production of true statements about value. In a similar vein, sociologists have argued that in situations of uncertainty economic actors do not behave as economic theory would have them do. In such situations, rational decision-making based on the calculation of expected benefits and costs gives way to judgements whose formation can be understood only by resorting to extra-economic explanations. For example, Karpik (1989) argued that when it comes to ‘incomplete’ goods and services, which appear as ‘promises whose reality can only be tested by time’ (p. 206), markets 279

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

operate through social mechanisms such as trust and networks. Another interesting example is provided by studies of entrepreneurs’ use of business plans: Karlsson and Honig (2009), for example, explain the time and resources devoted to planning the future—an activity doomed to failure because of uncertainty—as the search for legitimacy or as mimetic behaviour. Beckert (2016) argues convincingly that sociology has lagged behind economics in the role it assigns to the future, and urges us to take seriously the concept of uncertainty, the ‘openness of the future’, and the contingency of actors’ expectations about this open future. This chapter argues that developing such a sociological understanding of the future requires denaturalizing uncertainty and examining what it means in practice. ‘Uncertainties’, as Power (2007, 9) writes, ‘do not exist sui generis but must of necessity be organized, ordered, rendered thinkable, and made amenable to processes and practices of intervention’. This chapter proposes to envisage uncertainty in such a way: not as an analytical category that can be readily borrowed from economists (such as Knight 1921 or Shackle 1949), but as the very object of analysis on which the sociologist should focus. The following questions then arise. How is uncertainty defined when actors envisage the future? Which portions of the future are labelled ‘uncertain’, as opposed to risky or in some sense certain? How is the uncertainty thus defined taken into account in future-oriented calculative tools such as DCF? And how is it mobilized in narratives about streams of value, and claims about where these streams come from and where (and to whom) they should go? Discounting and its formulas provide a fruitful entry point to explore these questions. By stating that value comes from the future, DCF engages economic actors in crafting projections about events that will take place years ahead and are therefore said to be uncertain for they cannot be known in advance. By assuming that the future is worth less than the present, DCF faces economic actors with a theory of value that produces puzzling consequences. First, uncertainty appears as an external threat to DCF because the method’s orientation towards the future makes its calculations unreliable. But, secondly, uncertainty is also embedded in the formula itself. Indeed, by looking inside the formula, this chapter locates different expressions of uncertainty that entail different forms of future. More generally, this chapter argues that DCF forms the future insofar as it makes visible and gives shape to certain forms of uncertain future. In this perspective, uncertainty is no longer an external parameter, a threat faced by discounting, but a product thereof, whose characteristics can be described only through empirical analysis. The argument explored here is that the uncertain future is consubstantial with the instrument of valuation. This argument is developed through two case studies: forestry in the middle of the nineteenth century, and drug development today. Forests were among

280

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Discounting and the Making of the Future

the first non-financial assets to which the principles of discounting future costs and revenues were applied. The calculations that two German foresters and mathematicians proposed in 1849 are regarded as one of the earliest formulations of DCF, developed almost half a century before American economist Irving Fisher (1906) generalized this theory of value and expressed it in a formula that claimed to measure the value of anything, including forests (Doganova 2014). Examining the writings of these forester-mathematicians can help us to characterize the form of reasoning that DCF implies, the alternative tools with which it competes, and the effects it produces. One of these effects is precisely an upsurge in uncertainty. DCF appears in this case as a problematic form of reasoning: it produced unexpected and implausible results and its conclusions clashed with contemporary practices of forest management. Putting DCF in this historical context reveals uncertainty to be a distinctive characteristic of the particular form of future envisaged by this novel instrument. Today, DCF is firmly established in firms’ practices. The pharmaceutical industry is no exception. However, valuing drug development projects on this basis raises a number of problems, which are regularly discussed among practitioners. Drug development, commentators stress, is a long and uncertain process: it takes more than ten years to develop a new drug; the chances that the drug candidate will move successfully through all the phases of clinical and preclinical trials are low; in the early stages of a drug development process, little is known about the characteristics of the future drug, and making projections about the revenues that this drug might generate ten or more years ahead hardly makes sense. Such a situation, in which we do not know the basic categories and features of a novel product yet to be invented, echoes the type of ‘ontological uncertainty’ described by Bronk (2011). Building on analysis of the literature specializing in valuation in the life sciences (textbooks, academic and practitioner-oriented articles and reports) and on interviews conducted with managers and consultants in this field (see Doganova 2015), the second case study characterizes distinct expressions of uncertainty in DCF and discusses the forms of future that they entail. Uncertainty is treated in three ways: as risk, when measured through probabilities and averages; as an investor’s concern, when the discount rate is defined as the cost of capital (including a ‘risk premium’); and, thirdly, as the attempt to keep open a multiplicity of possibilities—a function that becomes visible when attention turns from the content of the formula to its use by managers in their everyday activities. The case study suggests that we can envisage uncertainty not only as an inherent characteristic of the future, but also as a contingent possibility of the present.

281

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

Discounting in Forest Management A Novel Theory of Value The first attempts to apply discounting to non-financial assets date back to the nineteenth century. These attempts are of interest for the argument developed in this chapter as problematic situations in which discounting encounters its objects and tries to grasp them through its theory of value. What happens when forests are being valued through the discounted future yields that they are likely to produce? This section explores this issue by delving into the writings of two German foresters and mathematicians, Edmund Franz von Gehren and Martin Faustmann, who are credited with having pioneered DCF in the field of forestry. It focuses in particular on three articles that they published in the German journal Allgemeine Forst- und Jagdzeitung in 1849. Von Gehren and Faustmann set about providing a solution to the problem of ‘determining the money value of bare forest land’ (von Gehren 1849 [1968], 19). This problem was raised by the implementation of legislation requiring that areas of forest should be converted into agriculture. What compensation should landowners receive for the property they were supposed to sell? This question entailed another one: how much is forest land worth? Von Gehren proposed the following example. Consider bare land suitable for Scots pine grown on a rotation of eighty years. The land will produce a series of yields, with thinnings every ten years and the final cut in eighty years’ time. The volume of wood thus produced can be converted into monetary units: for example, computes von Gehren, ‘in the 20th year thinnings, 200 cu. ft. of billets and faggotwood, at 5 Pf. = 1000 Pf.’. The present value of these flows of future revenues, when summed and discounted at a rate of interest of four per cent per annum, indicates the value of the plot of forest land. While von Gehren’s proposition may appear trivial, it reveals a form of reasoning and a theory of value that were novel at the time when he wrote these lines. In von Gehren’s view, the value of forest land stemmed neither from the past (for example, the efforts put into caring for the land and trees) nor from the present (for example, what is the current market price of wood?), but from the future (that is, what will the land produce if put to a certain kind of use?). Value was not something that land had, regardless of time and purpose, but something that land achieved, within a scenario of use in which it was placed in a stream of future actions consisting of specific tasks (namely thinning and cutting trees). The introduction to the English translation of von Gehren’s and Faustmann’s papers emphasizes the originality of this view: [von Gehren] saw clearly that any particular piece of ground did not have a single inherent value of its own attributable to its own unique qualities and location. . . .

282

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Discounting and the Making of the Future The value of the land is an attribute of the use to which it is put and, therefore, must be calculated from the value of the output of forest products or other crops which can be grown on it. More precisely, it is the costs and receipts incurred in growing timber that decide the forestry value of the land. (Gane 1968, 6)

In his article published in the journal’s next issue, Faustmann extended von Gehren’s proposition by applying it not only to the value of bare land, but to that of the trees standing on it: Before maturity, the stands . . . should be regarded as a product of the land which is not fully ripe, the harvesting of which causes loss to the forest owner in the same way as cutting wheat before time does to the farmer. Just think of a Scots pine stand say 10 years old, whose present market value is indisputably smaller than that which it possesses as the bearer of a future final yield. The latter is the economic value of the stand which we can express by a money capital, just like the economic value of the land. (Faustmann 1849 [1968], 32)

To the ‘market value’ of trees, derived from the price at which wood can be sold on the market now, Faustmann thus opposed an ‘economic value’ derived from the prospect of yields in the future. Similarly, he argued that the value of bare forest land should not be derived from its current market price, but from the yields that it is likely to produce in the future, divided by the interest rate to the power of the rotation length.

Uncertainty and Haste What happens to the forest, and to the forester, when value is derived from future yields, rather than past costs or current prices? A first consequence of this shift lies in the advent of uncertainty. From the very beginning of his article, von Gehren warned the reader that the valuation method he was about to propose resulted in ‘uncertainties’ and ‘absurdities’. When determined from the timber yields that could be produced on it, as he suggested, the value of bare land depends on a large number of factors: the species and type of management, whether the system is intermittent or sustained, on the length of the rotation, especially on the ratio of the wood assortments (timber, cordwood, etc.) and the money obtained from them, and finally on the method of calculating the interest. (von Gehren 1849 [1968], 19)

All these factors depend on future decisions and events and cannot be set with certainty at the moment when the calculation is made. Many different choices can be made, all resulting in a different value for the same piece of land, none of which is a priori more correct than the others. This variability—the sensitivity to assumptions used—embarrassed von Gehren. Moreover, he found that in some cases his calculation produced ‘absurdities’, such as a negative value for land. 283

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures The effect of these factors is so great that small changes in the assumptions, which are quite in the bounds of possibility, may make the land value either excessive or insignificant or even negative; such large fluctuations may arise that it is not possible to find even a mean figure which is reliable. (von Gehren 1849 [1968], 19)

These ‘uncertainties’ and ‘absurdities’ led von Gehren to abandon the valuation principle he had announced. His solution was a kind of compromise: keep the idea that the value of land should be derived from its future yields, but imagine these yields within a scenario in which the future is more regular, less fluctuating. He thus suggested that land should be valued, even if it currently carried timber, as if it was going to be converted into agriculture: from this standpoint, ‘its annual yield and, from this, its capital value can be better measured’. This scenario was indeed likely, for the very problem to which von Gehren set about providing a solution was the calculation of the compensation that should be paid to forest owners for converting forest land into agriculture. But notice how problematic and difficult this particular form of reasoning—looking to the future in order to decide about the present—was for those who invented it. A second consequence of the shift to the future is related to the form of future that discounting produces. The value of the forest is determined from its future yields, but these yields are discounted, that is, divided by the interest rate to the power of the rotation length. Hence, the greater the rotation length, the less future yields count. This is the very idea of the discount rate: there is a ‘cost of time’. A piece of land and the trees it carries are a form of capital that is ‘locked’ into a given scenario while it could have been engaged in an alternative scenario—the use of the rate of interest as the discount rate indicating that this alternative scenario is placing money at the bank—a sacrifice for which the investor (in this case the forester) should be rewarded. The immediate consequence of giving time a cost is a precipitation and haste: suddenly it appears that there is a need to cut trees earlier than previously thought, since the long term evokes a loss of value.1 This is visible when the new valuation method proposed by von Gehren and Faustmann is compared with the other methods that prevailed in forestry at that time, namely valuing forests by looking at current market prices, and managing forests so as to obtain the greatest annual income without considering that there is a cost of time. Faustmann acknowledged that the ‘market value’ of a forest, obtained by looking at current market prices, is usually higher than its ‘economic value’,

1

This effect of discounting is key in current controversies in environmental economics, which highlight tensions between the idea that the future should be discounted, on the one hand, and concern for the long term, sustainable development, and future generations, on the other.

284

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Discounting and the Making of the Future

derived from discounting future yields. In another article published in the same journal and in the same year,2 he took the example of a ‘forest of 95 Morgen (approximately 24 ha), comprising six pine stands in different site classes’, and calculated its value under two alternative management scenarios: clearcutting and selling on the market, or continuing with forestry to infinity (Viitala 2006, 138). The former would yield 26 935 florins, while the latter would yield 14 095 florins; that is, a little more than half the value. That discrepancy was a puzzling observation. Faustmann insisted on the critical role here of the choice of the discount rate, and noted the sensitivity of future cash flows to the prices of inputs and outputs in the future, which could not be known in advance. However, unlike von Gehren, he did not let such ‘absurdities’ and ‘fluctuations’ discourage him or invalidate the relevance of computing the economic value of forests. In contemporary practice, forests were managed so as to obtain not the greatest present value, calculated by discounting future yields, but the greatest annual income, that is, without introducing a discount rate supposed to compensate the investor for ‘locking’ his capital into the scenario of planting and cutting trees (Gane 1968). Faustmann claimed that his novel method was relevant not only to valuing forests, but also to managing them. If the value of land and timber can be derived from a number of parameters (future yields, the interest rate, and rotation length), then those parameters that fall within the forester’s scope of action can be fine-tuned so as to maximize the value of land and timber. In particular, Faustmann ‘presented an equation which described the present value of standing timber as a function of cutting dates and used differential calculus to derive the appropriate first- and second-order conditions for the maximum’ (Viitala 2006, 137). Faustmann thus provided an answer to a problem that had long caused discussions among foresters: the optimal age at which trees should be cut (Vatin 2008). But his conclusions were at odds with contemporary practices. Because it accounted for the ‘cost of time’, his method resulted in shorter rotation lengths. This discrepancy triggered controversies that suspended the implementation of DCF in forestry practice3 and were still live more than a century afterwards, as the issue of rotation length continued to set ‘forestry experts and the general public’ in opposition to ‘professional economists and profit-conscious businessmen’ (Samuelson 1995, 115).

2 This article is signed ‘F.’ but is attributed to Faustmann (Viitala 2006). This section relies on Viitala’s account of the article. 3 Today ‘Faustmann formula’, which expresses the value of a forest as the sum of discounted cash flows, and the related ‘Faustmann rule’, which states when a tree should be cut, are considered fundamental in forestry economics and forest management.

285

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

Forming the Future The early encounter between DCF and forests sheds light on what it means to look at the future to decide how to act in the present. It entails a problematic form of reasoning that confronts those who venture into it with a number of conundrums, including the uncertain outcomes of future courses of action, value fluctuations, and the clash between established norms and the results produced by discounting when it comes to inferring optimal courses of action. It entails, also, espousing a peculiar theory of value that focuses exclusively on the future, while defining this future from the point of view of an investor who compares alternative scenarios and thinks in terms of the opportunity cost of decisions. In line with previous works in the history of capitalism, the forestry case study discussed here suggests that looking at the future is not a natural disposition of economic actors, but a difficult form of reasoning that they only gradually learned to master. Calculative instruments such as DCF played a key role in learning to look at the future. The future is not only unknown but also invisible to the naked eye. The form it takes is that of the lens through which one looks at it. When forests are judged according to their current market value, the future fades away and it is the present that matters. When forests are instead appraised according to their potential to maximize annual income, the future takes the form of the passage of time, with each year bringing an incremental increase in the volume of timber and amount of revenue for the forest owner. But when forests are valued according to their potential to maximize the sum of discounted cash flows, the future becomes a range of alternative scenarios, which extends beyond the steady and lingering growth of trees, to include putting money in the bank (this is the meaning of the interest rate at which future cash flows are discounted) or using land for other purposes, such as agriculture—the very problem that triggered the attempts to determine the value of forest land described in this section. The future formed by discounting is a range of alternative scenarios. It is a future firmly anchored in the present insofar as the purpose of valuation is not to know what will happen (an impossible task, indeed), but to make a decision now by choosing which scenario to engage in. The economist Irving Fisher, who transformed the German foresters’ intuition into a theory of value based on the capitalization of future revenues, sealed the link between discounting and decision-making. According to Fisher (1930), discounting the future and comparing different values is what one does (or should do) when deciding the purpose to which land should be put: forestry, agriculture, or mining. As the rest of this chapter will show, the development of DCF went hand in hand with an extension of the range of alternative scenarios: putting money in the 286

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Discounting and the Making of the Future

bank or growing trees, but also buying equity, mining, farming land, operating a factory, or investing in, say, biotechnology or in information technology, and so forth—scenarios that have little in common except that they are all encompassed by the valuating gaze of the investor driven by the urge to make the right (that is, the most profitable) decision. Uncertainty thus appears as a distinctive feature of the particular forms of future that are produced through the instrument of discounting. Rather than being seen as an exogenous parameter, uncertainty can be seen as a consequence of valuing things through the various possible cash flows that they could produce for the investors who take a stake in them. Uncertainty, that is, becomes a practical problem when the future is envisaged as a range of alternative scenarios from which the investor has to choose. It matters to actors when it is organized within decision rules and models such as DCF designed to structure choice. This chapter sketches a pragmatist approach that envisages uncertainty not as a lack of knowledge entailed by distance in time, but rather as a concern that appears in certain places and at certain moments, is shaped through narratives and calculations, and engenders effects in the present. (For a pragmatist approach to the problem of valuation, see Hennion 2015.) This approach shifts attention to how uncertainty is produced, what it does, and when and for whom it matters. This will be developed by delving into another case, the use of discounting in the valuation of drug development projects.

Discounting in Drug Development Drug development is fraught with uncertainty: this is a phrase that those who study the pharmaceutical industry have often heard, and repeated. The length of the development process, the high level of innovation and failure rates, the instability of regulatory frameworks, and so forth, seem to challenge any attempt at planning the future.4 Nevertheless, discounting remains the most widespread tool for valuing drug development projects5. This section analyses how uncertainty is expressed in the discounting tools that equip managerial 4 To take one of many possible examples, the opening plenary at the last BIO-Europe meeting (a major event that gathers actors from the biotechnology industry twice a year) in March 2017 focused on the ‘mounting uncertainty [that] threatens research and innovation in medicines’ (Righetti 2017). Speakers talked, ‘on top of the usual financial and biological risks of drug development’, about uncertainties such as the possibility of regulatory reform, Brexit, cuts in research funding, Donald Trump, immigration, the ‘crisis of ignorance’, firms ‘agonizing over scenario plans around uncertainty’, political action, and resistance. 5 In a survey of valuation practices in the biotechnology industry, 85–100 per cent of interviewed managers declared that they used DCF for drug development projects that are at the stage of clinical trials; and 59–76 per cent of interviewed managers declared that they used DCF for early-stage projects (Hartmann and Hassan 2006).

287

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

decision-making in the field of drug development. It identifies three distinct expressions of uncertainty and characterizes the forms of future that they entail.

Uncertainty as Risk One way in which uncertainty is dealt with lies in the introduction of probabilities. In textbooks devoted to valuation in drug development, the following DCF formula is presented as the main tool for calculating the value of a drug development project by discounting the costs and revenues that the future drug is supposed to generate: NPV ¼

T X

pt ∗CFt ∗ð1 þ rÞt

t ¼0

According to this formula, the net present value (NPV)6 of a drug development project is equal to the sum of the cash flows (CF) that the project will produce at different points in time (t). The cash flows are reduced by a discount rate (r) due to their distance in time and the uncertainty with which they are estimated. They are further reduced by a probability (p) corresponding to the likelihood that they may not occur at all. The inclusion of probabilities is meant to take account of what textbooks call ‘technical uncertainty’: the possibility that the development project might fail because the scientific hypotheses that were to be tested do not prove valid, or the compounds under investigation turn out to be unable to go through the different phases of clinical trials. Probabilities here correspond to historical ‘success rates’. Thus, uncertainty is treated as if it were measurable risk. Conventional wisdom holds that only around ten per cent of drug development projects make it all the way from the first phase of clinical trials to approval by regulatory bodies. A growing body of academic and grey literature, which builds on the statistical analysis of past data, tends to confirm this number and provides, furthermore, success rates by phase of development and by therapeutic area (see, for example, Kola and Landis 2004). In a report published by BIO, the biotechnology industry association, one can read, for example, that, for a drug candidate in oncology, the probability of moving from Phase I to Phase II clinical trials is 62.8 per cent (BIO 2016). One might question whether reducing uncertainty to measurable probabilities in this way is not confusing uncertainty with measurable risk, and therefore missing Knight’s famous distinction. Indeed, there is no reason to 6 More precisely, this adapted version of DCF calculates what is called the risk-adjusted net present value (rNPV). The difference between standard NPV and rNPV is the inclusion of probabilities in the formula. For the sake of simplicity, and following actors’ practices, this section uses NPV to refer to rNPV.

288

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Discounting and the Making of the Future

believe that uncertainty about the technological breakthroughs and scientific advances that are likely to occur in the future could be amenable to systematic regularities that can be measured and transformed into probabilities. The ‘flaw of averages’ (Savage 2009) is not unknown to practitioners. But what is of interest for the argument here is that translating uncertainty into risk through the inclusion of probabilities is one way (be it correct or not) in which managers do in fact engage with the future. Of interest here is the form of future that is produced by the inclusion of probabilities in discounting. The probabilities considered translate the unknown characteristics of a future drug into a series of phases that may or may not lead to success. The future is thus treated as a linear process, which consists of a sequence of events that can be chronologically located and whose attributes can be known ‘on average’. The image of the pipeline, omnipresent in drug development, embodies this vision of the future: the drug candidate has to move through a number of phases—lead optimization, preclinical testing, clinical phase 1 trials, clinical phase 2 trials, clinical phase 3 trials, and approval. Average numbers, drawn from the records of past drug development projects, are associated with each phase: for example, phase I trials last between eighteen and twenty-two months and cost between 1 and 5 million dollars (Bogdan and Villiger 2007). The future then comes to be seen as no more than the reproduction of the past: the pipeline for a drug’s development is both a synthesis of what has happened so far and a projection of what will happen, using past data and patterns to project the path of a particular project that is about to begin. A similar process can be observed when it comes to estimating future sales: for early stage projects, when little is yet known about the characteristics of the future drug and its potential market, textbooks suggest resorting to averages: the average sales of a basket of comparable drugs that are currently on the market, or even the average sales of all drugs that are currently on the market. Thus, in the estimation of costs, success rates, and sales alike, the specificities of the drug under exploration are translated into broader categories (development phases, comparable drugs) and valuation is anchored to averages derived from the observation of the past (previous drug development projects) or the present (drugs currently on the market). The innovation is treated as if it would behave like the average of past innovations. One might wonder, then, whether DCF is just being used to act as if uncertainty does not exist and the future is knowable. This does not appear to be the case, as the reference to uncertainty is omnipresent in drug development. Stories about over-estimated or under-estimated sales abound in the pharmaceutical industry (a well-known example is Pfizer’s Viagra, whose indication and sales potential were unforeseen). Retrospective studies comparing sales forecasts to actual sales show that most forecasts are indeed wrong: 289

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

for example, in a study of analysts’ forecasts conducted by McKinsey, more than sixty per cent of the forecasts examined ‘were either over or under by more than forty percent of the actual peak revenues’ (Cha et al. 2013). Probabilities and averages produce one form of uncertain future—a linear, chronological future that cannot be known but can, thanks to certain assumptions, be given a present value that is good enough as a pragmatic basis for action, even though actors are aware that the predictions are most likely wrong. This future co-exists with other forms of uncertain future embedded in the instrument of discounting, which will be discussed in the rest of this chapter.

Uncertainty as the Investor’s Concern A second mechanism through which uncertainty is addressed in the DCF formula lies in changing the definition of the discount rate from the interest rate which is supposed to measure the cost of time, to a different number, called ‘the cost of capital’. This type of uncertainty is concerned not with a point probability that an event (a cash flow) will occur in the future, but with the variability of possible outcomes—that is, the possibility that the characteristics of this future event (the amount of the cash flow) might deviate from its estimated characteristics (for example, the sales of the future drug might be lower than expected): Discounting must . . . compensate not only for the loss of value over time, but also for the impending difference between the expected and the actual return. Consequently, uncertain investments should reward the investor at a higher rate than safe investments. (Bogdan and Villiger 2007, 24)

To account for the uncertainty related to the accuracy of estimates, the discount rate is increased—that is, a spread is added to the interest rate, which can range from zero up to twenty per cent in the field of drug development. Typically, the discount rates that firms use to value their projects correspond to their ‘cost of capital—that is, the rate of return that investors require in exchange for providing capital to the firm. In practice, these discount rates are calculated by firms’ financial departments.7 In some cases, they are made public: for example, according to its annual reports, Genentech, one of the first and best-known biotechnology companies, used a discount rate of 20–28 per cent in 1990 and 16–19 per cent in 1999 (Bogdan and Villiger 2007). Analysts in the biotechnology industry observe that the cost of capital ranges from twenty per cent for companies that are still at the 7

The cost of capital is calculated through a formula known as WACC, the weight-adjusted cost of capital, which averages the cost of debt and the cost of equity, that is, the returns required by the firm’s bondholders and shareholders.

290

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Discounting and the Making of the Future

discovery stage, to 8–10 per cent for companies that have a project on the market (Avance 2008).8 The difference with the case of forests is striking, in relation both to the level of the discount rate (referring to the interest rate, von Gehren and Faustmann used a discount rate of four per cent) and its very meaning. The case study on forest management showed that, in contrast to other instruments of valuation, discounting transforms the future into a range of alternative scenarios that are encompassed by the gaze of the investor. When the discount rate moves from the rate of interest to the cost of capital, the range of alternative scenarios spreads out to include all ‘other investments subject to the same degree of risk as the project under consideration . . . all the economy’s real and financial assets’ (Myers and Shyam-Sunder 1996, 209). When defined as the cost of capital, the discount rate translates the unknowable future as an investor’s concern. From the investor’s point of view, the future is something yet to be created, and for the creation of which a return, including a ‘risk premium’, is required. Moreover, the uncertainty that matters to the investor is only loosely connected to the specific drug development project in which their money is put. According to corporate finance, because investors hold projects in diversified portfolios, projectspecific risks fade away. The high cost of capital observed in pharmaceutical R&D (as compared with other industries) is explained by the observation that drug development projects face ‘a future liability’ in so far as additional investment might be required before any revenues are generated (Myers and Shyam-Sunder, 1996). For the investor, uncertainty is not the measurable probability that a given project could fail, but the worrying prospect of having to spend more money before receiving any return. Like the future expressed by probabilities and averages (which treats uncertainty as if it were risk), the future that concerns the investor—as captured in the discount rate—is not the open future, fraught with uncertainty, that is often associated with drug development and, more generally, with radical innovation (Shackle 1972). Instead, in everyday practice firms calculate as if the cost of capital were pricing in the ontological uncertainty associated with an innovation. While defined as the reward that investors need to receive in compensation for the risks they are taking by engaging their money in a given project, the cost of capital is in fact determined by the

8 Research in corporate finance, which uses publicly available data and the capital asset pricing model (CAPM) to infer pharmaceutical companies’ cost of capital has found similar levels. Myers and Shyam-Sunder (1996), for example, found that in 1990 cost of capital was, on average, fifteen per cent for major pharmaceutical companies and nineteen per cent for smaller biotechnology companies. Harrington (2012) found slightly lower numbers for the periods 2001–5 and 2006–8: respectively, ten per cent and nine per cent for pharmaceutical companies, and fourteen per cent and twelve per cent for biotechnology companies.

291

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

firm’s financial department, following the models and formulas of corporate finance. For the most part, the cost of capital is defined at the firm level, and project-specific discount rates are rarely calculated.

Uncertainty as the Multiplicity of Possibilities A third expression of uncertainty, radically different from the two expressions discussed so far, appears when one looks at how the DCF formula is used by managers in their everyday activities.9 What managers actually use is not a mathematical formula like the one reproduced on page 288, but Excel sheets and/or specialized software in which the valuation techniques, namely DCF, are built in. In an Excel sheet, discounting appears as a table, whose columns represent time, organized as a series of years grouped by the different stages of drug development (for example, preclinical, phase I, phase II, phase II, market), and whose rows represent yearly estimated cash flows (costs and revenues) and their probabilities. The Excel format allows managers to combine data coming from various sources, and manipulate the data in order to explore different scenarios. The aim of specialized valuation software is to make these two operations even easier. The software is usually accompanied by a database including success rates and averages on sales and costs categorized by diseases and phases. The interface allows the user to enter the input parameters, and then to add or remove time, increase or decrease the discount rate, correct the probabilities, change the estimates of sales and costs, and so forth, in order to observe how the value of a drug development project varies accordingly. Discounting thus produces a range of values, tied to different scenarios of drug development and marketing. What one observes, then, when looking at how discounting is mobilized in valuation practices, is not the pursuit of a value that would be more truthful than other values—that is, the pursuit of a forecast that would bear a greater resemblance to what will actually happen in the future. What one observes, instead, is managers experimenting with different assumptions and scenarios. DCF does not serve to indicate the value of a future drug, but to produce a multiplicity of possible values and possible outcomes. What kind of therapeutic indications will the drug aim at? Will it be above or below average? What kind of clinical trials will be necessary to know this? How does this affect costs and sales? And what if another indication was aimed at? And so on, and so forth. Imagining and estimating the nature of a future drug, and valuing it, are intertwined in the very same process. This process can be analysed as the concomitant exploration of ontologies and values. 9

292

This section draws on empirical material and analysis presented in Doganova (2015).

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Discounting and the Making of the Future

Another form of the uncertain future thus appears. The future is no longer something that will happen but cannot be known, uncertainty stemming from the divergence between what Luhmann (1976) called the ‘present future’ and the ‘future present’. The future is something that emerges from the process of valuation and through material operations such as discounting in Excel. It takes the form of a range of possible future entities and their values. Uncertainty then hinges on the multiplicity of these possibilities and on the indeterminacy of chosen assumptions and scenarios. An interesting illustration can be found in the following observation that a manager from a pharmaceutical company expressed during an interview.10 He observed that since discounting grants little value to exploratory projects that are in the very early stages of development (because such projects entail higher risk and the positive cash flows that they are likely to generate are distant in time), it paradoxically keeps these projects in the state of indeterminacy needed for exploration to proceed: If the investment isn’t that high, then you don’t need . . . to say exactly what it [the drug] will reach on the market, you just need to be able to justify that it will make sense: there is potential for it scientifically, it is interesting, risk-wise we haven’t identified anything out of the way, so we have to spring it into the portfolio and explore.

To recap, in firms’ practices, discounting produces uncertainty in at least two ways. On the one hand, embedded in valuation software that automates calculation, discounting is used as a means to navigate, back and forth, from assumptions to values, and from the future to the present, thereby producing multiple possibilities. Managers seem to be concerned not so much with what will happen, say, ten years from now (for example, how much sales the drug will achieve), but with making sure that various options have been considered, analysed, and ruled out or kept in. On the other hand, by positing that the future is worth less, discounting somehow loosens the need to know that future, and focuses attention on the ongoing process of making the future. Envisaging uncertainty in such a way—that is, as the contingent product of local experimental practices equipped with discounting calculations—urges us to reverse the question this chapter started from. What matters here is not to reduce the uncertainty of the future by making it somehow knowable (albeit through the imperfect means of probabilities and averages). What matters, rather, is to maintain the uncertainty of the present, by making multiple options visible and debatable.

10

See Doganova (2015).

293

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

Conclusion Knight’s distinction between risk and uncertainty has been widely discussed in economics and sociology. Much attention has been focused on the cognitive consequences of uncertainty, namely the shift from rational calculation to judgement and opinion in decision-making processes. Much less attention has been paid to two other consequences of uncertainty that Knight evoked. The first one we can call ‘political’, for it relates to the distribution of wealth among different categories of people, namely the entrepreneur and other economic actors.11 Uncertainty, Knight noted, ‘gives the characteristic forms of “enterprise” to economic organization as a whole and accounts for the peculiar income of the entrepreneur’ (p. 232). A second consequence is related to the social organization of markets. Uncertainty, Knight observed, is the raison d’être of a whole industry engaged in the production of information and instructions for the guidance of managers’ conduct, triggering ‘a veritable swarming of experts and consultants in nearly every department in industrial life’ (p. 262). This chapter’s analysis of the use of DCF in forest management and drug development resonates with Knight’s observations, but also takes them in a different direction. Uncertainty has, indeed, cognitive, political, and social consequences. It is intertwined with the figure of the investor and with associated claims about the rewards that are embedded in the cost of capital that firms use as a discount rate when they value investment projects. It is treated with probabilities and averages, derived from the observation of past activities and of the present state of markets, which are produced by valuation experts, consultants, or academics. The possibility of calculating the value of a drug development project by discounting future costs and revenues hinges on the availability of data and on the existence of a whole industry engaged in the production of knowledge about the future through the production of knowledge about the past and the present. This, in turn, hinges on the willingness of practitioners to espouse a vision of the future as reproducing the past, while at the same time acknowledging that predictions are doomed to fail and that calculated figures could not be correct representations of future events. Taking Knight’s argument in this novel direction, inspired by a pragmatist approach to valuation, entails questioning the relevance of his distinction

11 It should be noted that Knight’s point was not to discuss the issue of the distribution of wealth, but to explain how profits can persist in a free market. It is this chapter’s reading of Knight’s conclusions that suggests their political dimension. Moreover, while Knight was concerned with the figure of the entrepreneur (and their ‘peculiar income’), the discounting techniques discussed in this chapter point to a different figure, which is that of the investor (and the ‘rewards’ they should receive in compensation for uncertainty). On the entanglement between these two figures in processes of capitalization, see Doganova and Muniesa (2015).

294

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Discounting and the Making of the Future

between uncertainty and risk. What happens when uncertainty is treated not as an analytical category but as an object of analysis amenable to empirical description? Through such a lens, uncertainty appears not merely as an inherent characteristic of the future, which economic actors inevitably face when they make decisions, and which hinders their calculations. In the case of forest management, uncertainty is the result of a historically situated and problematic encounter between forests and an instrument of valuation that brought a novel vision of the future as a range of alternative scenarios encompassed by the gaze of the investor. In the case of drug development, uncertainty is multifaceted and contained in the instrument of valuation, simultaneously expressed as a risk, an investor’s concern, and the multiplicity of possibilities. Discounting is an instrument of valuation that forms the uncertain future in a variety of ways: a linear process that is unknown but can be rendered knowable on average if considered as a replication of the past; an infinite range of alternative scenarios encompassed by the gaze of the investor; or a locally engendered and contingent product of valuation processes and material operations.

Bibliography Avance. 2008. ‘Discount Rates for Biotech Companies’. News in Avance: Valuation in Life Sciences, January 2008, no 1. http://www.avance.ch/newsletter/docs/discount_1.pdf. Beckert, Jens. 2016. Imagined Futures: Fictional Expectations and Capitalist Dynamics. Cambridge, MA: Harvard University Press. BIO (Biotechnology Innovation Organization). 2016. Clinical Development Success Rates 2006–2015, June 2016. https://www.bio.org/sites/default/files/Clinical%20Develop ment%20Success%20Rates%202006-2015%20-%20BIO,%20Biomedtracker,% 20Amplion%202016.pdf. Birch, Kean. 2016. ‘Rethinking Value in the Bio-Economy: Finance, Assetization, and the Management of Value’. Science, Technology, & Human Values 42 (3): pp. 460–90. Bogdan, Boris, and Ralph Villiger. 2007. Valuation in Life Sciences: A Practical Guide. Berlin: Springer. Bourdieu, Pierre. 1963. ‘La société traditionnelle: Attitude à l’égard du temps et conduite économique’. Sociologie du Travail 1: pp. 24–44. Bronk, Richard. 2011. ‘Epistemological Difficulties with Neo-Classical Economics’. Southern Economic Association Conference Paper, 19–21 November, Washington D.C. http://eprints.lse.ac.uk/39423/. Cha, Myoung, Bassel Rifai, and Pasha Sarraf. 2013. ‘Pharmaceutical Forecasting: Throwing Darts?’. Nature Reviews Drug Discovery 12: pp. 737–8. Doganova, Liliana. 2014. ‘Décompter le futur: la formule des flux actualisés et le manager-investisseur’. Sociétés Contemporaines 93: pp. 67–87.

295

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures Doganova, Liliana. 2015. ‘Que vaut une molécule? Formulation de la valeur dans les projets de développement de nouveaux médicaments’. Revue d’Anthropologie des Connaissances 9 (1): pp. 17–38. Doganova, Liliana, and Fabian Muniesa. 2015. ‘Capitalization Devices: Business Models and the Renewal of Markets’. In Making Things Valuable, edited by Martin Kornberger, Lise Justesen, Anders Koed Madsen, and Jens Mouritsen, pp. 109–25. Oxford: Oxford University Press. Fisher, Irving. 1906. The Nature of Capital and Income. New York: Macmillan. Fisher, Irving. 1930. The Theory of Interest, as Determined by Impatience To Spend Income and Opportunity To Invest It. New York: Macmillan. Freeman, Christopher. 1986. The Economics of Industrial Innovation. 2nd edn. Cambridge, MA: MIT Press. Faustmann, Martin. 1849 [1968]. ‘Calculation of the Value which Forest Land and Immature Stands Possess for Forestry’. In Martin Faustmann and the Evolution of Discounted Cash Flow: Two Articles from the Original German of 1849, edited by Michael Gane, pp. 27–55. Institute Paper, no 42, Commonwealth Forestry Institute, University of Oxford. Gane, Michael. 1968. ‘Introduction’. In Martin Faustmann and the Evolution of Discounted Cash Flow: Two Articles from the Original German of 1849, edited by Michael Gane, pp. 5–18. Institute Paper, no 42, Commonwealth Forestry Institute, University of Oxford. Harrington, Scott E. 2012. ‘Cost of Capital for Pharmaceutical, Biotechnology, and Medical Device Firms’. In The Oxford Handbook of the Economics of the Biopharmaceutical Industry, edited by Patricia M. Danzon and Sean Nicholson, pp. 75–99. New York: Oxford University Press. Hartmann, Marcus, and Ali Hassan. 2006. ‘Application of Real Options Analysis for Pharmaceutical R&D Project Valuation: Empirical Results from a Survey’. Research Policy 35 (3): pp. 343–54. Hennion, Antoine. 2015. ‘Paying Attention: What is Tasting Wine about?’. In Moments of Valuation: Exploring Sites of Dissonance, edited by Ariane Berthoin Antal, Michael Hutter, and David Stark, pp. 37–56, Oxford: Oxford University Press. Karpik, Lucien. 1989. ‘L’économie de la qualité’. Revue française de sociologie 30 (2): pp. 187–210. Karlsson, Tomas, and Benson Honig. 2009. ‘Judging a Business by its Cover: An Institutional Perspective on New Ventures and the Business Plan’. Journal of Business Venturing 24 (1): pp. 27–45. Keynes, John Maynard. 1936. The General Theory of Employment, Interest and Money. London: Palgrave Macmillan. Knight, Frank. 1921. Risk, Uncertainty, and Profit. Boston, MA: Houghton Mifflin Company. Kola, Ismail, and John Landis. 2004. ‘Can the Pharmaceutical Industry Reduce Attrition Rates?’. Nature Reviews Drug Discovery 3: pp. 711–16. Luhmann, Niklas. 1976. ‘The Future Cannot Begin: Temporal Structures in Modern Society’. Social Research 43 (1): pp. 130–52.

296

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Discounting and the Making of the Future Muniesa, Fabian. 2016. ‘Setting the Habit of Capitalization: The Pedagogy of Earning Power at the Harvard Business School, 1920–1940’. Historical Social Research 41 (2): pp. 196–217. Muniesa, Fabian, Liliana Doganova, Horacio Ortiz, Alvaro Pina-Stranger, Florence Paterson, Alaric Bourgoin, Véra Ehrenstein, Pierre-André Juven, David Pontille, Başak Saraç-Lesavre, and Guillaume Yon. 2016. Capitalization: A Cultural Guide. Paris: Presses des Mines. Myers, Stewart C., and Lakshmi Shyam-Sunder. 1996. ‘Measuring Pharmaceutical Industry Risk and the Cost of Capital’. In Competitive Strategies in the Pharmaceutical Industry, edited by Robert B. Helms, pp. 208–37. Washington, D.C.: American Enterprise Institute Press. Pollard, Sidney. 1965. The Genesis of Modern Management: A Study of the Industrial Revolution in Great Britain. London: Edward Arnold. Power, Michael. 2007. Organized Uncertainty: Designing a World of Risk Management. Oxford: Oxford University Press. Righetti, Erin. 2017. ‘Mounting Uncertainty Threatens Research and Innovation in Medicines’. http://insight.ebdgroup.com/featured-content/mounting-uncertaintythreatens-research-and-innovation-in-medicines. Samuelson, Paul Anthony. 1995. ‘Economics of Forestry in an Evolving Society’. Journal of Forest Economics 1 (1): pp. 115–49. Savage, Sam L. 2009. The Flaw of Averages: Why We Underestimate Risk in the Face of Uncertainty. Hoboken, NJ: Wiley. Shackle, George. 1949. ‘Probability and Uncertainty’. Metroeconomica 1 (3): pp. 161–73. Shackle, George. 1972 [1992]. Epistemics and Economics: A Critique of Economic Doctrines. New Brunswick, NJ: Transaction Publishers. Vatin, François. 2008. ‘L’esprit d’ingénieur: pensée calculatoire et éthique économique’. Revue Française de Socio-Économie 1 (1): pp. 131–52. Viitala, Esa-Jussi. 2006. ‘An Early Contribution of Martin Faustmann to Natural Resource Economics’. Journal of Forest Economics 12 (2): pp. 131–44. Von Gehren, Edmund Franz. 1849 [1968]. ‘On the Determination of the Money Value of Bare Forest Land’. In Martin Faustmann and the Evolution of Discounted Cash Flow: Two Articles from the Original German of 1849, edited by Michael Gane, pp. 19–26. Institute Paper, no 42, Commonwealth Forestry Institute, University of Oxford.

297

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

14 The Dilemma between Aligned Expectations and Diversity in Innovation Evidence from Early Energy Technology Policies Timur Ergen

Introduction There are few domains in which capitalist societies’ relationship with the future has been as elaborately debated as technological innovation. How can organizations, states, and societies prepare for and shape the impact of technologies whose future is highly uncertain or whose very nature is as yet unclear? And how can they stimulate and direct the development of practices, products, and services they do not yet know about? One of the main structuring axes of debates about technological innovation is the question of how to allocate resources, organizational structures, and institutional supports between the improvement of existing technologies and work on those that are not yet known, fully developed, or commercialized. Decisions ranging in scale from small-firm research and development projects through to the general legitimacy of state support for basic research are shaped by this trade-off. While widely varying proposals for desirable allocations of corporate and public resources to the two activities have been presented for a long time now, few thinkers have doubted in principle the desirability of dedicated, yet undirected, search efforts for new technologies.1 In the specific way they relate to the future, such institutional conditions for technological innovation are 1 Insightful historical peaks in debates about the desirability of such dedicated search efforts can be found in Kleinman’s (1995) history of the struggle to establish the National Science Foundation and in Hoddeson’s (1981) and Mowery and Rosenberg’s (1989) accounts of the emergence of industrial research in the United States and in the United Kingdom.

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Expectations and Diversity in Innovation

closely related to Beckert’s and Bronk’s argument in the introduction of this volume to the effect that the ‘negative capability’ of remaining ‘in uncertainties, mysteries, doubts’ about the future is key to creative economic action. Richard Lester and Michael Piore (2004) have developed a useful typology for situating such explorative, open-ended search efforts. They distinguish between two major types of activity in design and development: analysis and interpretation. The former—prominent in engineering, economics, and political discourse—consists of focused work to design efficient solutions to well-defined technological problems. The latter—demonstrably essential for innovation, yet rarely articulated in best-practice thinking about technological development—consists of exploratory processes that identify technological problems, needs, and possibilities in the first place. Interpretation is open-ended and, in its structure, resembles everyday conversations: The way that new designs came to be initiated [by the subjects of our case studies, TE], the way that new styles emerged or trends in style were ‘recognized,’ the way that problems came to be identified and clarified to the point where a solution could be discussed was through conversations among people from different backgrounds and with different perspectives. Communication during this conversational phase is often punctuated by misunderstandings or ambiguities; indeed, an accepted vocabulary to describe the new product may not even exist. Yet this ambiguity in the conversation is the resource out of which new ideas emerge. And something is lost if that conversation is closed off too soon. (Lester and Piore 2004, 51)

Lester’s and Piore’s ideas about ‘closing off ’ conversations ‘too soon’ should not be mistaken for an argument in favour of radically free-market innovation policies. In fact, intense economic competition is, for them, the social basis of efficiency-enhancing analysis rather than of boundary-crossing interpretation, since competition reinforces instrumental rationality, secrecy, and focused business organization. They rather support the maintenance of developmental communities without clear-cut design objectives, probably best exemplified by the organization of the corporate laboratories of giant US firms during the three decades following World War II. The call for a renewed appreciation of such communities permeates many recent statements on technology and industrial policy (Block 2008; Piore 2008; Rodrik 2004; Schrank and Whitford 2009). This literature builds on the argument that today’s industrial reality, marked by fractured supply chains and accelerated technological change, is served neither by hierarchical models of industrial policy nor by free-market policies. Instead, it thrives on the basis of state activity or other non-market forms of coordination that facilitate, in Lester’s and Piore’s words, continued conversations within and across firms and industrial sectors. There is much to learn from and agree with in these calls for updated innovation policies. Most importantly, they spell out that ‘creativity’ in the 299

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

economy has a concrete organizational and institutional basis that requires continuous political and corporate nurturing. Nevertheless, they also replicate a problematic division between ‘pre-market’ research and development that creates technological options and the later-stage domain of business that exploits these options and eventually ‘picks winners’. Commercialization, production, and marketing are treated as unproblematic features of the innovation process susceptible to focused engineering and entrepreneurial exploitation. Moreover, in many emerging technological fields—and particularly in manufacturing—interpretation and analysis stand in a more interdependent relationship than the one described by Lester and Piore. In many cases, the general potential and use-value of technologies is accessible only during or after industrial upscaling and feedback from marketing and usage—post-analysis, so to say. Indeed, collective industrial upscaling is often the precondition for creative development: by attracting talent, resources, and supporters; by confronting innovators with buyers and users; and by giving them the possibility for trial-and-error processes and serendipitous discoveries. If it is true that certain technologies do indeed require this collective leap of faith to become the ground for collective explorations of ambiguity, but at the same time the early foreclosure of technological options needed for this collective action weakens the undirected search movements that Lester and Piore describe, then there exists an unavoidable and persistent dilemma between the merits of openness and of closure of technological development and policy options. It is this dilemma that this chapter aims to illuminate. It argues that interpretation and analysis remain in ‘perpetual tension’ (Lester and Piore 2004, 121) because both contribute to innovation, and yet thrive in conflicting social conditions. The commercialization of new technologies depends on analysis and is dependent on continuous support from manifold actors and organizations. But such support structures are difficult to maintain without a certain coordination of expectations and commitment. At the same time, aligned expectations and focused work undermine the diverse exploration activities that are also crucial to technological innovation. The chapter brings together arguments from economics, sociology, and political economy to show that innovation processes are characterized by this dilemma between the advantages of aligned expectations and those of diversity. To illustrate the argument, the chapter discusses a historical case involving one of the largest coordinated peace-time attempts to hasten technological innovation in the history of capitalism, namely the US energy technology policies of the 1970s and 1980s. At a time of increasing uncertainty about future resource supplies and the future direction of societal development, the state, industry, and activists experimented with large-scale 300

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Expectations and Diversity in Innovation

support programmes for virtually every energy technology known at the time. Closer examination of the commercialization of photovoltaics and synthetic fuels serves to demonstrate both sides of the dilemma between diverse and shared expectations in innovation: openness but possible stagnation, on one hand, and cohesion but possible premature lock-in, on the other. The remainder of the chapter is divided into three parts. Firstly, it develops the conceptual argument about the troubled relationship between cohesion and exploration. This argument is then illuminated with a closer look at two technology policies that failed resoundingly: solar photovoltaics and synthetic fuels in the 1970s and 1980s. The conclusion connects the argument to broader questions of sociological research on expectations in the economy.

Technological Innovation, Organized Diversity, and the Alignment of Expectations Lester’s and Piore’s distinction between analysis and interpretation mirrors other well-known distinctions in business, economic, and organizational studies, including exploitation versus exploration (March 1991), ‘administration of existing structures’ versus creation of new ones (Schumpeter 1942 [1975]), decision-making under conditions of ‘risk’ versus ‘uncertainty’ (Knight 1921), or habitual reaction and innovation (Winter 1971). Equally popular has been the conceptualization of these activities as two sides of a resource conflict. Entrepreneurs, firms, and, for some authors, political economies as a whole, may commit resources either towards calculable short-term improvements of their existing business models, products, and technologies or towards innovative and uncertain long-term ventures. Besides standing in a trade-off with regard to resources, the two types of activity are characterized by contradictory relationships with the future. While calculable short-term improvements thrive on the basis of firmly anchored and broadly aligned expectations about coming technological pathways, long-term innovation benefits from a diversity of outlooks and an openness about the future. While this contradiction does not cause problems for the organization of technological development in linear paths of innovation, in which a phase of open-ended search gives way to a phase of focused commercialization (as discussed by Lester and Piore 2004, 100), it does give rise to a genuine dilemma in situations in which innovative design and development build on a degree of anticipatory lock-in. In these cases, aligned expectations about future pathways enable technological development but weaken open-ended search, while the alternative of maintaining diversity threatens to stymie developmental activities at an early stage. 301

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

This section discusses two lines of research in the literature that highlight the two sides of the dilemma between organized diversity and alignment of expectations. The first discusses the organizational, institutional, and social conditions that support exploratory activities, while the second demonstrates that a certain degree of collective lock-in is a precondition of complex processes of technological development. Both lines of research overlook the other’s main point and thereby fail, as this chapter argues, to account for the dilemma between diversity and aligned expectations. Recent conceptual work on the organizational, institutional, and social conditions of creative development challenges a previously long-standing line of reasoning in the sociology of innovation. With few exceptions—and in line with decades of sociological theories of social order—researchers sought until recently to discover the social conditions of overcoming entropy and uncertainty in innovation processes, and asked the following question: how and when can actors overcome the dissipation and uncertainty endemic in new ventures? Creativity and the discovery of innovative technological possibilities, by contrast, were often explained with reference to deviant individuals or organizations. One important recent conceptual reflection on the social conditions of creative development is that of David Stark (2009). Building on John Dewey, Stark argues that uncertainty, friction, and dissonance—or what he calls ‘perplexing situations’—play a productive role in economic action by fostering specific kinds of reflective search and rethinking of routines: We sense that there is a difference between occasions when we look for solutions within a set of established parameters and other occasions . . . rife with uncertainty and yet, precisely because of that, also ripe with possibilities . . . Stated as recognition of the incognita, the process of innovation is paradoxical, for it involves a curious cognitive function of recognizing what is not yet formulated as a category. It is one thing to recognize an already-identified pattern, but quite another to make a new association. (Stark 2009, 2, 4)

Based on that idea, Stark demonstrates that focused management, homogenous cognitive foci, and early top-down control stifle innovative discovery and that organizational forms that allow for conflicting foci, ambiguity, and uncertainty can prevent lock-in into routine ways of development. At the level of industries and sectors, Lester and Piore (2004) spell out why this may be the case. In their studies of development and design in garments, medical devices, and mobile telecommunication, they find that innovation depends on continued exchange across functional and organizational boundaries. ‘Interpretive processes’, they observe, are particularly vulnerable to . . . pressures [of organizational rationalization, TE] in the early stages of product development, before a rich language for exploring

302

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Expectations and Diversity in Innovation ambiguity has fully developed. At that early stage, genuine ambiguity is not easily isolated from simple confusion and misunderstanding, and the conversation is fragile and easily abandoned. (Lester and Piore 2004, 176)

Lester and Piore call for ‘sheltered spaces’ in universities and corporate laboratories or via various government instruments that intentionally prevent earlystage focus and compartmentalized development and keep conversations between different functions and organizations going. A similar point has been made in historical research on the structure of the US innovation system. Fred Block (2008), for example, argues for a reorientation of the debate on industrial policy from the targeting and nurturing of sectors and specific technologies by the state bureaucracy towards an assemblage of decentralized policies that he calls the Developmental Network State (see also Ó Riain 2004). Since the 1980s, he argues, advanced capitalist states have been institutionalizing structures that facilitate technological development around vaguely specified goals by providing funding for very early-stage ventures and by nurturing collaboration and the spread of information between firms, scientists, engineers, and state agents. As the challenge in today’s global economy is to promote and develop ‘product and process innovations that do not yet exist’ (Block 2008, 172), rather than to develop domestic counterparts to internationally leading firms or technologies, industrial policy has increasingly less to do with ‘picking winners’ and more with ‘making winners’ through education policies, the spread of information, and network activities (Ó Riain 2004, 98–105; Rodrik 2004; Schrank and Whitford 2009). The argument for diversity-enhancing policy designs has recently been generalized by Richard Bronk and Wade Jacoby (2016). Criticizing regulatory harmonization efforts, they argue that convergence on single solutions to regulatory problems in uncertain environments can be dangerous because of three processes. First, convergence might turn out to be premature if there are unforeseen changes in the environment. Communities would then be stripped of possible institutional building blocks to respond to new challenges. Second, institutional convergence can lead to cognitive convergence, undermining capacities for creative rethinking and recombination in the future. Third, the crafting of solutions to perceived problems must be understood as an ongoing process of discovery or trial and error. In dynamic environments, a certain degree of organizational redundancy or slack (Grabher 1994; March and Simon 1958) can outweigh short-term static efficiency losses by allowing for gains in dynamic adaptability. This chapter maintains that such arguments in praise of diversity and against early-stage lock-in are incomplete for two, related reasons. First, in many industrial fields, the division between early-stage innovation, which 303

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

thrives on uncertainty and ambiguity, and later-stage commercialization and production, which thrives on rational exploitation, is misleading. Second— and this is where a sociological view of the formation of expectations becomes relevant—the maintenance of heterogeneous developmental communities is difficult without a certain degree of cognitive, organizational, and institutional cohesion. The flawed nature of the analytical distinction between innovation and production has been highlighted by Mark Blaug (1962 [1990]). Admittedly writing at a time when the dominant innovative sector of advanced economies was manufacturing, he made the following observation from business history: The vital difference for an individual firm is not between known and unknown but between tried and untried methods of production. The convention of putting all available technical knowledge in one box called ‘production functions’ and all advances in knowledge in another box called ‘innovations’ has no simple counterpart in the real world, where most innovations are ‘embodied’ in new capital goods, so that firms move down production functions and shift them at one and the same time. (Blaug 1962 [1990], 704)

This observation is not limited to process innovations; nor is the relationship between production and innovation one of mere co-occurrence. As variously formalized in theories of ‘learning curves’, in many technological fields, firms can only shift production functions if they move down them, extend production, recoup resources, and learn in the process of production and marketing (see, for example, Rosenberg 1982, chapters 5–7). As documented repeatedly in the history of technology, it is often only in the process of commercialization and marketing that producers learn—through real-world use—about the actual categories their products might occupy and their various qualities (see, for example, Schwartz Cowan 1987). In their study of modern wind turbine development in the United States and in Denmark, for example, Garud and Karnøe (2003) show how it was incremental development and upscaling that over time led to breakthrough improvements in the technology, which redefined its use-cases, and ultimately changed which societal values it was able to cater to. Similar experience-curve logics exist in technology policy. As shown by Nick Ziegler (1997), the growth of technology- and industry-specific expertise in networks between state agents, firms, scientists, and stakeholders is a long-term process that is essential for the appropriate design and implementation of technology policies. It does not follow from this that innovations come naturally with the extension of production, or that the positions in praise of diversity discussed in the previous paragraphs are wrong. Rather, these observations challenge the notion that firms or states operating in fields in which innovation is 304

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Expectations and Diversity in Innovation

intimately related to production can even come close to certainty about the future of novel products before committing significant resources to them. The problem arises of how communities become bound to such technological paths—how they risk lock-in—before they can reasonably know how things will turn out. Put differently, the question is one of coalition- and institutionbuilding. This, in essence, is what Peter Evans described in rarely-cited passages of his seminal comparative study of IT-industry policies. Besides getting state-industry relations right, the ‘key to facilitating the growth of a new sector’, he observed, ‘was . . . creating the conditions that led entrepreneurial groups to identify their interests with the growth of the sector and commit resources to it’ (Evans 1995, 210). This problem has been discussed extensively in institutionalist policy analysis. Margaret Weir, for example, demonstrates how heterogeneous groups can have trouble unifying behind common causes over longer periods of time, because all of them have multiple alternative pathways to pursue their goals (Weir, 2006; Weir et al. 2009). Unity, seen in this way, has the character of a classic coordination problem. A good recent example is mass-market electric cars. The realization of affordable electric cars depends on the decades-long interlocking development efforts of hundreds of firms and institutions, each of which has numerous alternative development opportunities to pursue. For example, battery manufacturers could focus instead on improving batteries for other uses, while auto manufacturers could turn to the development of what used to be called ‘clean diesel’ engine technology. (The problem of complementary innovation is analysed by Gawer 2000.) Recent sociological research on early-stage technological development has spelled out how to conceptualize the emergence of such developmental coalitions. They can be understood as emerging based on shared imaginaries of the future, or, to put it differently, on a degree of cognitive lock-in with respect to expectations (Beckert 2016, chapter 7). The richest case-based analyses of these processes have been produced by scholars working in the tradition of science and technology studies. As described in a study by Harro van Lente of the emergence of the field of membrane technology, ‘projections, expectations and scenarios’ can create ‘prospective’ social structures that are ‘forms of coordination which can occur without commitment of actors to a shared project, while their outcome, even if not necessarily consensual, is to make the new scientific-technological field a “going concern” ’ (van Lente and Rip 1998, 224). Under the promissory ‘umbrella term’ of membrane technology, a field of supporting structures solidified that allowed scientists, state agents, and businesses to work on a wide array of cross-disciplinary technological problems from the 1960s. Cognitive lock-in with regard to images of the future does not, of course, preclude further creative development within emerging fields; nor are shared 305

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

images of the future as a rule self-fulfilling, as the large number of instances in which technological visions failed to materialize proves. Still, they often do diminish the chances of alternative technological pathways by altering the distribution of resources and by breaking apart coalitions for possible further options. To come back to the example of electric cars, a large-scale shift towards fuel cell systems or combustion engines running on renewable fuels, which at the end of the 1990s were arguably not severely disadvantaged vis-à-vis a large-scale shift towards electric cars, has become significantly more unlikely with the emergence of momentum in favour of the latter during the first two decades of the twenty-first century. More importantly, perhaps, with respect to the ideas discussed here, this entails a shift of a significant amount of advocacy, talent, and resources from possible activities searching for as yet unknown technological options towards activities focused on a single promise. It is this necessary sacrificing of openness and diversity for learning and cohesion that poses a dilemma in technological development and technology policy.

US Energy Technology Policies in the 1970s and 1980s The conflict between the coordination benefits of aligned expectations and the benefits of diversity in innovative fields is not merely a structural feature in technological development; it can also become the basis of factional conflicts when it comes to questions of how to distribute resources in organizations and across societies. As demonstrated in this section, development efforts can oscillate between focused development and open conversations, reflecting both the benefits and costs described earlier in this chapter. American energy technology policies in the 1970s and 1980s are a case particularly well-suited for studying the conflicts of different models of technological development. Initially started by an initiative of the federal government, an endless variety of search movements for technological solutions to the energy crisis of the 1970s emerged, ranging from local activistdriven attempts at technical tinkering with energy self-sufficiency through to Big-Science programmes led by the American military-industrial complex. This section briefly describes the origins and overall structure of these technology policies, before ‘zooming in’ on two failed initiatives: the attempts to commercialize solar cells and to create a mass-market for synthetic fuels. Peace-time government support for energy technologies was not new to the 1970s. However, the first oil crisis of 1973/1974 gave rise to a degree of focused state support unheard of in previous decades. Political engagement to hasten innovation in energy, rather than to combat the energy crisis with regulatory or

306

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Expectations and Diversity in Innovation

diplomatic instruments can be explained in part by the toughness of the challenge itself. It was also caused by political stalemate in the energy arena, as well as in the tumultuous 1970s more generally (Ikenberry 1988; Kitschelt 1983). In comparison with tax policies, liberalization measures, price adjustments, and the establishment of new regulatory constraints, technology policies had few immediate distributional losers. The conservative administrations of the first half of the 1970s, not to speak of Jimmy Carter’s speeches, escalated their rhetoric about the severity and permanence of the oil shortages from the early 1970s. In November 1973, Richard Nixon, who for years had issued scattered warnings about coming fuel shortages, formulated what became—at least in theory—the elusive goal of US energy technology policies for the coming decades: Today the challenge is to regain the strength that we had earlier in this century, the strength of self-sufficiency. Our ability to meet our own energy needs is directly limited to our continued ability to act decisively and independently at home and abroad in the service of peace, not only for America but for all nations in the world. . . . Let us set as our national goal, in the spirit of Apollo, with the determination of the Manhattan Project, that by the end of this decade we will have developed the potential to meet our own energy needs without depending on any foreign energy sources. (Nixon 1973)

Despite growing forces in the Nixon, Ford, and Carter administrations that championed a broad liberalization initiative in the energy sector to combat shortages with a kind of shock therapy through price rises (Jacobs 2016), energy independence in the 1970s was supposed to be achieved by numerous initiatives to raise domestic fuel exploitation, as well as a large-scale attempt to hasten the development of ‘technological fixes’. The reason for the prominence of the technological medium- and long-term lay in the growing conviction among large sections of society and experts that shortages and price-hikes for energy were just a prelude to coming extreme turbulence in the energy sector, due ultimately to the scarcity of global reserves. In what were in part chaotic political battles after the OPEC Oil Embargo, competencies for energy technology policy were centralized in a giant newly created federal agency called the Energy Research and Development Administration (ERDA) in 1974. At the beginning, ERDA was staffed with 7200 direct employees, mainly seconded from the non-regulatory and non-military parts of the Atomic Energy Commission and various institutions in support of mining and exploration. It had an initial budget of $3.6 billion. ‘ERDA’s job is to throw money at the Energy Crisis’, a contemporary journalistic account concluded about the new agency (Alexander 1976). Over the following years, the administration, the research and business communities, and

307

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

Congress massively expanded the technological options that could fulfil the promise of the (continuously delayed and relativized) energy independence the ERDA was meant to support. These ranged from experiments with photosynthesis and waste-recycling through to geothermal power plants and nuclear fusion. Table 14.1 gives an overview of ERDA’s programme funds for the years of its full operation, before it was eventually merged into the newly created Department of Energy (DOE.) in 1977. Figure 14.1 presents funding levels for different technological paths for a longer time-period and helps to put the expansion of energy technology support in the 1970s into perspective. Despite growing pressures on the research and development complex to come up with technological breakthroughs that would lessen the economic, environmental, national security-related, and societal strains of the 1970s, ERDA and hundreds of related programmes and laboratories maintained stable structures for technological experimentation. Indeed, ERDA, when fending off political calls for immediate commercialization programmes for specific technologies, often described itself as a kind of virtual market-place in which different communities of scientists, firms, and developers would be able to compete for resources. In part, the breadth of the initiative had to do with similar convictions on the side of planners and policy-makers; in part, it was the result of a growing bureaucratic susceptibility to pork-barrel politics. Besides being pulled into ever more technological ventures by regional interests, the initiative was home to conflicts over what exactly the problems with energy were. For many environmental groups, solar energy supporters, and small firms, for example, the energy crisis did not signify merely the depletion of cheap fuels, but the fact that the energy sector had been monopolized by Table 14.1. US Energy Research and Development Administration, budget for research and development, 1975–7

Nuclear safety and fuel cycle Conservation Geothermal energy Nuclear fusion Nuclear fission Solar energy of which: photovoltaics Fossil fuels Environmental technologies Total (technology support) Total (including basic research)

1975

1976

1977

120 21 21 151 538 15 2.6 138 7 1011 1324

163 55 32 224 522 86 16.4 333 12 1427 1800

282 91 50 304 709 116 24.3 442 16 2010 2413

Notes: Outlays in million US dollars. ‘Solar Energy’ was a synonym for renewable energy technologies until the early 1980s. Sources: US Energy Research and Development Administration. 1976. A National Plan for Energy Research, Development and Demonstration: Creating Energy Choices for the Future, Vol. 1: The Plan. Washington, D.C., 15 April, 37; US Energy Research and Development Administration, 1976: A National Plan for Energy Research, Development and Demonstration: Creating Energy Choices for the Future, Vol. 2: Program Implementation. Washington, D.C., 30 June, 103.

308

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Expectations and Diversity in Innovation 2012 US dollars (millions) 10000 9000 8000 7000 6000 5000 4000 3000 2000 1000 0 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 Cross-cutting

Others

Nuclear power

Renewable energies

Fossil fuels

Energy efficiency

Figure 14.1 Annual federal spending for energy technology research and development, 1974–90 Source: Data compiled by the International Energy Agency.

big corporations.2 Mainly through Congress and the media, they vocally criticized subsidies for research in corporate laboratories and funding for large-scale technological solutions, particularly in nuclear energy and the fossil fuels. Supporters of fossil fuel research and nuclear energy, in turn, heavily criticized environmental and other regulations, which they argued were the main causes of rising prices and the slow increase of supplies. Despite the sometimes ‘anarchic’ political battles on the energy issue, ERDA did in fact develop several more serious commercialization efforts. The remainder of this chapter takes a closer look at two of them—photovoltaics and Synfuels— as they illustrate the dilemma between the merits of diversity and cohesion.

Photovoltaics The photovoltaics commercialization initiative evolved from a niche programme at the beginning of the decade into one of the most-cited stars of 2 Good examples from the technology policy debate in this period are Commoner (1979); Hammond and Metz (1977); and Singular (1977).

309

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

the energy technology policies in the second half of the 1970s. It ultimately lost its dynamic when the support coalition fractured in conflicts about proper timelines, technological foci, and support instruments. In the first discussions after the 1973 oil embargo, the extremely expensive semiconductor technology, which had been developed in the space programmes since the end of the 1950s, received sparse support. Environmentalist and progressive groups at first focused on simpler and more advanced technologies, especially heat-based solar appliances. Big-science representatives and ERDA elites, on the other hand, focused on available high technology options, in nuclear energy as well as in mining, plant design, and conversion questions for fossil fuels. The rise of photovoltaics into one of the most promising renewable energy technologies in the 1970s can be explained by the focused development effort of a small community of dedicated supporters. This community was formed at a conference in late 1973 in Cherry Hill, New Jersey, at the invitation of NASA’s Jet Propulsion Laboratory (JPL). Collectively questioning years of consensus that the use of photovoltaics for large-scale energy conversion was contingent on breakthroughs in basic research, the group of entrepreneurs, state agency representatives, and scientists developed a belief in the feasibility of the coordinated industrial upscaling of long-known and comparatively simple crystalline silicon photovoltaics. In the words of one of the central figures in the emerging photovoltaics networks, William Cherry, the state would have to jump start production to unlock a lasting industrial dynamic: Definitely the government has got to do some pump priming. The semiconductor industry got started in the same way . . . [I]f you would look at the cost of semiconductors, you could see that there wasn’t much of a reduction over the years during the fifties. But as soon as the large amounts of government expenditures dropped off, the prices started coming down; the competition went up; and those who could make it for the price stayed in the field. The same thing is going to happen with us. (Jet Propulsion Laboratory 1973, 57)

The upscaling route of photovoltaics development was managed by the JPL. It unified most actors engaged in applied research and manufacturing with the goal of having the technology ready for large-scale mass production in 1986. The JPL consolidated the fragmented sector with numerous networking activities, feasibility studies, and pilot machine-tool contracts. It also engaged in direct industry support via systematic (and often fairly sizeable) block buys from manufacturers, systematic testing, and the dissemination of bestpractice knowledge in the industry. By 1977, the sector had tripled manufacturing capacity and cut the costs of the technology roughly in half, which brought further public and political attention to the technology and helped to secure sizeable annual increases in federal funding. In 1977, an author 310

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Expectations and Diversity in Innovation

for Science magazine, for example, declared: ‘The Semiconductor Revolution Comes to Solar’, and reported: [T]he federal photovoltaic research effort is credited by many observers as being perhaps the best conceived and most successful of the government solar programs . . . Not only is it achieving . . . reductions in the costs of silicon solar cells at a more rapid rate than that projected by its plan, but it also appears to have stimulated private industry into activity. (Hammond 1977, 445)

Euphoria around photovoltaics development at the time echoed a more general belief in the viability of a government-sponsored green-energy revolution during the Carter administration. Energy Secretary Schlesinger told Business Week in 1978 that ‘We certainly can declare the age of solar energy now more appropriately than we declared the atomic age 25 years ago’, while an adviser to the Governor of California added that ‘[we] see solar as a $4 billion to $7 billion industry by the late 1980s, with a labour force of more than 50,000 . . . It will be a bigger employer than electricity and gas’ (Business Week 1978, 90, 94). The projected large-scale commercialization of photovoltaics did not, in fact, materialize until the late 1990s and early 2000s. Instead, the sector abolished a large part of the industrial upscaling dynamic in the last years of the 1970s and returned to a more research-intensive mode. An important reason for this were growing uncertainties about the future of the technology and resulting industrial stagnation. Searching for an explanation of why industry was reluctant to invest in further capacity expansions, a JPL programme manager reported: In summary, anticipated rapid technological change delays or prevents investments, biases facilities toward labor intensive processes and increases product prices. Thus, any attempt on the part of government to increase (say double) R&D expenditures will increase the tendencies [to delay or prevent investment]. (Smith 1978, 19, emphasis added)

Ironically, the anticipation of ‘rapid technological change’ that was supposed to be preventing further investment emerged only as a result of the influx of funds into the sector, which in turn was based on the promise of rapid coordinated upscaling. These internal sectoral dynamics resonated with an increase in warnings against the incremental pathway towards commercial maturity, which eventually led to a questioning of the earlier unifying scenarios. Actors in industry as well as in government became wary of risking ‘premature’ technological lock-in, and this wariness cost them much of the dynamic of imagined medium-term readiness. After Congress passed an already weakened support bill in 1978, Jimmy Carter publicly criticized the initiative, suggesting that it was ‘still too early to concentrate on commercialization of photovoltaics. Photovoltaic systems hold great promise, but in the short run we must emphasize research and development, including 311

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

fundamental work on the physical properties of these systems, so that this promise can be realized’ (Carter 1978). At the beginning of the 1980s, after neither industry nor government made decisive investments, a programme manager reported on the future direction of photovoltaics support under the Reagan administration that ‘market development expenditures and associated commercial readiness targets have been deleted. In addition, all technical readiness goals have been dropped. In their place, a set of “technical feasibility targets” limited to selected, high-risk PV components and processes will be substituted’ (US Congress 1981, 95). In terms of industrial development, the sector, it seemed, was not that far away from its situation in 1972.3

Synfuels Synthetic fuel commercialization, in a sense, took the opposite direction from photovoltaics. After decades of failed attempts, proponents managed to establish a state-owned Synthetic Fuel Corporation after the second oil crisis that was meant swiftly to commercialize the technology and make alternatives to oil imports available. Broad Synthetic Fuels development—an umbrella term capturing conversion technologies for fossil fuels, such as coal gasification, shale to gas conversion and others—had been proposed by the defence sector and allies of the coal industry since the end of World War II. While parts of Congress managed to include provisions for loan guarantees and the establishment of a government corporation for coal gasification in the early Ford and Carter bills for energy technology policy, systematic funding for rapid commercialization was blocked by various groups until 1980 (Ikenberry 1988, 129–31). Entrepreneurial members of Congress and a change of stance in the Carter administration made the establishment of the Synfuels Corporation possible. Carter himself called for a $88 billion funding commitment in 1979, promising 2 million barrels of synthetic fuels a day by 1992 (worldwide consumption of oil in 1980 was roughly 63 million barrels a day) and asking Congress for the establishment of ‘an independent, government-sponsored enterprise with Federal charter’ (quoted in ibid., 133). Supporting members of Congress managed to establish the Synfuels Corporation with $3 billion in initial funding and an estimated commitment to the programme as a whole of $92 billion until 1992 in an overarching energy and defence policy package. The resulting Synfuels programme consisted of purchase agreements by the Department of Defense, various large-scale demonstration and pilot plants, up to ten-figure loan guarantees for exploration activities, and further research commitments 3

For a more extensive account of the US photovoltaics industry in this period, see Ergen (2017).

312

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Expectations and Diversity in Innovation

by the Department of Energy (a full overview is given by Anadón and Nemet 2014). After the passage of the measures, Carter issued euphoric statements. The ‘keystone of our national energy policy is at last being put into place’, he declared, and promised ‘70,000 jobs a year to design, build, operate and supply resources for synthetic fuel plants’ (New York Times 1980). These promises can be viewed as part of what Beckert and Bronk in their introduction to this volume discuss as the management of expectations. The promises represented efforts by the administration to popularize problem perceptions and possible future solutions conducive to getting bold programmes under way to deal with the energy crisis. Even before the second oil crisis of 1979 raised the pressure on the government to demonstrate its capacity to react to the turbulence, the administration repeatedly tried to secure the legitimacy of path-breaking energy policies by publicizing promises about the benefits of future energy independence and by issuing warnings about the dangers of medium-term inaction. This is the immediate context of the energy-related parts of Jimmy Carter’s famous ‘crisis of confidence’ speech, in which he declared coping with the energy crisis to be the ‘moral equivalent of war’ and reasoned that the Synfuels Corporation would be the manifestation of concerted action in the energy arena. ‘Just as a similar synthetic rubber corporation helped us win World War II’, Carter promised, ‘so will we mobilize American determination and ability to win the energy war’ (Carter 1979). For the first years of the programme, the Synfuels Corporation benefited from the fact that the support coalition hung together. The Corporation received more than enough applications for demonstration plants and, for a short time, industrial brand-names engaged in coal liquefaction and gasification and in shale exploration. The initiative suffered, however, from the massive fall in oil prices during the 1980s, the so-called Oil Glut, and from repeated cost overruns for its demonstration plants. As Deutch and Lester (2004, 203–4) explain, the success of synthetic fuels was contingent on widely agreed predictions of the price of oil at the end of the decade. Expected prices of up to $100 a barrel would have made synthetic fuels roughly competitive; oil prices in 1990 were, however, closer to around $20. This largely unexpected change of the environment—coupled with a range of organizational scandals and permanent environmentalist attacks on the Synfuels effort—is why Congress and the administration had few opponents when they eventually abolished the Synthetic Fuels Corporation in 1986. Critical observers of the history of the American Synthetic Fuels effort have suggested that a more cautious and temperate approach in relation to oil price forecasts, and the absence of fixed production goals, could have saved government and industry from failure (Deutch and Lester 2004). ‘Softer’ state instruments, such as the generation and spread of information and R&D support, 313

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures

might have kept the effort flexible, while still providing some insurance against the possibility of continuously rising oil and gas prices. While the merits of the initiative are of course debatable retrospectively, critics neglect the important point that lock-in into the scenario of competitive synthetic fuels—and the connected emphasis on ramping up production—were constitutive for the effort. It is difficult to imagine that a credible initiative to discover whether synthetic fuels could play a significant role in the mid-1980s energy provision would have emerged on the basis of widespread doubt about the technology’s potential.

Conclusion When assessing hits and misses in technological development one has to be careful not to fall for the fallacy of analysing past decisions on the basis of hindsight and current technological knowledge. This danger exists in all historical research that analyses past expectations about coming futures. The true industrial potential of crystalline silicon photovoltaics was demonstrated only in the 2000s by ramping up production to a once unthinkable scale and with the risk of failure on a Synfuel-scale. Similarly, the oil glut of the 1980s was truly unexpected. The strong cyclicality of oil prices, now almost common knowledge, was rarely a natural way to think about the future of oil prices at the end of the 1970s. Moreover, neither initiative was a full-blown ‘failure’. As is often the case in technological innovation, the ERDA programmes left behind building blocks for decades of technological change, for example for photovoltaics development in the 1990s and 2000s and for the shale revolution in recent years. Seen in this way, the designs of developmental initiatives at the end of the 1970s were decisions taken under genuine uncertainty. And it is exactly in such situations that the dilemma between aligned expectations and diversity in innovation becomes relevant for thinking about technological development. This chapter does not argue for or against one particular model of developmental organization. Rather, it suggests that the recently proposed model of a less intrusive developmental state, which restricts itself to the creation of technological options, but does not forcefully pick them, might be incomplete. Particularly in fields in which technological advances are not just ‘nice to haves’, promising prosperity and employment, but rather are key to solving major societal problems—as in the case of the environmental issues related to energy production—‘old-fashioned’ focused state support has merits of its own. It can help temporarily to suspend doubts about future developments and thereby lay the foundations for discovery. 314

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Expectations and Diversity in Innovation

The existence of a dilemma does not imply that organizations or states should refrain from constantly making judgements about how to balance the drawbacks of lock-in and openness, or that they should avoid developing strategies to mitigate the drawbacks of both. For example, the dilemma can be obviated to some extent through organizational ‘heterarchy’, as described by David Stark (2009), within focused activities or through an organizational model of a broad coverage of focused activities, as exemplified by the early ERDA and later DOE as a whole. As an influential older literature in industrial research suggests, there exist organizational and institutional configurations that allow for what used to be called ‘flexible specialization’ (Piore and Sabel 1984; Sabel et al. 1989). By maintaining and cultivating general-purpose resources and skills, organizations and networks of organizations might indeed be able to engage in focused development without full-blown lock-in.4 Such strategies are likely to turn out to be more difficult in cases of complex technologies whose development can occupy large parts of relevant organizations or sectors. Merely through scale and complexity, these technologies often develop what Thomas Hughes, in his comparative study of electricity systems, called ‘an inertia of directed motion’ (Hughes 1983, 15). This inertia undermines the diversity of open-ended search movements for technological alternatives, while concerned actors have little opportunity to find out whether such lock-in was worth it without committing to it in the first place. The challenge for the organization of technological development, then, does not lie in finding rarely available best-of-both-worlds approaches or making futile optimality calculations under genuine uncertainty about the future. Rather, it is about manoeuvring through situations in which every alternative for action might have significant drawbacks in the future.

Bibliography Alexander, Tom. 1976. ‘ERDA’s Job Is to Throw Money at the Energy Crisis’. Fortune, July: pp. 152–62. Anadón, Laura Díaz, and Gregory F. Nemet. 2014. ‘The U.S. Synthetic Fuels Corporation: Policy Consistency, Flexibility, and the Long-Term Consequences of Perceived Failures’. In Energy Technology Innovation: Learning from Historical Successes and Failures, edited by Arnulf Grubler and Charlie Wilson, pp. 257–72. Cambridge: Cambridge University Press. Beckert, Jens. 2016. Imagined Futures: Fictional Expectations and Capitalist Dynamics. Cambridge, MA: Harvard University Press.

4 Lester and Piore (2004, chapter 5) document and discuss various strategies of combining analysis and interpretation in their empirical case studies.

315

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures Blaug, Mark. 1962 [1990]. Economic Theory in Retrospect. Cambridge: Cambridge University Press. Block, Fred. 2008. ‘Swimming Against the Current: The Rise of a Hidden Developmental State in the United States’. Politics and Society 36 (2): pp. 169–206. Bronk, Richard, and Wade Jacoby. 2016. ‘Uncertainty and the Dangers of Monocultures in Regulation, Analysis, and Practice’. MPIfG Discussion Paper 16/6, Max Planck Institute for the Study of Societies, Cologne. Business Week. 1978. ‘The Coming Boom in Solar Energy’. 9 October: pp. 88. Carter, Jimmy. 1978. ‘Solar Photovoltaic Energy Research, Development, and Demonstration Act of 1978 Statement on Signing H.R. 12874 Into Law’. 4 November 1978, Washington, D.C. Carter, Jimmy. 1979. ‘Crisis of Confidence’. Speech, 14 July 1979, Washington, D.C. Commoner, Barry. 1979. ‘I–The Solar Transition’. New Yorker, 23 April: pp. 53–98. Deutch, John M., and Richard Lester. 2004. Making Technology Work: Applications in Energy and the Environment. Cambridge: Cambridge University Press. Ergen, Timur. 2017. ‘Coalitional Cohesion in Technology Policy: The Case of the Early Solar Cell Industry in the United States’. MPIfG Discussion Paper 17/7, Max Planck Institute for the Study of Societies, Cologne. Evans, Peter. 1995. Embedded Autonomy: States and Industrial Transformation. Princeton, NJ: Princeton University Press. Garud, Raghu, and Peter Karnøe. 2003. ‘Bricolage versus Breakthrough: Distributed and Embedded Agency in Technology Entrepreneurship’. Research Policy 32 (2): pp. 277–300. Gawer, Annabelle. 2000. ‘The Organization of Platform Leadership: An Empirical Investigation of Intel’s Management Processes Aimed at Fostering Complementary Innovation by Third Parties’. PhD Thesis, Sloan School of Management, Massachusetts Institute of Technology, Cambridge, MA. Grabher, Gernot. 1994. Lob der Verschwendung: Redundanz in der Regionalentwicklung; Ein sozioökonomisches Plädoyer. Berlin: Sigma. Hammond, Allen L. 1977. ‘Photovoltaics: The Semiconductor Revolution Comes to Solar’. Science 197 (4302): pp. 445–7. Hammond, Allen L., and William D. Metz. 1977. ‘Solar Energy Research: Making Solar After the Nuclear Model?’ Science 197: pp. 241–4. Hoddeson, Lillian. 1981. ‘The Emergence of Basic Research in the Bell Telephone System, 1875–1915’. Technology and Culture 22 (3): pp. 512–44. Hughes, Thomas P. 1983. Networks of Power: Electrification in Western Society, 1880–1930. Baltimore, MA: Johns Hopkins University Press. Ikenberry, G. John. 1988. Reasons of State: Oil Politics and the Capacities of American Government. Ithaca, NY: Cornell University Press. Jacobs, Meg. 2016. Panic at the Pump: The Energy Crisis and the Transformation of American Politics in the 1970s. New York: Hill and Wang. Jet Propulsion Laboratory. 1973. ‘Workshop Proceedings: Photovoltaic Conversion of Solar Energy for Terrestrial Applications. Workshop Proceedings, Vol. I. Working Group and Panel Reports’. Cherry Hill, NJ.

316

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Expectations and Diversity in Innovation Kleinman, Daniel Lee. 1995. Politics on the Endless Frontier: Postwar Research Policy in the United States. Durham, NC: Duke University Press. Kitschelt, Herbert. 1983. Politik und Energie: Energie-Technologiepolitiken in den USA, der Bundesrepublik Deutschland, Frankreich und Schweden. Frankfurt am Main: Campus. Knight, Frank H. 1921. Risk, Uncertainty, and Profit. Boston, MA: Houghton Mifflin. Lester, Richard K., and Michael J. Piore. 2004. Innovation: The Missing Dimension. Cambridge, MA: Harvard University Press. March, James G. 1991. ‘Exploration and Exploitation in Organizational Learning’. Organization Science 2 (1): pp. 71–87. March, James G., and Herbert A. Simon. 1958. Organizations. New York: Wiley. Mowery, David C., and Nathan Rosenberg. 1989. Technology and the Pursuit of Economic Growth. Cambridge: Cambridge University Press. New York Times. 1980. ‘Synthetic Fuels Bill Signed’. 1 July: pp. 57, 60. Nixon, Richard M. 1973. ‘Address to the Nation about Policies to Deal with the Energy Shortages’, speech on 7 November. Washington D.C. Ó Riain, Séan. 2004. The Politics of High-Tech Growth: Developmental Network States in the Global Economy. Cambridge: Cambridge University Press. Piore, Michael J. 2008. ‘Learning on the Fly: Reviving Active Governmental Policy in an Economic Crisis’. Conference Paper, Conference on ‘How Will a New Administration and Congress Support Innovation in an Economic Crisis?’. Economic Policy Institute, Washington, D.C. Piore, Michael J., and Charles F. Sabel. 1984. The Second Industrial Divide: Possibilities for Prosperity. New York: Basic Books. Rodrik, Dani. 2004. ‘Industrial Policy for the Twenty-First Century’. KSG Working Paper no. RWP04–047. Rosenberg, Nathan. 1982. Inside the Black Box: Technology and Economics. Cambridge: Cambridge University Press. Sabel, Charles F., Gary B. Herrigel, Richard Deeg, and Richard Kazis. 1989. ‘Regional Prosperities Compared: Massachusetts and Baden-Württemberg in the 1980s’. Economy and Society 18 (4): pp. 374–404. Schrank, Andrew, and Josh Whitford. 2009. ‘Industrial Policy in the United States: A Neo-Polanyian Interpretation’. Politics & Society 37 (4): pp. 521–53. Schumpeter, Joseph A. 1942 [1975]. Capitalism, Socialism, and Democracy. New York: Harper. Schwartz Cowan, Ruth. 1987. ‘The Consumption Junction: A Proposal for Research Strategies in the Sociology of Technology’. In The Social Construction of Technological Systems: New Directions in the Sociology and History of Technology, edited by Wiebe W. Bijker, Thomas P. Hughes, and Trevor J. Pinch, pp. 261–80. Cambridge, MA: MIT Press. Singular, Stephen. 1977. ‘Solar Energy Battle Heats Up in California’. Chicago Tribune, 23 October: pp. 241–4. Smith, Jeffrey L. 1978. ‘The Industrialization of Photovoltaic Systems’. Jet Propulsion Laboratory, Pasadena, CA. Stark, David. 2009. The Sense of Dissonance: Accounts of Worth in Economic Life. Princeton, NJ: Princeton University Press.

317

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Uncertain Futures US Congress. 1981. ‘Administration Budget Cuts in Conservation and Solar Programs. Hearings before the Subcommittee on Energy Conservation and Power of the Committee on Energy and Commerce’. House of Representatives, Ninety-Seventh Congress, First Session, 1–3 December. van Lente, Harro, and Arie Rip. 1998. ‘The Rise of Membrane Technology: From Rhetorics to Social Reality’. Social Studies of Science 28 (2): pp. 221–54. Weir, Margaret. 2006. ‘When Does Politics Create Policy? The Organizational Politics of Change’. In Rethinking Political Institutions, edited by Ian Shapiro, Stephen Skowronek, and Daniel Galvin, pp. 171–86. New York: New York University Press. Weir, Margaret, Jane Rongerude, and Christopher K. Ansell. 2009. ‘Collaboration Is Not Enough: Virtuous Cycles of Reform in Transportation Policy’. Urban Affairs Review 44 (4): pp. 455–89. Winter, Sidney G. 1971. ‘Satisficing, Selection, and the Innovating Remnant’. Quarterly Journal of Economics 85 (2): pp. 237–61. Ziegler, J. Nicholas. 1997. Governing Ideas: Strategies for Innovation in France and Germany. Ithaca, NY: Cornell University Press.

318

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Contributor Biographies

Jenny Andersson is CNRS Research Professor and the Co-Director of MaxPo, Max Planck Sciences Po Center on Coping with Instability in Market Societies, at Sciences Po in Paris. Her book, The Future of the World: Futurology, Futurists and the Struggle for the Post Cold War Imagination, was published in 2018 by Oxford University Press. Jens Beckert is Professor of Sociology and Director at the Max Planck Institute for the Study of Societies in Cologne. He is the author of Imagined Futures: Fictional Expectations and Capitalist Dynamics, Harvard University Press, 2016. In 2018, Beckert received the Gottfried Wilhelm Leibniz Prize for his work in reinvigorating the social sciences with an interdisciplinary perspective, especially at the intersection of sociology and economics. Natalia Besedovsky is a postdoctoral researcher at the University of Hamburg. She studied in Cologne (Germany) and Princeton (USA) and received her PhD at the Humboldt University of Berlin (Germany) before working at the Institute for Intercultural and International Studies at the University of Bremen. Her research areas include the sociology of finance, practices of knowledge production, risk conceptions, knowledge intermediaries, and practice theory. Her article ‘Financialization as Calculative Practice: The Rise of Structured Finance and the Cultural and Calculative Transformation of Credit Rating Agencies’ was published in 2018 in Socio-Economic Review. Robert Boyer is an economist and currently a member of the scientific committee of the Institute of the Americas in Paris. He has been a senior researcher at CNRS (National Center for Scientific Research) and professor at EHESS (School for Advanced Studies in the Social Sciences). He has been active in regulation theory devoted to the analysis of the long-run transformations of capitalism. Benjamin Braun is a senior researcher at the Max Planck Institute for the Study of Societies in Cologne, Germany. His work on central banking has been published, among others, in the British Journal of Politics and International Relations, Economy and Society, and the Review of International Political Economy. His 2017 report on the transparency and accountability of the European Central Bank has been published by Transparency International. Richard Bronk is a visiting fellow in the European Institute at the London School of Economics and Political Science (LSE), where he taught political economy from 2000–7. Before joining the LSE, Bronk spent seventeen years in the City of London, including senior positions in fund management and as Adviser on European capital markets at the Bank of England. He is the author of The Romantic Economist: Imagination

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Contributor Biographies in Economics, Cambridge University Press, 2009; and his research now focuses on the role of imagination and language in economics, the dangers of analytical and regulatory monocultures, and the epistemology of markets. Liliana Doganova is associate professor at the Center for the Sociology of Innovation, MINES ParisTech. At the intersection of economic sociology and Science and Technology Studies (STS), her work has focused on business models, the valorization of public research, and markets for bio- and clean-technologies. She has published in journals such as Research Policy, Science and Public Policy, and the Journal of Cultural Economy, and she is currently preparing a monograph on the historical sociology of discounting. Timur Ergen is a research fellow at the Max Planck Institute for the Study of Societies (MPIfG) in Cologne, Germany. He obtained his PhD from the University of Cologne in 2015 and held an IMPRS Fellowship from the MPIfG from 2010 until 2014. His research focuses on innovation and technology policy, deindustrialization in a historical perspective, and the sociology of expectations. Elena Esposito is Professor of Sociology at the University Bielefeld (Germany) and at the University of Modena-Reggio Emilia (Italy). She has published many works on the theory of social systems, media theory, memory theory, and sociology of financial markets; and her current research focuses on a sociology of algorithms. Esposito’s recent publications include The Future of Futures: The Time of Money in Financing and Society, Edward Elgar, 2011; ‘Performativity and Unpredictability in Economic Operations’, Economy and Society 42, 2013; and ‘Artificial Communication? The Production of Contingency by Algorithms’, Zeitschrift für Soziologie 46, 2017. Martin Giraudeau is Assistant Professor in Accounting at the London School of Economics and Political Science. His research focuses on the sociology and history of accounting, and more broadly of management, organizations, and capitalism. Drawing inspiration from Science and Technology Studies (STS), Martin has especially explored the forms and roles of a specific entrepreneurial instrument, the business plan, since the late eighteenth century. Andrew G. Haldane is the Chief Economist at the Bank of England. He is also Executive Director, Monetary Analysis, Research and Statistics. He is a member of the Bank’s Monetary Policy Committee. He also has responsibility for research and statistics across the Bank. Douglas R. Holmes teaches anthropology at the State University of New York at Binghamton. He is known primarily for an ethnographic trilogy: Cultural Disenchantments: Worker Peasantries in Northeast Italy, Princeton University Press, 1989; Integral Europe: Fast-capitalism, Multiculturalism, Neo-fascism, Princeton University Press, 2000; and Economy of Words: Communicative Imperatives in Central Banks, University of Chicago Press, 2013. He continues to work closely with George E. Marcus on a project of ‘re-functioning ethnography’ for the purposes of investigating cultures of expertise, with settings ranging from science labs to alternative art spaces. Olivier Pilmis is a Research Fellow in Sociology at the French National Center for Scientific Research (CNRS) and a member of the Centre de Sociologie des Organisations (Sciences Po—CNRS) in Paris. His research applies economic sociology, organizational

320

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

Contributor Biographies sociology, and sociological theory to the study of macroeconomic forecasting. Pilmis’s work focuses on the emergence of beliefs, the production of legitimate discourses about the future, and the social structure of the market for forecasting. Werner Reichmann is a Privatdozent and postdoctoral researcher at the University of Konstanz, Germany. He is a Science and Technology Studies (STS) scholar and works on the production, visualization, and impact of knowledge within the discipline of economics. He is the author of Wirtschaftsprognosen—Eine Soziologie des Wissens über die ökonomische Zukunft [Economic forecasts—A Sociology of Knowledge about the Economic Future], Campus, 2018. David Tuckett is a psychoanalyst, economist, and medical sociologist who is Director of the Centre for the Study of Decision-Making Uncertainty and Professor in the Psychoanalysis Unit and the Department of Science, Technology, and Engineering applied to Public Policy at University College London (UCL). His book, Minding the Markets: An Emotional Finance View of Financial Instability (Palgrave Macmillan, 2011) describes the decision-making of fifty leading asset managers and develops the theory of emotional finance. He now collaborates with colleagues at the Bank of England and UCL to develop a new approach to modelling economic behaviour under uncertainty, and he leads the EPSRC CRUISSE network.

321

OUP CORRECTED PROOF – FINAL, 25/5/2018, SPi

OUP CORRECTED PROOF – FINAL, 9/6/2018, SPi

Index

accounting 263, 265, 270–1 adaptive heuristics, and conviction narratives 70, 71 administrative procedures, role of formalities in 271–4 advertising 2, 6 agent-based models (ABMs) 15, 150–67 costs and benefits of 151–8 definition of 150 emergent behaviour and uncertainty in 6, 15, 152–3 heterogeneous agents 6–7, 150 heuristic behaviour and 150, 153–4, 159 interactive agents 15, 150 and multiple equilibria 15, 150, 154 and non-normal behaviour 154–8 real-world application of 150–1, 158–67 Akerlof, George 24 ambivalence, and Conviction Narrative Theory 71, 75–7 American Research and Development Corporation (ARD) 17, 259–75 project appraisal process 269–75 analysis, and technological innovation 299, 300, 301 Andersson, J. 27, 83–101 animal spirits 20, 54, 197 and conviction narratives 23, 63, 74–5, 77, 78 and grand narratives 48 Keynes on 21, 74, 279 see also emotions Arctic, the 27, 83–98 claims on the future of 27, 83, 84–7 and climate change 27, 84, 85 de-iced future of 84, 87, 88–90, 91–3 and environmental research 92, 95–6, 98 role of scenarios and forecasts in 85, 96 as site for competing expectations 84–7 and sustainability 86, 90, 91–3 Sweden as Arctic nation 27, 84, 87, 88, 89, 90, 93–7 Arctic Council 88, 90, 91, 92, 93, 96 Bank of Canada 180–1, 200 Bank of England 6 and agent-based models 15, 158–67

DSGE modelling framework 147 and forecasting ‘errors’ 126 and forward guidance 179–80, 203 network of agencies 22, 121, 188–90 One Bank Research Agenda 158, 185–7 and quantitative easing 203 and the UK housing market 158–62 Bank for International Settlements (BIS) 195 Bartlett, Frederic Charles 69 Beck, Ulrich 239 Becker, H.S. 135 Beckert, Jens vii–x, 1–36, 85, 86, 125, 240, 252, 279, 280, 299 on ‘instruments of imagination’ 261–2 behavioural economics vii, 9, 75 behavioural rules of thumb and agent-based modelling 150, 153–4, 159 see also heuristics Bernanke, Ben 26, 174, 201 Bernard, Claude 267 Besedovsky, Natalia 13, 19–20, 51, 236–56 biotechnology 287, 290–1 BIS (Bank for International Settlements) 195 Black Swans 131 Blaug, Mark 304 Blinder, Alan 177, 200 Block, Fred 303 Black-Scholes formula 223, 226, 227, 229, 230 bounded rationality 5, 54, 151 Bourdieu, Pierre 119, 279 Boyer, Robert 24–5, 29, 39–61 brain 77 and Conviction Narrative Theory 69, 70–1, 74, 75 and emotions 72–3 new connections in 3 Suddendorf on 107 see also neuroscience Braun, Benjamin 26, 32, 47, 175–6, 189, 194–216 Braun-Munzinger, Karen 164, 167 Brexit 16, 26–7, 133–4 Bronk, Richard vii–x, 1–36, 240, 261, 281, 299, 303 Bruner, Jerome 68, 69, 72 Buchanan, James 5, 6

OUP CORRECTED PROOF – FINAL, 1/6/2018, SPi

Index business plans viii, 13, 21, 28, 40, 46 at American Research and Development Corporation (ARD) 272–3, 280 calculation, rational 41 and imaginaries viii, 10, 13, 17, 32 use in conditions of uncertainty 9, 14, 21, 23, 51, 65, 74, 221, 232, 261, 279 versus storytelling in forecasting 134, 137, 139 calculative devices viii, 5, 10, 33 as constitutive of markets 18–20 as diagnostic tools 14–15, 16–17 as example of ‘reasoning imagination’ 13–14 as instruments of the imagination 13–18 performative impact of 28, 29 as props for decision making 18, 227, 228, 262 as social justifications for action 18–20 widely shared 12, 32 calculative practices, and different concept of risk 241–4 Callon, Michel 28, 185 Canada and the Arctic 90, 92–3, 94 Bank of Canada 180–1, 200 capitalism and capitalist systems 49, 58 and competition 1–2 future orientation of 1, 42, 279, 286 and fictional expectations 4, 21, 26, 85 and maximization 1, 2 and novelty vii, 2 and uncertainty 2–3, 5–10, 32, 41–3, 252 Carney, Mark 185–7 Carruthers, Bruce G. 251 Carter, Jimmy 307, 311–12, 313 Cartwright, Nancy 145–6 Castel, Robert 245 CDO markets 19, 77, 221, 245, 249, 253 ‘celestial mechanics’ metaphor 6, 145 central banking bending the yield curve 205, 207, 208, 211, 212 and central planning 26, 32, 194–216 ‘communism of models’ in 199–201, 211 distributional consequences of 195, 211 DSGE modelling framework 147, 200 and economic forecasting 111, 136, 200, 203, 207–8, 211 epistemic authority of 26, 200–1, 203, 206–8 forward guidance viii, 22, 26, 178–80, 190, 201, 203–4, 205, 206–8, 210, 211 governability paradigms of 175–6, 196–9 and governments 47, 48 hydraulic instrument, use of 26, 175–6, 195–6, 201–3, 205, 208–10 and inflation targeting 175–7, 195–9, 206

324

monetary-policy stories 178, 180–4 narratives and conversation, use of 22, 26, 173–90, 207 networks, use of 188–90 performativity in 176, 178, 185, 195, 199–203, 205, 210, 211 and quantitative easing (QE) 26, 32, 40, 53, 178, 183–4, 201, 203–5, 208–10, 211 and the rational expectations hypothesis 177–8, 199–203, 211 relations with the public 173–90 role in coordinating expectations 177, 200 see also Bank of England; Bank of Canada; European Central Bank (ECB); New Zealand, Reserve Bank of; Riksbank; US Federal Reserve central planning and central banks 26, 32, 194–212 Keynes and Hayek on 31 see also indicative planning Cherry, William 310 China 52 and the Arctic 88, 89, 94, 95, 97 Chong, Kimberley 23, 63, 64, 66 climate change and the Arctic 84 and the green economy 53 re-imagining of 91–3 role of models in understanding 15 Cocteau, Jean 46 Coeuré, Benoît 181, 209 cognitive bias 9 cognitive lock-in, and technological innovation 305–6 Cold War, and the Arctic 95 competition 1–2, 42, 55, 90, 132, 299 Complex Adaptive System model 151 complex economic systems 9 emergent properties of 6, 152–3 and need for simple narratives 54–5 uncertainty as property of 6, 145, 152–3 complexity economics 7, 12 Complexity Research Initiative for Systemic Instabilities (CRISIS) 151 confidence, instilled by narratives or stories viii, 23, 24, 68, 74, 178 conversations and innovation 303, 306 as input to central bank policy 22, 178, 187–90 Conviction Narrative Theory (CNT) 23, 62–82 and ambivalence 71, 75–7 and animal spirits 23, 63, 74–5, 77, 78 divided states and 75–7 four functions of a conviction narrative 69–72 money management case study 63–7 role of emotions in 23, 65, 68, 71–5, 77

OUP CORRECTED PROOF – FINAL, 1/6/2018, SPi

Index coordination of expectations benefits of 24–6 destabilizing effect of 29, 55 forward guidance to achieve 208 role of markets in 45–6, 47, 55, 56 role of narratives in viii, 5, 25, 26, 46–58, 75, 85, 177, 208 trade-offs between costs and benefits of 30–2, 300, 306 coordination problems entailed by indeterminacy of beliefs 25 in innovation 31, 305 market and non-market solutions to 41, 194–5 narratives and predictions as solutions to 24–8, 85, 125–6 counter-performativity 29, 226–7, 229–33 Creative Destruction 2 creativity 1, 4, 85 in entrepreneurship 261–2, 265, 273 market as creative process 2, 5–6 and technological innovation 299–300, 302 see also imagination; innovation credit rating 236–56 and cardinal measures of risk 19–20, 246, 249 diagnostic versus technical conception of 236–9, 244–50, 252 and illusion of control 19 and ordinal ranking of risk 19, 244, 252 sovereign (country) ratings 13, 238, 242, 244–5, 247, 248, 249–50, 251–2, 253 in structured finance 238–9, 244, 245–51 credit rating agencies (CRAs) 19, 236–7, 243–53 credit risk models 243–52 see also credit rating; finance models; risk; risk management; structured finance dappled world 31, 145–6, 151, 167 and agent-based modelling 154–8 see also Cartwright, Nancy decision-making calculative devices as props to 13–18, 228 central bank influence on 208 and Conviction Narrative Theory 23, 62–78 fictional expectations or imaginaries as basis for 4–5, 10–12, 21, 125, 201, 228, 262 risk calculations exempting from responsibility for 248 nature of in conditions of uncertainty vii, viii, 3, 14, 32–3, 194, 261, 262 role of emotions in 23, 71–5, 269–70 role of formalities in 271–4 role of institutions and planning in 44 role of judgement in ix, 14, 17, 178, 240, 268, 279, 280, 294, 315

role of narratives in 23, 32, 62–78 use of discounted cash flow analysis for 278, 288 demand management Keynesian 175–6, 195, 197, 198 and quantitative easing 176, 205 Dequech, David 6 derivative markets 19, 43, 51 cardinal risk assessments in 20 trading in based on volatility calculus 20, 221, 222–3, 225, 226, 227 see also structured finance; volatility calculus Deutch, John M. 316 Developmental Network State 303 Dewey, John 302 Diamond, Peter 30 Digital Equipment Corporation 272 discounted cash flow (DCF) analysis viii, 13, 17–18, 278–97 in forestry management 17, 280–1, 282–7, 291, 294, 295 in the pharmaceutical industry 17, 281, 287–93, 294, 295 making multiple options visible 17, 293 and risk premium 281, 291 as theory of value 278–9 and uncertainty 278–95 valuation software 292, 293 discourse analysis 23, 77–8, 139 dissonance, cognitive benefits of 28–32, 189, 302 divided vs integrated mental states 75–7 Doganova, Liliana 13, 17–18, 278–97 dominant mood, and economic forecasting 120 Dominguez, Kathryn M. 131 Donham, Dean Wallace B. 265–7 Doriot, Georges F. 17, 259–60, 262–76 dot.com crisis 50, 51 Draghi, Mario 22, 184, 185, 203–4 drug development see pharmaceutical industry Durkheim, Émile 125 Dynamic Stochastic General Equilibrium (DSGE) model 16, 146–8, 154, 159, 199, 200, 211 ECB see European Central Bank (ECB) eco-modernization paradigm, and Arctic futures 87 econometric models and techniques 39, 110–11, 112, 113, 115, 121, 128, 129, 132, 133, 136, 181 economic crisis 7, 24, 32, 39–40, 41, 43 failure to model uncertainty causing 63 failure of models to predict 149–50 grand narratives as cause of 39, 50, 53–6 see also financial crisis

325

OUP CORRECTED PROOF – FINAL, 1/6/2018, SPi

Index economic forecasting viii, ix, 8, 15–16, 105–70 and Brexit 16, 27, 133–4 consistency in 115–16, 136, 153 as coordination device 25, 26, 125 criticisms of 15–16, 124–6 as diagnostic tool 15, 16, 108, 133 dominant mood and 120 and econometric models 110–11, 113, 115, 129, 132, 133, 136 epistemic quality of 111–14 epistemology of 132–6 explaining errors in 124–43 ‘foretalk’, role of in producing 16, 22, 108–18 German forecasting institutes and 22, 105, 109–10, 112–18, 119 herding and 119–20 interactional foundations of 22, 105–23 magic compared with 16, 125, 126, 127, 133, 135, 137–8, 139, 140 narrative economics as new tool for 23, 77–8 narrative scenarios in 16, 22, 25, 132, 133, 135, 139 point forecasts versus narratives 128–9, 133, 140 social legitimacy of 105, 111, 121 use by central banks 111, 136, 200, 203, 207–8, 211 see also macroeconomic forecasting economic growth, limits to 48, 54 economic institutions, devices for guiding expectations and behaviour 44 economic models borrowed from epidemiology 24 incompleteness of 30 political significance of 27–8 standard versions of vii, ix, 8, 9, 10, 42, 62, 75, 78, 147–50, 152, 153, 154, 167, 168 and uncertainty vii, 8, 44, 63 see also agent-based models; complexity economics; Dynamic Stochastic General Equilibrium (DSGE) models; econometric models; rational expectations theory economic systems disequilibrium in 6–7, 145 see also complex economic systems economics heterodox vii, 125, 131, 145, 167 standard vii, ix, 8, 10, 40, 42, 43, 44, 56, 62, 75, 78, 154, 167 efficient market hypothesis 8–9, 211 electric cars 305, 306 emergent behaviour 6, 15, 152–3 see also complexity economics emotions 4, 41, 69 ‘approach’ versus ‘avoidance’ 23, 68, 71–4 contagion of 24, 77

326

dangers of 270 evoked by conviction narratives 23, 71–4 group nature of 24, 29, 76, 77 and money management 65–7 and project appraisal 269–70, 271, 274–5 role in decision-making 23, 71–5 see also animal spirits energy technology development in the US 25, 31, 298–315 photovoltaics 309–12, 314 synthetic fuels 312–14 epistemic authority, in central banking 26, 200–1, 203, 206–8 epistemic culture 19, 238–9, 242, 251, 252 epistemological uncertainty 6, 240 equilibrium 42, 43, 145–6, 147, 267 multiple equilibria in agent-based modelling 15, 150, 154 concept of in economics 5, 6, 9, 12, 40, 55, 56, 75, 76, 154 see also Dynamic Stochastic General Equilibrium (DSGE) model ERDA (Energy Research and Development Administration) 307–9, 310, 314 Ergen, Timur 25, 31, 298–318 ergodic futures 8, 10, 14, 20, 56, 106, 118, 241 see also non-ergodic futures Esposito, Elena 20, 40, 219–35 EURACE 151 European Central Bank (ECB) 22 epistemic authority of 201, 206–8 and the euro crisis 184–5 and forward guidance 203–4, 207, 208, 209, 211 post-financial crisis planning 205–8 quantitative easing 204–5, 208–11 European Union (EU) and the Arctic 88, 89, 90, 94 referendum on Brexit 26–7 Evans, Peter 305 Evans, Robert 110 exogenous shocks 7, 32, 153, 160 as excuse for errors in economic forecasting 129–30 expectation regimes chronology of 54 contingent nature of 40–1 definition of 41 see also socio-economic regimes expectations vii–ix, 3, 4, 8 and Arctic futures 27, 83–7, 92–3, 96 central bank influence on 22, 26, 176, 177, 179, 182, 185, 190, 195, 198, 200–5, 207, 209 contingent nature of vii, viii, ix, 6, 11, 12, 15, 25, 29, 32, 280 co-produced by imagination and reason 4–5, 13–14

OUP CORRECTED PROOF – FINAL, 1/6/2018, SPi

Index and economic forecasts 105, 113, 118, 120, 121 and emotions 66, 74 impact of models on 27–8, 226–7, 232 and institutions 11, 44 and interests 27, 84–90, 97 market coordination of 45–6, 47, 55, 56 performative effect of 28–30 and politics 9, 28, 86–7, 97 and power 26, 27, 28 role of narratives in coordinating or socializing viii, 5, 25, 26, 46–58, 75, 85, 177, 208 social construction of 11–12 stabilization of 30, 31, 32, 182 technological innovation and the alignment of 300–6, 313 see also coordination of expectations; fictional expectations; rational expectations theory/hypothesis Farrell, John 94 fat tails in probability distributions 51, 154, 158, 161–2, 167 Faustmann, Martin 282–3, 284–5, 291 fictional expectations 5, 10, 15 Beckert on 3, 4, 10–12, 13, 85, 86 and calculative devices 13–17 interactional basis of 109 and macroeconomic forecasting 125–6 and performativity 28 as opposed to rational expectations 10, 32 uncertainty and 10–11, 21 fictions calculated 20, 46, 228 literary 10, 228 shared 4, finance models viii, ix, 28 see also credit risk models; structured finance; volatility calculus financial crisis 124, 129, 144, 167 and agent-based modelling 162–4 and credit rating practices 19, 20, 236, 250–2 and forecasting errors 124, 129–30, 148–9 and modelling monoculture 29, 200 grand narratives as cause of 39, 51, 53–6, 58 link to narrative shifts or revisions 24, 40 post-crisis central bank interventions and planning 26, 175–6, 178, 181, 187, 202–4, 205–10 and radical uncertainty 63 and structured finance 220, 227 and the unpredictability of risk 230 see also economic crisis financial markets 19–20, 56–8 agent-based models of 162–7 and the green economy 52–3 and imaginaries 51–2

money management and conviction narratives 63–7 and narratives 24, 41, 51–2 performativity of structured finance models in 227, 228 risk in 153, 222, 238, 241, 243, 248 Finland 90 Fischer, Stanley 201 Fisher, Irving 131, 177, 281, 286 flexible specialization 315 forecasting see economic forecasting; prediction forestry management, discounting in 17, 280–1, 282–3, 291, 294, 295 foretalk, and economic forecasting 16, 22, 106, 108–9, 111, 115, 116–17 forward guidance viii, 22, 26, 178–80, 201, 203–4, 205, 206–8, 210, 211 see also central banking Foucault, Michel 28 framing biases 5 France indicative planning 41, 44–5 macroeconomic forecasting 128 freedom, and the indeterminacy of the future 1, 2, 219, 220 Freeman, Christopher 279 Friedman, Milton 195 Frisch/Slutsky impulse-propagation model 152 future, the 6, 8, 280 claims on the future 84–7 discounting of 278–95 open versus closed versions of 1, 2, 20, 27, 86, 87, 97–8, 219–20, 227–8 orientation towards as feature of capitalism 1, 42, 279, 286 processing 259–77 ‘sense’ of 262–8, 270 see also ergodic futures; indeterminacy of future ‘future present’ versus ‘present future’ 11–12, 222–3, 227, 228–9, 232, 293 futures markets 41, 45–6, 56 ‘futurity’ 42, 46 Garud, Raghu 304 GDP (gross domestic product) agent-based modelling of 154–8 forecasting GDP growth 110, 128, 135, 138–9, 148–9 Gehren, Edmund Franz von 282–4, 291 Genentech 290 Germany Bundesbank 180, 185 forecasting institutes in 22, 105, 109–10, 112–18, 119 Gibson, David 108–9, 112 Giddens, Anthony 239

327

OUP CORRECTED PROOF – FINAL, 1/6/2018, SPi

Index Giraudeau, Martin 13, 17, 259–77 globalization 24, 43, 56–8, 117–18, 198 Goffman, Erving 106 Goldman Sachs 209 Goodhart’s law 185 governability paradigms, in central banking 175–6, 196–9 Gramsci, Antonio 194, 208 grand narratives and financial crises 39, 51, 53–6, 58 and socio-economic regimes 24, 48–56, 58 Great Depression 77, 131, 144 green economy 52–3 Greenland 90, 91 Greenspan, Alan 174 Grömling, Michael 119 groupfeel 24, 76, 77 groupthink 76, 77, 189, 195 habitus 12, 237 Haldane, Andrew G. viii, 6–7, 15–16, 31, 51, 144–70 Hamilton, Valerie 13 Harvard Economic Service 131 Harvard School of Business Administration 265–6 Hayek, F.A. vii, 5, 11, 18, 31–2, 121, 195, 201 Hazlitt, William 4, 13–14 Henderson, Lawrence J. 267 herding 119–20, 195 heterodox economics vii, 125, 131, 145, 167 heuristics 70, 71, 72, 153 see also behavioural rules of thumb Holmes, Douglas R. 22, 26, 111, 121, 173–93 Honig, Benson 280 housing market 158–62 Hubert, Henri 125, 126, 127 Hughes, Thomas 315 Iceland 88, 89, 90, 94 imaginaries viii, ix, 2, 3, 4–5, 7, 12, 26, 32, 33, 42, 48, 51, 58, 98, 105, 253, 270 chronology of 54 and emotions 12, 73 and fantasy 4, 12, 228 hegemonic 39, 51–2 and interests 84, 87 informed 18 and models 13–14 and narratives 3, 11, 32, 39, 51–2, 53 and political power 4, 7, 27, 97 public images 11, 12 and rational calculation viii, 10, 13, 17, 32 role of foretalk and interaction in formation of 16, 105, 108, 111, 114, 118, 121 as templates for action 3, 5 technological 49, 270, 305

328

as tool for coping with uncertainty ix, 3, 4, 5 see also fictional expectations imagination 3–5 as basis of sympathy 3 calculation devices as instruments of 13–18, 125, 261–2 and conviction narratives 68–9 definition of 3–4 and fictional expectations 10–12 as force for evil 4 and knowledge 14, 17, 261, 265–8, 274–5 and memory 107 relationship to reason 4–5, 13–14, 17, 261, 266–7 role in entrepreneurship 262, 265, 266 and sense of future 14–15, 17, 30, 261–2 and uncertainty 2, 3, 4, 45, 261 and valuation 11, 292 see also negative capability; reasoning imagination incommensurable values 7 indeterminacy of future vii, viii, ix, 1, 3, 6, 8, 11, 12, 14, 19, 21, 32, 62, 178, 248, 253 in discounted cash flow (DCF) analysis 293 and macroeconomic forecasting 129–32, 134–5, 140 ontological 6 politics as source of 7 as result of innovation vii, 3, 6, 9, 229 and volatility calculus 219–33 see also uncertainty indicative planning 41, 44–5, 47, 56 inflation targeting 176–8, 196–9, 201–2 information asymmetries vii, 5, 6, 9 information and communication technologies (ICT) 50–1 information economics vii Ingres, Stefan 183 innovation vii, ix, 2–3, 5, 20, 48, 50, 85, 130, 264, 267, 289, 298–315 alignment of expectations as precondition for 31, 300, 301–6 Lester and Piore on 299–300, 302–3 link with production 304–5 organized diversity and 300, 301–6 role of models and narratives in coordinating 21, 25, 31 role of dissonance and open-ended search in 30, 31, 301 and uncertainty vii, 2–3, 5–10, 24, 32, 39, 42, 54, 56, 63, 129–30, 223, 229, 287, 291, 302, 314, 315 see also creativity; technological development/progress institutions, role in reducing uncertainty 11–12, 44

OUP CORRECTED PROOF – FINAL, 1/6/2018, SPi

Index integrated and divided mental states, and conviction narratives 76–7 interaction in economic forecasting 109–21 external networks 111–14 formal and informal networks 116–18 internal interaction patterns 114–16 interpretation, and technological innovation 299, 300, 302–3 interests in Arctic futures 83–101 heterogeneity of 41 mutual constitution of expectations and 27, 84–7 and selective sorting of images of the future 7, 86 intuitive judgement 133, 181, 261, 268, 273, 274 Inuit Circumpolar Council (ICC) 92–3 Inuit nations, and the Arctic 89 Jacoby, Wade 303 Japan economic model 48–9, 54 indicative planning 45 judgements 10, 14, 72, 244, 315 dynamic 178 ‘holistic’ and ‘diagnostic’ 19, 244 intuitive 133, 181, 261, 268, 273, 274 role of in decision-making ix, 10, 14, 17, 64, 181, 182, 183, 240, 268, 279, 280, 294, 315 role of in forecasting 22–3, 111, 112, 126, 127, 128, 133 Kahneman, Daniel 72 Kalthoff, Herbert 242 Karlsson, Tomas 280 Karnøe, Peter 304 Karpik, Lucien 279–80 Keats, John 3, 76 Keynes, John Maynard vii, 11, 28, 144, 177 on animal spirits 21, 23, 63, 74–5, 77, 78, 279 beauty contest metaphor 64, 118–19, 120, 225 Keynesian demand management 175–6, 195–6, 197, 198, 202, 209, 211 on uncertainty 8, 31–2, 74 King, Mervyn 6, 21, 201 Knight, Frank vii, 7–8, 18, 221, 225, 228, 232, 240, 241, 248, 253, 267–8, 279, 294–5 Knorr Cetina, Karin 180, 242, 244 knowledge 32, 55, 118, 132, 175, 207, 221, 226, 239, 294 decentralized 11, 31, 55, 121, 200 diagnostic form of 267–8 Hayek on 11, 31, 195 illusion of 20 and imagination 14, 17, 261, 265–8, 274–5 indigenous knowledge and the Arctic 91, 92

and information and communication technologies (ICT) 50–1 interactional foundations of 109, 111, 112, 116, 117 Keynes on precariousness of 8 Popper on 146 pretence-of-knowledge syndrome 195, 201 risk as a category of 238, 241 role in entrepreneurship 17, 260, 262–3, 264, 270, 271, 273 role in expectations 12 subjective 72, 74 Whitehead on 260, 266–8 knowledge problems vii, 5, 121 and performativity 28 related to innovation and indeterminacy 6 related to shortcomings of knowing agents 5 see also information asymmetries; bounded rationality; framing biases Kydland, Finn E. 202 Lane, David A. 21–2 Latour, Bruno 274 lean production 49 Lehman Brothers collapse 52, 53, 130–1 Lester, Richard 299, 300, 301, 302–3, 316 Levin, Carl 250 Linnaeus, Carl 94–5 Lucas, Robert 8, 147, 149–50, 154, 199, 202 Luhmann, Niklas 239, 293 McGoey, Linsey 131 MacKenzie, Donald 20, 28, 185 macroeconomic forecasting 13, 15, 16, 124–43 dealing with ‘errors’ 124–43 forecasters’ claims of professionalism 136–9 GDP growth forecasting errors 148–9 and the ‘reality test’ 124–5, 127, 138–9, 140, 141 see also economic forecasting macroeconomic modelling 110, 144–70, 175 Dynamic Stochastic General Equilibrium (DSGE) models 16, 146–8, 154, 159, 199, 200, 211 pros and cons of standard approaches 62, 146–50, 152 rocking horse metaphor of 147, 154, 167 see also agent-based models (ABMs) macroeconomics and central bank planning 194–212 dilemma between coordination and diversity in 30, 31–2 and expectation regimes 40–1 four Ts of macroeconomic state agency 197–9 indicative planning versus delegation to markets 41, 47 new microfoundations for 63, 75–7 status of 39–40

329

OUP CORRECTED PROOF – FINAL, 9/6/2018, SPi

Index magic and economic forecasting 125, 126, 127, 133, 135, 137–8, 139, 140 emphasis on following rituals 127, 129, 135 Hubert and Mauss theory of 125, 126, 127, 139 and ‘will to believe’ 16, 125 Malthus, Thomas 48 market instability, causes of 20, 24, 29, 31, 39, 51, 56, 58, 63, 76, 77–8, 231 market internal risk 230 market value 8, 56 contingency of 10–12 versus net present value in discounted cash flow models 282–3, 284–5 markets calculative devices as constitutive of 18–20 nature of uncertainty in innovative markets 5–10 Marx, Karl 42 Mauss, Marcel 125, 126, 127 Maxfield, Robert R., 21–2 maximization 1, 2, 199, 202, 285–6 see also optimization Mayo, Elton 263 McCloskey, Deirdre N. 187 media, and dominant narratives 47 memory, and imagination 107 mental time travelling 106, 107–8, 111, 113, 115, 116 Merquior, J.G. 28 microfoundations of economic models alternative version for conditions of uncertainty vii, 10, 32, 40, 63, 75–8 standard 8, 74, 147, 152, 202 mimetism, rational 55–6 MINSKY model 151 models climate change 15 ‘communism of’ 199–201, 211 decision-making models 72–4, 287 Diamond on 30 as shared fictions 228–9 and imaginaries 13–14, 16–17 incompleteness of 30, 129 performative impact of 15, 20, 27, 28, 29, 44, 185, 196, 226–7, 229, 241 pluralism of versus shared 30–1 shared mental models 29, 31, 32 use as diagnostic tools 10, 12, 14–15, 16–17, 30 see also agent-based models (ABMs); credit risk models; discounted cash flow models; economic models; finance models monetary policy see central banking money management, and conviction narratives 63–7

330

monocultures, analytical or methodological 29, 76, 145, 148, 200 destabilizing impact of ix, 24, 63, 78 Morson, Gary Saul 23 Muniesa, Fabian 279 Muth, John 199 narrative economics 20–8 and central banks 22, 174 Shiller on 23–4 narrative framing 69 narratives viii, 3, 4, 5, 7, 20–8, 32, 33, 68–74 of the Arctic future 84, 87, 88 contingency of 29, 178 as coordination device 5, 24–8, 71, 75, 85, 177 contagion of 12, 24, 77 and central banking 22, 26, 173–90, 207 and decision-making 68–9 and fictional expectations 10–12 and forecasts 136, 137, 139 functions of 21–5, 69–72 and group emotions 4–5, 23, 71, 74, 76–7 hegemony of a single narrative 55–6 and imaginaries 3, 11–12, 51 and institutions 11–12 making sense of uncertain futures 21, 22, 23, 54–5, 58, 68–9 performativity of viii, ix, 28–9, 178 and politics viii, 7, 24, 26–8 and power ix, 4, 24–8, 87 providing confidence and conviction viii, 23, 24, 71, 73, 74 providing logic of action or script 21, 25, 71, 72 quantitative study of 23, 77 revision of as cause of instability 24, 40 and socio-economic regimes 24, 39–61 as self-fulfilling prophecies 28, 47, 51, 57 socialization of expectations by 46–8 widely shared ix, 4, 11, 12, 23, 29, 46–8 see also Conviction Narrative Theory (CNT); discourse analysis; fictions; grand narratives; new era stories; stories natural sciences versus social sciences 145–7 see also Cartwright, Nancy negative capability 3, 76, 299 neoclassical economics 27–8, 145, 175, 196–7, 200 neoliberalism 41, 45–6 net present value (NPV) 17, 278, 282, 285, 288, 290 neuroscience 23, 62, 70 and brain imaging 107 new economy narratives 47, 50–1, 54 new era stories viii, ix, 9, 21, 24, 41

OUP CORRECTED PROOF – FINAL, 9/6/2018, SPi

Index Newtonian physics, and macroeconomic modelling 146, 147 New Zealand, Reserve Bank of 176–7, 200 Nixon, Richard 307 non-ergodic futures 9, 10, 19, 106, 114, 118, 121, 130 non-normal behaviour, and agent-based modelling 154, 158 North, Douglass 2–3, 11 North Pole 83 see also Arctic, the Norway, and the Arctic 89, 90 novelty ix, 14, 85, 111, 121 as cause of indeterminacy and uncertainty vii, 3, 6, 7, 10, 32 as key feature of capitalist systems vii, 2, 42, 85 see also innovation oil crises (1970s) 48, 53, 306, 310 OMTs (outright monetary transactions) 184 ontological indeterminacy or uncertainty 6, 21, 62, 128, 129–32, 240, 281 open versus closed future 1, 2, 20, 27, 86, 87, 97–8, 219–21, 227–8 optimization 2, 153 options markets 41, 223, 224–5, 226–7, 230–1 Orléan, André 125 Paris Conference on Climate Change 53 Parker, Martin 13 perfect-competition hypothesis 132 performativity and central banks 176, 178, 185, 195, 199–203, 205, 210, 211 and counter-performativity 29, 226–7, 229–33 and expectations 28–30, 185 limits of 28–32 MacKenzie on 28, 185, 227 of theories and models 28, 185, 227, 229 and volatility calculus 226–7, 229 Persons, Warren 131 pharmaceutical industry, use of discounted cash flow 17, 281, 287–93, 294, 295 photovoltaics 309–12, 315 physics, and macroeconomic modelling 146, 147 Pill, Huw 209 Pilmis, Olivier 13, 16, 25, 124–43 Piore, Michael 299, 300, 301, 302–3 Polanyi, Karl 7 Polar Year, the 93, 96 policy analysis, and macroeconomic modelling 147–8

politics as battle between narratives 26–7, 41 as cause of indeterminacy or uncertainty 7, 309 of expectations 28, 29, 84, 86–7 political instability 24 Pollard, Sidney 279 Popper, Karl 146 portfolio insurance systems 231 post-war Golden Age and indicative planning 41, 44–5 reinterpreted 49 power economic 3, 86 market ix, 67 narratives and theories as instruments of viii, 4, 24–8 political ix, 3, 26–7, 86, 97, 114 Power, Michael 18, 241, 242, 280 Praet, Peter 207 prediction and Conviction Narrative Theory 69, 70, 72 Gramsci on 194, 208 impossibility of at macroeconomic level viii, 15–16, 106, 124, 125, 131 ‘informed’ predictions 8, 199 judgement rather than mechanical prediction ix, 14, 22 to manage uncertainty or foreclose open future 85, 86, 98 pattern predictions 10, 14, 15, 22, 133 and reflexivity 231 see also economic forecasting, probability assessments or forecasts Prescott, Edward C. 202 probability assessments or forecasts 2, 8, 14, 18, 19, 46, 221, 290 normal distribution versus fat tails in 51, 154, 158, 161–2, 167, 230–1 see also ergodic futures project appraisal or evaluation, in venture capital 12, 259–76 Prometheus unbound, economic vision of 48 promissory stories 21, 25 psychology vii, 11 and Conviction Narrative Theory 23, 62, 63, 69, 71, 72–4, 78 and mental time travelling 108 public choice theory 27–8 quantitative easing (QE) 26, 32, 40, 53, 178, 183–4, 201, 203–5, 208–10, 211 see also central banking quants 51, 64, 237–8

331

OUP CORRECTED PROOF – FINAL, 9/6/2018, SPi

Index Rajan, Raghuram 26, 174, 201 rational expectations theory/hypothesis vii, 5, 8, 9, 10, 24, 32, 39, 40, 43, 55, 75 and central banking 177–8, 199–203, 211 and the efficient markets hypothesis 8–9 as opposed to fictional expectations theory/hypothesis 10 reason relationship to imagination 4–5 see also calculation, rational reasoning imagination 4, 13–14, 261 Rebonato, Riccardo 14 reflexive modelling 12, 229 reflexivity 29, 64, 77, 202, 229–33 regulation theory 44 Reichmann, Werner 13, 16, 22, 105–23 Reis, Ricardo 177 rhetoric viii, ix, 10, 12, 22, 26, 51, 174, 177–8, 208, 307 Riksbank, Sveriges 181–4 risk assessment of as social or market practice 241–52, 288–90 cardinal measures of 19, 20 as category of knowledge 241 as danger 245 diagnostic versus technical conceptions of 236–9, 244–50 Knight on 7–8, 18, 221, 225, 232, 240, 241, 248, 253, 279, 294–5 measurable risk versus radical uncertainty vii, 7, 8, 14, 19, 54, 153, 199, 221–2, 223, 226, 229, 233, 240 ordinal rankings of 19, 244 ‘risk taking channel’ 209 in sociological literature 232, 239–42 unpredictability of 230–1 see also credit risk models; probability assessments risk management viii, 19, 20, 238, 241–2 risk premium 291 Rundrechnung, and economic forecasts 115–16 Russia, and the Arctic 89, 90, 94, 97 Sami people 90, 95–6 Sargent, Thomas 199 Savage, Leonard J. 75 scenario analysis 18, 27, 32, 96, 291, 292, 295 Schacter, Daniel L. 70 Schumpeter, Joseph A. 2, 42, 264, 267 Science and Technology Studies (STS) 25, 86 and Arctic futures 85 Shackle, George vii, 2, 3, 5, 9–10, 11, 144–5, 167 Shapin, Steven 261 Shapiro, Morton 23 shared mental models 29 Shiller, Robert J. 23–4, 77

332

shocks to economy 24, 43, 147, 150, 165–6, 195, 230, 245 endogenous disequilibrium 2, 6, 15, 32, 129, 130, 152, 161, 239 external or exogenous 7, 32, 129, 130, 153, 160 and macroeconomic forecasting 16, 129–31 Smaghi, Bini 208 Smart, Graham 180–1 Smelser, Neil J. 75–6 social sciences, versus natural sciences 145–6 social foundations of expectations and imaginaries 11–12, 25, 98, 105, 118, 119, 136, 302–3 socio-economic regimes 39–58 chronology of imaginaries and expectation regimes 54 and grand narratives 24, 48–56 see also expectation regimes sociology viii, 5, 16, 44, 62, 63, 68, 71, 75, 85, 124, 125, 136, 233, 237, 238, 266, 267, 278, 279–80, 301–3, 305 risk in sociological literature 232, 239–42 Sociology of Expectations 25 Sörlin, Sverker 88 Sutton, John 14 Stark, David 30, 302, 315 stewardship, and Arctic futures 84, 92–3 Stinchcombe, Arthur 274 stories viii, 20–4, 25, 39, 41, 43, 47, 66, 87, 88, 107, 108, 177, 188, 190 contingency of 11 and economic forecasts 16, 134 monetary policy stories 178–9, 180–5 see also narratives; new era stories; promissory stories structured finance 220, 221, 222, 233 and credit rating 236, 237, 238–9, 243–4, 245–52, 253 and production of unpredictability 227–9 see also volatility calculus Suddendorf, Thomas 107–8, 109 Svensson, Lars 181–2 Sweden as Arctic nation 27, 84, 87, 88, 89, 90, 93–4, 96–7 Arctic Strategy 94–7 Riksbank 181–4 synthetic fuels 312–14 Taleb, N.N. 124 Taylor, John 195 technological determinism 49 technological development/progress 25, 289, 300, 301, 302, 303, 305, 314 distribution of resources for 25, 298, 306

OUP CORRECTED PROOF – FINAL, 9/6/2018, SPi

Index information and communication technologies (ICT) 50–1 and shared imaginaries viii, 49, 270, 305 threat to the social system 267 upscaling and 300, 304, 310, 311 see also energy technology development in the US; innovation technological innovation, institutional conditions for 298–300, 305 Tetlock, Philip 127 Tomasello, Michael 69 Tuckett, David 23, 29, 30, 62–82 Turner, Adair 19 ‘two cultures’ divide 5 uncertainty 1–2, 5–10 and capitalism 2–3, 5–10, 32, 41–3, 252 central bank attempts to reduce 182, 194, 201, 207–8 and contingency of expectations vii, viii, 6, 11, 25, 29, 32, 42 and coordination problems 9, 25, 194–5 decision-making and vii, viii, 3, 14, 32–3, 194, 261, 262 and diagnostic knowledge 14, 268 and discounted cash flow analysis 278–95 in complex systems 6, 145, 152–3 expressed as cost of capital 281, 290–2 fictional expectations and 10–11, 21 and financial markets 42–3, 46, 227–9 first-order and second-order 6, 25, 54, 56 and globalization 24, 43 ignored by standard economics 8, 19, 43, 199 ignored by structured finance rating models 248, 253 imaginaries as tool to cope with ix, 3, 4, 5 and innovation vii, 2–3, 5–10, 24, 32, 39, 42, 54, 56, 63, 129–30, 223, 229, 287, 291, 302, 314, 315 institutions as partial solution to 11–12, 44 judgement, reliance on in conditions of ix, 10, 14, 19, 23, 182, 240, 259, 279, 294 Keynes on 8, 31–2, 74 King on 6 Knight on 7–8, 18, 221, 232, 240, 241, 248, 268, 279, 294–5 and macroeconomic modelling 144–67 as multiplicity of possibilities 292–3 narratives as help to overcome 21, 22, 23, 54–5, 58, 68–9 ontological 6, 21, 62, 129–32, 240, 281 politics as source of 7, 309 and probabilities 221, 288–90 radical vii, 7, 21, 23, 39, 40, 44, 56, 58, 62–3, 68, 69, 71–2, 74, 75, 106, 153, 221–2

rational calculation in conditions of 9, 14, 21, 23, 51, 65, 74, 221, 232, 261, 279 Shackle on 2, 3, 5, 9, 145 as source of profit 8, 221, 222 technical 288 versus measurable ‘risk’ vii, 7, 8, 14, 19, 54, 153, 199, 221–2, 223, 226, 229, 233, 240 see also indeterminacy of future United Kingdom (UK) Brexit 16, 26–7, 133–4 GDP growth 148–9, 154–8 housing market 158–62 United Nations (UN) Copenhagen Climate Change Conference 91 Law of the Sea 88–9 United States 109, 131, 250–1 and the Arctic 90, 94, 96, 97 energy technology development 25, 31, 298–315 ERDA (Energy Research and Development Administration) 307–9, 310 industrial decline 48–9 information and communication technologies (ICT) 50 US Federal Reserve 26, 174, 203, 204 Value at Risk models 19 van Lente, Harro 305 Vanberg, Viktor 5, 6 venture capital and administrative proccedures 17, 269, 271–5 project appraisal in 17, 259–76 role of knowledge versus imagination 261–2, 274–5 volatility advanced 225 historical 224, 225 implied 220, 225–6, 229 and the unpredictability of risk 230–1 the ‘volatility smile’ 230, 231, 233 volatility calculus 20, 219–33 and the production of unpredictability 227–9 Walrasian theory 40 water lily model of economic growth 48 Weber, Max 4, 275 Weir, Margaret 305 Whitehead, Alfred North 260, 266–7 Wicksell, Knut 177 Woodford, Michael, Interest and Prices (MWIP) model 178–9 Yellen, Janet 201 Ziegler, Nick 304

333