Behavioral Public Economics: Social Incentives and Social Preferences 0367362414, 9780367362416

Behavioral Public Economics shows how standard public economics can be improved using insights from behavioral economics

812 143 13MB

English Pages 218 [219] Year 2021

Report DMCA / Copyright


Polecaj historie

Behavioral Public Economics: Social Incentives and Social Preferences
 0367362414, 9780367362416

Table of contents :
Half Title
Title Page
Copyright Page
Table of Contents
List of Figures
1 Introduction: Why ‘Behavioral’ Public Economics?
1.1 Public Economics and Policy
1.2 Behavioral Economics and Policy
1.3 Why Social Preferences Matter?
1.4 From Economic Incentives to Social Incentives
2 Preferences, Utility, and Welfare
2.1 Preferences and Choices
2.1.1 Revealed Preference Theory
2.1.2 Critique of Revealed Preference Theory
2.1.3 Behavioral Economics and Revealed Preference Theory
2.2 Utility
2.2.1 What Is Utility?
2.2.2 Expected Utility Theory
2.3 Welfare: Personal and Social
2.4 Well-Being and Happiness
3 Economic Incentives in Public Economics
3.1 Taxation
3.2 Sin Taxes
3.3 Imperfect Information and Contracts
3.4 Tax Compliance
3.5 The Environment
4 Behavioral Economics and Public Policy
4.1 Bounded Rationality
4.1.1 Herbert Simon and Bounded Rationality
4.1.2 Routine
4.1.3 Selective Rationality and X-Efficiency
4.1.4 Ecological Rationality
4.2 Perception and Entrepreneurship
4.2.1 Uncertainty
4.2.2 Schumpeterian Entrepreneurship
4.2.3 Kirznerian Entrepreneurship
4.3 Errors and Biases
4.4 Gains and Losses
4.5 Nudges
4.6 Addiction and Self-Control
5 Social Preferences and Moral Economy
5.1 Social Incentives and Social Preferences
5.2 Social Preferences and Fairness
5.3 Altruism and Envy
5.4 Public Goods and Charitable Giving
5.5 Philanthropy and Corporate Social Responsibility
6 Social Incentives and Interaction
6.1 Society and Coordination Problems
6.2 Social Norms
6.3 Norm Compliance
6.4 Identity and Culture
6.5 Social Preferences and Internal Moral Constraints
7 Governing the Commons With Social Incentives
7.1 The Tragedy of the Commons
7.2 Social Preferences and Environmental Sustainability
7.3 Internal Moral Constraints and Environmental Sustainability
7.4 From Homo Economicus to Homo Moralis

Citation preview


Behavioral Public Economics shows how standard public economics can be improved using insights from behavioral economics. Public economics typically lists four market failures that may justify government intervention in markets—​imperfect competition (or natural monopoly), externalities, public goods, and asymmetric information. Under the rational choice paradigm (‘agents choose what is best for them’), public economics has examined the welfare effects of policy. Recent research in behavioral economics highlights a fifth market failure—​individuals may make mistakes in pursuing their own well-​being. This book calls for a rethinking of assumptions of individual behavior and provides a good foundation for public economic theory. Key features: 1. Introduces behavioral perspectives into public economics. 2. Explains why economic incentives often undermine social preferences. 3. Reveals that social incentives matter for public policy. This book will be an invaluable resource for researchers and postgraduate students in public economics, behavioral economics, and public policy. Shinji Teraji is Professor of Economics at Yamaguchi University, Japan. His research is mainly concerned with behavioral economics, institutional economics, and economic methodology.

BEHAVIORAL PUBLIC ECONOMICS Social Incentives and Social Preferences

Shinji Teraji

First published 2022 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN and by Routledge 605 Third Avenue, New York, NY 10158 Routledge is an imprint of the Taylor & Francis Group, an informa business © 2022 Shinji Teraji The right of Shinji Teraji to be identified as author of this work has been asserted by him in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. British Library Cataloguing-​in-​Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging-​in-​Publication Data Names: Teraji, Shinji, author. Title: Behavioral public economics : social incentives and social preferences / Shinji Teraji. Description: Milton Park, Abingdon, Oxon; New York, NY: Routledge, 2022. | Includes bibliographical references and index. Identifiers: LCCN 2021017264 Subjects: LCSH: Economics–Sociological aspects. | Economics–Psychological aspects Classification: LCC HM548.T468 2022 | DDC 306.3–dc23 LC record available at ISBN: 9780367362409 (hbk) ISBN: 9780367362416 (pbk) ISBN: 9780429344817 (ebk) DOI: 10.4324/​9780429344817 Typeset in Bembo by Newgen Publishing UK


List of figures  Preface  1 Introduction: why ‘behavioral’ public economics?  1.1 1.2 1.3 1.4

vii ix 1

Public economics and policy  1 Behavioral economics and policy  7 Why social preferences matter?  14 From economic incentives to social incentives  20

2 Preferences, utility, and welfare 


2.1 Preferences and choices  25 2.1.1 Revealed preference theory  25 2.1.2 Critique of revealed preference theory  29 2.1.3 Behavioral economics and revealed preference theory  32 2.2 Utility  35 2.2.1 What is utility?  35 2.2.2 Expected utility theory  38 2.3 Welfare: personal and social  41 2.4 Well-​being and happiness  46

3 Economic incentives in public economics  3.1 3.2 3.3 3.4 3.5

Taxation  52 Sin taxes  57 Imperfect information and contracts  62 Tax compliance  66 The environment  70


vi Contents

4 Behavioral economics and public policy 


4.1 Bounded rationality  77 4.1.1 Herbert Simon and bounded rationality  77 4.1.2 Routine  82 4.1.3 Selective rationality and X-​efficiency  84 4.1.4 Ecological rationality  85 4.2 Perception and entrepreneurship  88 4.2.1 Uncertainty  88 4.2.2 Schumpeterian entrepreneurship  89 4.2.3 Kirznerian entrepreneurship  90 4.3 Errors and biases  93 4.4 Gains and losses  95 4.5 Nudges  98 4.6 Addiction and self-​control  102

5 Social preferences and moral economy  5.1 5.2 5.3 5.4 5.5

6 Social incentives and interaction  6.1 6.2 6.3 6.4 6.5


Society and coordination problems  138 Social norms  141 Norm compliance  151 Identity and culture  161 Social preferences and internal moral constraints  165

7 Governing the commons with social incentives  7.1 7.2 7.3 7.4


Social incentives and social preferences  108 Social preferences and fairness  112 Altruism and envy  115 Public goods and charitable giving  122 Philanthropy and corporate social responsibility  125


The tragedy of the commons  171 Social preferences and environmental sustainability  173 Internal moral constraints and environmental sustainability  181 From Homo economicus to Homo moralis  186

Bibliography  Index 

192 202


1 .1 1.2 5.1 6.1 6.2 6.3 6.4 6.5 7.1 7.2 7.3 7.4 7.5 7.6

Negative externality  Demand curves for light and heavy drinkers  Population dynamics  Payoff matrix  Two utilities  The relationship between b and p  Increase in p  Decrease in p  An equilibrium: 1 –​xk and 1 < (1/​n)F  No equilibrium: 1 –​xk and (1/​n)F < 1  An equilibrium: 1 –​xk and (1/​n)F < 1  An equilibrium: 1 –​m –​ y(k –​ m) and 1 < (1/​n)F  No equilibrium: 1 –​m –​ y(k –​ m) and (1/​n)F < 1  An equilibrium: 1 –​m –​ y(k –​ m) and (1/​n)F < 1 

5 6 119 143 144 157 158 159 177 178 178 183 184 184


In essence, economics is about behavior. Economics has enhanced our understanding of how people make decisions and behave. Neoclassical economic models appeal to axioms of rationality in which individuals have stable, consistent, and context-​ independent preferences. However, people’s actual choices do not appear to be explained by neoclassical economic models. People do not come into the world with immutable preferences, do not always act rationally in the neoclassical economic sense, and sometimes cannot handle all the information that comes their way. Behavioral economics tries to offer a potentially richer set of tools to understand individual behavior than does neoclassical economic theory. Rather than assuming that people generally know what is best for them and make decisions consistent with that knowledge, behavioral economics acknowledges that people often do not act rationally in the traditional economic sense, make myopic decisions based on an inadequate understanding of the alternatives, and do not necessarily learn from their mistakes. Work in behavioral economics suggests that there are systematic errors that people regularly make because they are ‘human.’ Behavioral economics has pointed out the limits on human behavior: bounded rationality, bounded willpower, and bounded self-​interest. By assuming a more realistic picture of individual behavior, behavioral economics makes it possible to alter the design of effective public policy under conditions of bounded rationality, bounded willpower, and bounded self-​interest. The aim of this book is to provide an analytical framework for ‘behavioral’ public economics. Behavioral economics can provide valuable insights on psychological regularities. These insights can be used to increase effectiveness of interventions in public policy. Simply not assuming that people are making choices that are in their own best interests opens up a world of alternative policy interventions beyond just manipulating prices via taxes and subsidies. Behavioral public economics is the set of attempts to modify standard public economics so that it is better aligned with

x Preface

insights from behavioral economics. This book systematically assesses the meaning of these behavioral insights for policy prescription. First, this book revisits many results in standard public economics that are based on the standard model of choice and considers the robustness of these results to an alternative model of behavior that grounds on more realistic psychological foundations. Second, this book argues that, by starting with a model of behavior that grounds on more realistic psychological foundations, we may be able to provide an expanded toolkit of policy instruments. A behavioral perspective presents a new challenge for economic analysis.This book is effective in showing how standard public economics can be improved using insights from behavioral economics. The main contribution of this book is to provide the implications of social incentives and social preferences for public policy. This book calls for a rethinking of assumptions of individual behavior and provides a good foundation for public economic theory. Human behavior is not wholly explicable in terms of rationality and selfishness. The implications of such behavior for public economics are examined in this book. The coronavirus disease 2019 (COVID-​ 19) pandemic represents a massive global health crisis.To decrease the transmission rate of COVID-​19, a massive public health campaign is under way: increasing handwashing, reducing face touching, wearing masks in public, and social distancing. The COVID-​19 pandemic may be one of the greatest societal challenges that require large-​scale cooperation.The capacity to cooperate in large groups with nonrelatives has enabled humans to develop social networks, build public works, and create social institutions. Cooperation requires people to bear an individual cost to benefit others. It is supported by social preferences, including a concern for the welfare of others and a preference for equity, which mediates conflicts between self-​interest and other-​interest. People are more likely to cooperate when they believe that others are cooperating. Human cooperation is regulated by social norms that establish standards for how people should behave in particular situations. Economics is not about private profit only. Institutions have to be constructed to reconcile, as far as possible, the interests of the individuals with the social interest. Recent developments in behavioral economics focus on altering some of the cornerstones of economics and opening new areas of research.The Homo-​economicus model in economics has crowed out consideration of important prosocial aspects of human nature. People, in various cases, behave as if they are intrinsically motivated rather than stimulated by any extrinsic, monetary reward. The introduction of Homo moralis into economics enables us to consider empathy, altruism, and trust as important factors of decision-​making. This book is divided into the following seven chapters: Chapter 1 sets out economic and social problems that are discussed in this book. Many economists recognize that there are situations in which competitive markets generate unsatisfactory outcomes (market failures) and government intervention can lead to better outcomes. Public economics typically lists four market failures that may justify government intervention in markets: imperfect

Preface  xi

competition (or natural monopoly), externalities, public goods, and asymmetric information. Recent research in behavioral economics highlights a fifth possible market failure—​individuals may make mistakes in pursuing their own well-​being. Systematic errors constitute a behavioral market failure, which brings us to the realm of behavioral public economics. What makes people better off is not the satisfaction of revealed preferences, but true preferences which may not always be observed through choice. Chapter 2 discusses preferences and choices in economics. Conventional economic approach relies on the theory of revealed preferences, in which the concept of preferences is entirely defined in terms of choice and has no psychological meaning. Behavioral economists have questioned the satisfaction of revealed preferences as the normative foundation of welfare economics. Psychological factors, such as limited attention, inferior cognitive abilities, or lack of self-​control, can operate against the satisfaction of individual’s true preferences. If individuals make mistakes, and if these mistakes are systematic, it might be possible to make them better off by restricting their choices and putting attractive bad choices off limits. Thus, it is feasible to improve welfare by restricting or altering individual choice. If the true preference set is reconstructed, the satisfaction of these preferences could be used as a basis for welfare-​enhancing policy proposals. Chapter 3 covers economic incentives (taxes, fines, and rewards) in standard public economics. Tax systems have been discussed in terms of at least three different criteria: (i) the need for taxes to be fair, (ii) the need to minimize administrative costs, and (iii) the need to minimize disincentive effects. The trade-​offs between efficient and equitable outcomes are evaluated by a social welfare function in optimal taxation theory. The social welfare function may take various forms, depending on society’s taste for equity.Tax systems have to achieve a range of public policy objectives in a complex and changing environment. In the rational choice approach, tax compliance is in one’s self-​interest. Policymakers try to raise the perceived costs of violation by increasing the certainty of detection and strengthen the severity of punishment.Thus, the rational choice approach predicts that the frequency of violation will decrease if the perceived costs associated with offending decisions are increased. Chapter 4 deals with behavioral economics as policy tools. As a result of behavioral biases, people fail to satisfy well-​defined preferences. Public policy is directed at better preference satisfaction. Behavioral economics is fundamentally concerned with the questions of how people actually behave in decision-​making situations and how their choices can be improved so that their welfare is enhanced. Preferences are affected by whether the decision problems are framed in terms of gains or losses. A framing effect is said to be present when different ways of describing the same choice problem change the choices that people make. Systematic biases lead to poor choices. There is a scope for policymakers to correct the cognitive shortcomings. Nudges are designed to effect behavioral change on people. Nudges tend to involve relatively small additions or changes to the decision environments that encourage, but do not force, changes in behavior.


xii Preface

Chapter 5 explains why social incentives and social preferences matter in behavioral public economics. A factor that might increase the performance of prosocial activities (e.g., volunteering, civic duty, charitable donations, or other social contributions) concerns people’s incentives. People exhibit prosocial behavior when they do not always make choices only based on economic incentives. Extrinsic rewards to encourage prosocial behavior can sometimes backfire and decrease the desired behavior by crowding out intrinsic motivation to act altruistically. In the case of blood donations, introducing payments led to a decrease in giving. Economic incentives can manipulate people into being more selfish. Social incentives will give prominence to the role of nonpecuniary drivers of prosocial behavior. Ethical and moral aspects cannot be ignored even in a market economy. Chapter 6 argues that social norms bring about social incentives. Social norms are shared understanding about actions that are obligatory, permitted, or forbidden. Unlike legal rules, social norms are not supported by formal sanctions. Why do social norms not simply collapse from the violation? This chapter examines several mechanisms on norm compliance. The incentive to comply with social norms derives not only from the enforcement of costly punishment by others but also from reputation building for oneself. People are concerned about social approval or disapproval in their interactions with others. Social norms are sustained through social emotions such as shame, guilt, or embarrassment. Social norms can direct people to undertake actions that are inconsistent with selfish actions. People may feel ashamed from acting in their self-​interest, which fosters prosocial behavior. Chapter 7 discusses the issue of governing the commons with social incentives. Common-​pool resource exploitation is known as the tragedy of the commons.This metaphor is based on the assumption that human behavior is driven only by self-​ interest. Selfish users are unwilling to pay the costs of resource conservation because the benefits of doing so are shared collectively. This chapter offers an explanation for the conservation of a common-​pool resource, namely individuals who consider the resource conservation to be their major concern are driven by social incentives. Besides material payoff, the pro-​environmental identity obtains psychological payoff related to the resource conservation. Further, this chapter considers the interplay between the pro-​environmental identity and the morally constrained one. The individual with internal moral constraints follows a moral rule imposing an extra cost on the resource extraction. A society can be sustainable to the degree that its members embody moral preferences and regulate themselves by moral rules. I have benefited from ideas and comments of many colleagues and experts in the fields. I am grateful to Morris Altman, Mark Pingle, and Mehmet Tosun for their thoughtful remarks and suggestions. The late John Tomer encouragingly accepted my idea to publish a book on behavioral economics and public policy. I would like to thank Yongling Lam, my editor at Routledge, who persevered with me in making this book possible. Shinji Teraji Yamaguchi March 2021

1 INTRODUCTION Why ‘behavioral’ public economics?

1.1  Public economics and policy Most economists hope that their research will have an impact for public policy. A market system, as the fundamental task, determines which goods and services are produced, how they are produced, and who gets them once they have been produced. Yet, many economists recognize that there are some situations in which competitive markets generate unsatisfactory outcomes (‘market failures’) and government intervention can lead to better ones. Public economics indeed has provided policymakers with vital tools. A major concern of public economics is to remove preexisting distortions, recognizing the limitation of free markets. Public economics is expected to provide us with objective economic analysis and advice on the development and implementation of public policy issues. It provides causal knowledge of the consequences of policies to enable policymakers to choose effective means toward their ends. A useful framework is offered for thinking about taxes, subsides, transfer programs, regulation, health-​care reform, the social security system, and so on. The success of public economics demonstrates that policymakers are willing to look to academic fields for guidance in setting their policies. In public economics, the concept of Pareto efficiency (or Pareto optimality) is introduced as an essential criterion to evaluate public policies. A discussion is then given on how markets achieve Pareto efficiency, in the sense that no resources are wasted and all Pareto improvements have been exhausted.The possibility of making one person better off without making someone else worse off implies the existence of a potential Pareto improvement. A Pareto-​efficient allocation is thus one from which there exists no potential Pareto improvements. Public economics evaluates whether an allocation is ‘better’ or ‘worse’ for a particular ordinal utility function from modern demand theory. The first and second fundamental theorems link the concept of a Pareto-​efficient allocation to the Walrasian competitive equilibrium. DOI: 10.4324/​9780429344817-1

2 Introduction

The first fundamental theorem states that every competitive equilibrium is Pareto efficient. The second fundamental theorem states that any Pareto-​efficient allocation of resources can be achieved by a combination of competitive equilibrium and some set of lump-​sum transfers (taxes and/​or subsidies). Public economics typically lists four (traditional) market failures that may justify government intervention in markets—​imperfect competition (or natural monopoly), externalities, public goods, and asymmetric information. The invisible hand needs a helping hand. Many industries exhibit multiple market failures. An efficient market requires a large number of buyers and sellers who are able to compete on price and quantity terms in the market. If there is only one producer (a monopoly) or a few producers (an oligopoly) in the market, there will be attempts to extract profits by setting prices higher than in a competitive market. This will be a major source of market inefficiency. The water industry is an example of a natural monopoly. Raw water is treated to remove natural and man-​made pollutants in order to make it fit for consumption. Water quality is most important for consumers’ health, but consumers cannot inspect the quality before drinking. There is tight regulation of standards. Regulation of prices would be simple if the regulator had as much information as the firm about demand and cost conditions. In practice, the regulator is unlikely to know as much as the firm about such conditions. The asymmetry of information exists between the firm and the regulator. We cannot expect a naturally monopolistic industry to have the correct incentive to increase the quality of water. If the price is fixed, the firm sets a low-​quality level for the given price. Furthermore, extra pollution in the water environment increases the costs of meeting a given standard of water quality. The information people possess affects their behavior in many situations. An efficient market requires complete, unbiased information. A form of the imperfect information problem arises when parties to a transaction have access to different levels of information. An informed market participant can make profits at the expense of other market participants who have limited information. This will produce sub-​optimal market outcomes. Consumers are often poorly informed about market conditions and product characteristics. Consumers may have great difficulty in evaluating the post-​purchase performance, based on observable product characteristics. The so-​called consumer protection laws are essentially justified on the argument that firms as compared to consumers are much better informed about the nature of goods they produce and, hence, may abuse their information superior position to the detriment of the consumers.The theory of asymmetric information has indeed transformed the way economists think about the functioning of markets. Information asymmetry is a cause of market failure and therefore necessitates state regulation of contracts. If improving the information held by consumers is difficult or costly, the problem of adverse selection may arise. That is, given competitive markets, producers may be unable to provide desirable products since consumers are unable to observe their superior characteristics prior to sale.

Introduction  3

If there are market failures, government policies are required to eliminate the failures by providing public goods or taxing external effects. Pure public goods have two properties: non-​excludable and non-​r ivalrous.1 Usually, non-​excludability means that a price cannot be charged for the good by restricting its use to those who pay for it, and non-​r ivalness means that the marginal cost of adding an additional consumer is zero.2 Then, people can contribute to the provision of the public good and enjoy its benefit, or they can keep their money in their pockets and enjoy its benefit anyway. Total contributions might be relatively small, and only a small quantity of the public good might be ultimately provided. Most people still rely on the government to provide public goods. They face several problems in their daily lives: educational facilities, medical facilities, libraries, drinking water, roads, electricity, pollution, crime, and so on. National defense is the collective provision of protection against external threats, laws and their enforcement against internal threats, and fire departments against fires. Clearly, if most of what one produces were stolen without compensation, no one would have incentives to produce to begin with. Therefore, the provision of law and order is essential for the market to function. Conventionally, it has been argued that the optimal provision of public goods is determined by the ‘Samuelsonian’ condition—​equality between the marginal rate of transformation and the sum of the marginal rates of substitution between a public and a private good (Samuelson, 1954). This condition lies at the foundation of normative public expenditure theory. Market mechanisms do not provide the required incentives in choosing an optimal provision of public good. Each individual is better off if all contribute to the provision of the public good than if all do not, and each is still better off if he or she does not pay for the good. For a private good, a consumer reveals one’s preference for it by being willing to buy a certain amount at a certain price. Unlike private goods, individuals have little incentive to voluntarily provide public goods when they can simply enjoy the benefits of public goods provided by others. The ‘free-​rider’ problem belongs to the sphere of public economics. If user charges are based on reported preferences, there is an incentive to underreport. Preventing such under-​provision of public goods is one of the primary economic rationales for government. Government intervention is usually required for the efficient allocation of public goods. The same rationale applies to environmental protection. Because individuals and firms face free-​r iding incentives for protecting the environment, public policies are often put in place to limit pollution, restrict resource exploitation, or create the right incentives to promote (or protect) environmental quality. On the other hand, many goods supplied or subsidized by governments may be supplemented by private-​market purchases: police protection may be augmented by private security services, and government-​funded health care may be also supplemented by private purchases. The provision of public goods by the government raises the questions whether they are provided efficiently and whether their provision justifies the costs involved. These questions arise in considering government provision of goods when private provision is also feasible. Especially, it has been pointed out that the problems of

4 Introduction

government failure—​associated with, for example, rent seeking, corruption, and inefficiency—​should be set beside those of market failure in assessing the desirability of government action. It may be argued that government failure is so severe in developing countries that the level of government action should be lower than in developed countries. Economic analysis of environmental policy is based on the idea that the potentially harmful consequences of economic activities on the environment constitute an externality. An externality is an economically significant effect of an activity, the consequences of which are borne (at least in part) by a party or parties other than the party who controls the externality-​producing activity. Appropriate compensation induces the generator of the externality to take into account the effects of one’s actions on others. An externality only occurs when appropriate monetary compensation is not made. Burning carbon has an external cost because it produces CO2 and other greenhouse gases that accumulate in the atmosphere. Eventually, unwanted climate change—​higher temperatures, climate variability, and possibly increases in sea levels—​will occur. This external cost is referred to as the social cost of carbon.When an externality is present, there is a divergence between private and social cost. Without interference in the market mechanism, some transactions that would be beneficial are not carried out. There still remain some interactions that ought to be internalized but which the market mechanism cannot cope with. Why is the market mechanism unable to make the emitter of an externality internalize the costs of one’s actions? The reason why self-​interested economic agents do not undertake these transactions is that the cost of carrying out the actual transactions is greater than the expected benefit. If market forces are, by themselves, unable to eliminate the remaining inefficiencies, government policies are required to eliminate them. This incentive-​based approach is well known as the Pigouvian tax scheme. The idea is levying a tax on an externality-​generating activity equal to its marginal social damage. This is a first-​best remedy which, in the absence of other distortions in the economy, moves the competitive equilibrium of the economy to its Pareto-​efficient frontier. Consider the following standard Pigouvian story, using Figure 1.1. Drivers produce one dollar’s worth of externalities for every mile driven in a private vehicle. They drive a total amount Q per year, which is higher than the optimum amount Q* they would drive if the externalities they produced were fully factored in. Therefore, the economic problem resulting from these negative externalities is that the output is higher than optimal.Then, decreasing output can create value. On the other hand, with positive externality, the output is lower than optimal, therefore increasing output can create value. There are two major camps in externality theory: the Pigouvian and the Coasean. Both Pigouvians and Coaseans use Pareto optimality, or Pareto improvement, as the basis for their different policy recommendation. The Pigouvian economists solve the externality problem by making the emitter liable for economic damages. From the Pigouvian perspective, making the emitter to impose an emission tax is

Introduction  5 $ Social marginal cost

Personal marginal cost

$1 Marginal benefit

O FIGURE 1.1  Negative





considered to create the necessary and sufficient conditions for restoring optimality. On the other hand, in the Coasean system, these problems are solved when agents bargain over internalizing externalities and reveal their willingness to pay. By focusing exclusively on negotiating agents, Coase (1960) claims that externalities are ‘reciprocal.’ Here, reciprocity means that causality is dual, both sides of the problem are liable and should be considered. The optimal solution is to minimize the damage. From the Coasean perspective, efficiency requires determining which party could change behavior most cheaply. In such a world, bargaining and merger, for example, become effective methods for internalizing externalities. If the existence of costless Coasean bargaining is assumed, sufferers from the externalities (such as those who object to the air pollution or excessive congestion) could bribe drivers so that they drive less. A bribe, or subsidy not to pollute, is just equivalent to a charge coupled with a lump-​sum transfer to the polluter. In this case, there would be no basis for imposing a Pigou tax. Even if bargaining is costly and transaction costs are positive, there could still be incentives to drive less or to use public transportation, depending on the level of transaction costs. Limitations on parking can be seen as aspects of a political set of Coasean bargains. People will change their behavior as a consequence of policy or incentive changes. Public economics is especially considered to be the analysis of incentives in the relationship between private economic agents and the government. For example, governments attempt to reduce problematic alcohol consumption through setting alcohol taxes. Excessive alcohol consumption has been associated with increased rates of crime (Carpenter, 2007). Under the economic approach, the purpose of the alcohol excise tax is to ensure that, at the margin, each drinker takes the external costs into account when making one’s drinking decision. Excise taxation

6 Introduction $



O FIGURE 1.2  Demand


curves for light and heavy drinkers

can lead drinkers to internalize negative externalities. Higher excise taxes would provide inappropriate signals about drinking.3 Governments have long used taxation to correct for the socially costly overdrinking. Following Pigou’s (1920) idea, marginal externalities are constant and equal across drinkers, a single tax rate on ethanol (or pure alcohol) can correct for negative externalities. However, the marginal externality associated with drinking can be heterogeneous across individuals. The majority of the external costs is created by a small number of heavy drinkers. The problem is that light drinkers do not generate external costs, but heavy drinkers do. The price response may not be constant: heavy drinkers’ demand for alcohol is less responsive to price than is the demand of light drinkers. Figure 1.2 displays two demand curves for light and heavy drinkers where the demand curve becomes less price responsive as we move from light to heavy drinking. The price increase induced by the tax may be too high for some consumers (those with a low marginal externality) and too low for others (those with a high marginal externality). There is a need to balance the reduction in harmful drinking through higher taxation against the loss in welfare of light drinking. When the marginal externality of drinking varies across individuals, a single tax rate can no longer achieve the first best. When there is heterogeneity in consumption patterns and the marginal externality of consumption across individuals, Diamond (1973) considers optimal corrective taxes. In the setting, the second-​best prescription is to set an alcohol tax rate equal to the average marginal externality across drinkers. Many governments vary tax rates across different types of alcohol. By levying a relatively high tax rate on strong spirits, the government can target a larger share of the alcohol purchases of heavy than light drinkers and encourage heavy drinkers to switch to less strong alcohol products, hence lowering their level of ethanol consumption.

Introduction  7

1.2  Behavioral economics and policy Under the rational choice paradigm (‘agents choose what is best for them’), public economics has examined the welfare effects of policy. In conventional economic modeling, it is assumed that people know what will make them best off and that they can achieve such a state, given the resources and wealth available to them. That is, individuals make choices that are most likely to improve their welfare, with predetermined tastes and preferences, using the information available and processing it appropriately. However, such a modeling of individual decision-​making has been questioned in the literature of behavioral economics. The use of a behavioral approach in social sciences can be traced back at least to Herbert Simon. Simon (1957) coined the term ‘bounded rationality’ to refer to choices, given the cognitive limitations of individual decision-​makers in terms of acquiring and processing information. For Simon, human beings are capable only of very approximate and bounded rationality. The concept of bounded rationality takes a procedural view on decision-​making. People intend to act rationally but are constrained by internal limitations. One way to view bounded rationality is to think of an agent as an information processor: the procedure is a reasonable compromise between the accuracy of the output and the difficulties involved in processing. Human beings do not know all the alternatives that are available for action, and they are unable to make the calculations that would support optimization. Therefore, people must use shortcuts, or heuristics, rather than engage in utility-​maximizing behavior. Conventional economic theory does not address the decision-​ making process because there is no interest in the process per se, only in the outcome. Bounded rationality is an alternative conception of rationality that incorporates the cognitive processes of decision-​makers more realistically. One of the key messages of behavioral economics is that people do not seem to behave rationally in the manner of conventional economic modeling. Behavioral economics combines the economics of incentives with insights from psychology about how people actually behave under real-​world circumstances. For example, psychological evidence shows that individuals exhibit a ‘status-​quo’ bias, which is just like it sounds: a preference for the current state of affairs. People prefer maintaining the default, rather than switching away from it. They tend to rely on the status quo or preset default.The program should capitalize on people’s tendency to select default options when making decisions. Many savings plans aim to enable people to save money for their retirement period. However, despite the obvious economic benefit of saving for retirement, many employees do not save nearly enough for their future (Benartzi and Thaler, 2004). Do people simply prefer to spend their money rather than save? Behavioral economics makes it clear that low savings rates are at least partly by the complexity in choosing between different savings options (Huberman et al., 2007) or the substantial impact of defaults in decision-​making (Samuelson and Zackhauser, 1988). It is not surprising that consumers (who have to make highly complex decisions) turn to rules of thumb and mental shortcuts to

8 Introduction

ease the burden. Public policies should make realistic assumptions about the mental capabilities of the people facing various decisions. How do economists’ prescriptions for public policies change in view of evidence that individuals are not fully rational? This book examines the implications of such behavior for public economics. It is important that public policies build on the apparently nonrational aspects of human decision-​making rather than assuming them away. People usually display bounded rationality, in the sense that they depart from the standard account in predictable and systematic ways. They should not be seen as irrational; rather, heuristics and biases as well as situational dependency present both challenges and opportunities for policymakers. As discussed in Section 1.1, public economics typically lists four (traditional) market failures that may justify government intervention in markets—​imperfect competition (or natural monopoly), externalities, public goods, and asymmetric information. The standard approach to public economics, based on the principle of revealed preference, constitutes a unified framework. By relying on the revealed preferences approach, neoclassical economic analysis does not distinguish between individual’s choices and well-​being. As a consequence, given his or her preferences, constraints, and available information, the economic agent would end up making the choice that guarantees him or her the maximum well-​being. The conventional economic approach would be useful if it is capable of jointly describing the effects of a policy on behavior and evaluating the effects on welfare. However, according to behavioral economics, observed choices do not necessarily reveal ‘true’ preferences. Most behavioral economists agree that it would be a mistake to accept the satisfaction of revealed preferences as the normative criterion of choice. What makes individuals better off is not the satisfaction of revealed preferences, but ‘true’ or ‘latent’ preferences which may not always be observed through choice. Experiments have consistently shown that individuals do not possess stable context-​independent revealed preferences. In this situation, it is very difficult to judge whether a particular policy makes an individual better off. It has been suggested that consumers undervalue energy efficiency in a variety of choice situations. For example, we frequently observe that consumers do not choose the most energy-​efficient appliance, even if this appliance is also the most cost-​efficient choice for them. The low adoption of energy-​efficient appliances despite apparently large cost savings is closely related to what Jaffe and Stavins (1994) call the ‘energy paradox’: the diffusion of apparently cost-​effective and energy-​efficient technologies is typically gradual. Is the low market share of energy-​ efficient appliances simply an expression of well-​informed preferences, or are consumers unaware of or inattentive to how much money they could save? Allcott and Taubinsky (2015) test the effects of information about energy cost on consumers’ choices of either compact fluorescent light bulbs (CFLs) or incandescent ones. Compared to standard incandescents, CFLs last much longer and use four times less electricity (a 60-​watt equivalent CFLs saves about $5 per year on average). Their results suggest that providing monetary information supports consumers in accounting for operating cost.The information is intended to help consumers make

Introduction  9

better decisions. Information provision is a key element of government energy efficiency policy. Information-​provision policies can be improved by making information easier to understand and use.This implies that consumers make sub-​optimal decisions because they lack information. However, consumers may not only lack information, but they may also misperceive or misuse the information they have. Recent research in behavioral economics highlights a fifth possible ‘market failure’—​individuals may make mistakes in pursuing their own well-​being. For example, limited self-​control may lead to a deterioration of health (e.g., drinking, smoking, overeating) or a decrease in wealth (e.g., compulsive credit card purchases, consumption of luxury goods, gambling). In these situations, individuals might benefit if they are induced to behave as if they had perfect self-​control. Systematic mistakes constitute a ‘behavioral’ market failure that, like other market failures, government can step in to alleviate. This new kind of market failure brings us to the realm of ‘behavioral’ public economics.The field has drawn from the growing studies (behavioral economics, judgment and decision-​making) formulated by scholars such as Daniel Kahneman, Amos Tversky, and Richard Thaler. Behavioral public economics offers the interdisciplinary analysis of public issues from the micro-​level perspective of individual behavior and attitudes by drawing on recent advances in our understanding of the underlying psychology and behavior of individuals. It focuses on the psychological processes underlying the perceptions, attitudes, and behaviors of individuals in the public sector. Contributing to our understanding of behavior requires not only developing well-​crafted theory but also challenging some of the assumptions of established theory. This book will reconsider many results in traditional public economics that are based on the standard modeling of choice and consider whether these results are robust to an alternative modeling of behavior that grounds on more realistic foundations. Furthermore, this book will argue whether, relying on the alternative modeling of behavior, policymakers can expand a toolkit of policy instruments. Since we consumers are not fully rational, it may be profitable for firms to exploit our decision errors, hence government interventions can improve welfare even absent other market failures. If people systematically violate rational choice, they can consistently make themselves worse off while often having little power to improve. Behavioral public economics is expected to provide us with public policy implications to prevent those mistakes in individual behavior.These public policies help to facilitate better individual choices, thereby raising individual welfare. Consumer information, education, and advice are the major policy tools to educate, enable, encourage, and empower consumers. Insights from research in behavioral economics could be very useful in informing policy decisions. Over the last several decades, the standard economic prediction of human behavior has been subject to challenges from empirical results qualified as anomalies (Thaler, 1988). Anomalies are often labeled as decision-​making failures or mistakes. In standard economic modeling, individuals’ preferences are assumed to be time-​consistent, independent of the framing of the decision, and affected only by own payoffs. In the field of behavioral economics, human judgments and choices

10 Introduction

often deviate substantially from the predictions of conventional economic models. Individuals are time-​inconsistent (Thaler, 1981), exhibit an attitude toward risk that depends on framing and reference points (Kahneman and Tversky, 1979), and show a concern for the welfare of others (Charness and Rabin, 2002). Economics is steadily incorporating the empirical realities of decision-​making into its models. Behavioral economics is the study of how individuals behave and how thinking and emotions affect individual decision-​making.4 The boundedness is sometimes defined as a deviation from the standard model. According to Mullainathan and Thaler (2000), behavioral economics deviates from the conventional economic modeling in three ways: (i) Under ‘bounded rationality’ conditions, humans are faced with limited cognitive abilities that constrain their problem-​solving abilities. (ii) ‘Bounded willpower’ illustrates that humans sometimes make choices that are not in their long-​term interests. (iii) ‘Bounded self-​interest’ shows that humans are often willing to sacrifice their own interests to help others. First, under bounded rationality conditions, humans are faced with limited cognitive abilities that constrain their problem-​solving abilities. Bounded rationality is caused by shortcoming due to limited attention, computational capacity, and biased reasoning. Across a wide range of contexts, actual judgments show systematic differences from unbiased forecasts.5 Second, bounded willpower illustrates that people sometimes make choices that are not in their long-​run interest.With respect to bounded willpower, the question concerns how to view a decision to spend rather than save or to consume desserts rather than salads. People often fail to follow through on the plans they make. Bounded willpower or self-​control focuses on time-​inconsistency, in particular, the discrepancy between planning and implementation due to behavioral tendencies (e.g., procrastination). Finally, bounded self-​interest is due to ‘other-​regarding’ preferences. It emphasizes that many people care about giving and receiving fair treatment in a range of settings. Bounded self-​ interest shows that humans are often willing to sacrifice their own interests to help others. Behavioral economics encourages economists to be more receptive to the results of experimental research. By adding insights from a wide range of disciplines (psychology, sociology, neuroscience, and evolutionary biology), behavioral economics tries to modify the conventional economic approach and to analyze how ‘flesh-​and-​blood’ people act in social contexts.The size of the deviation determines the extent to which the conception of economic choice needs to be modified. As one result of the behavioral economic approach, what is best for individuals from their actual behavior does not hold any longer. Behavioral economic studies show precisely that individuals do not always make optimal choices. Individuals systematically depart from neoclassical economic assumption of rationality. Psychological research teaches economists about ways to describe preferences more realistically, about biases in belief formation, and about ways it is misleading

Introduction  11

to conceptualize humans as attempting to maximize stable, coherent, and accurately perceived preferences. In Kahneman’s (2003, p. 1449) words, “[o]‌ur research attempted to obtain a map of bounded rationality, by exploring the systematic biases that separate the beliefs that people have and the choices they make from the optimal beliefs and choices assumed in rational-​agent models.” The idea of behavioral and cognitive modularity helps explicate a variety of otherwise anomalous observations. How closely we wish to interweave economics with psychology depends both on the range of questions we wish to answer and on our assessment of how far we may trust the assumptions of conventional economic theory as approximations. There are two different processes which the mind uses to interpret the world. Dual-​process models have their origins in the 1970s and 1980s, and they have become one of the most important theoretical developments in the understanding of human behavior (Evans, 1989). The human condition is marked by two, often conflicting, cognitive systems: System 1 and System 2. Our decision-​ making processes are the result of these two interacting mental systems. Kahneman’s book, Thinking, Fast and Slow (2011), is organized around the metaphor of System 1 and System 2. As the book title suggests, System 1 corresponds to thinking fast and System 2 to thinking slow. Our thinking is a combination of the two systems: System 1 processes are intuitive, automatic, unconscious, and effortless, while System 2 processes are deliberate, controlled, conscious, and effortful.6 System 1 is driven by emotions, intuitions, habits, and traditions; it derives humans to pursue their wants and satisfactions with an urgency requiring little or no reflection or contemplation. Thus, in the mode of System 1, we are ‘human’ actors who think about current situations and how to gratify immediate desires. The problem with System 1 is that individuals apparently often make decisions that seem sub-​optimal from the perspective of maximization of expected utility. In the mode of System 2, by contrast, we are ‘economic’ actors who think more long-​range about planning for outcomes. In utilizing System 2, we carefully calculate costs, benefits, and any other aspects salient to the situation before acting. Decisions made by System 2 would be consistent with conventional economic predictions. Additionally, System 1 processes are typically thought of as older, shaped by evolution. System 2 processes are typically thought to be evolutionary newer. Rather, many of our actual choices in life are System 1 choices and are, therefore, subject to substantial deviations from the predictions of the standard economic modeling. Behavioral public economics can be conceptualized as the set of all attempts to modify standard public economics so that it is better aligned with insights from behavioral economics and can be coherently applied in a world where individuals have limited computational capacities, attention, and willpower. Public policy is tightly linked to behavioral economics, in particular, by considering and predicting how people will change their behavior as a consequence of policy or incentive changes inherently involves psychology. Behavioral economics is now being used or seriously considered as a policymaking tool across a wide range of areas. Fundamentally, behavioral economics is concerned with the question of how

12 Introduction

people actually behave in decision-​making situations and how their choices can be improved so that their welfare is enhanced. If the deviations from perfect rationality are predictable, policymakers can craft policies to alleviate the adverse effects by inducing people to make better choices. Policy design applies insights from behavioral economics more broadly. Though it is a relatively recent endeavor, there are several behavioral public policy frameworks. Behavioral economic research might help create policies that would increase savings, decrease drunk driving, inhabit violence and aggressions, or increase the duration of marriages. Some of these approaches aim to educate people about their possible behavioral biases so that they may make more savvy decisions, and others instead aim to influence automatic decision-​making without appealing directly to deliberation. Pioneered by the Behavioural Insights Team (BIT) in the United Kingdom, international organizations and governments around the world are increasingly making use of behavioral insights in policymaking. In 2010, the BIT was created by the Cabinet Office to dedicatedly incorporate the behavioral economic approach into policymaking. One pioneering study from the BIT found that framing alterations in the wording of tax delinquency letters had a notable impact on payment of overdue taxes (Hallsworth et al., 2017). To date, the BIT has conducted hundreds of behaviorally informed field experiments with government partners, both in the United Kingdom and globally. Also, in 2014, the United States established the White House Social and Behavioral Science Team.With the establishment of governmental behavioral units, behavioral interventions are becoming part of the policy toolkit. In September 2015, the White House issued the Executive Order ‘Using Behavioral Science Insights to Better Serve the American People’ in an attempt to push further the implementation of behavioral policymaking.7 Furthermore, European Union (EU) member state governments and the European Commission are increasingly active in bringing behavioral insights into legislation and regulatory interventions as well as in establishing and developing organizations for the implementation. According to the European Commission Report ‘Behavioural Insights Applied to Policy’ published in 2016, there are more than 200 examples of behavioral policy initiatives in EU member states in fields such as competition, consumer protection, employment, energy, environment, health, finance, taxation, and transport.8 For any decision problem, there is a ‘framing.’ People often make decisions that are bad for then, and these decisions are strongly influenced by the framing. Many policy interventions can be used to alter people’s behavior in a predictable way. Because people exhibit systematic and predictable errors in judgment and decision-​making, public policy should help them make better choices in contexts where they are prone to error. Much discussion about behavioral policy tools have emphasized so-​called nudges: interventions that steer people in a particular direction while preserving freedom of choice (Thaler and Sunstein, 2008). Nudges should be assessed in terms of their impacts, like all policy instruments. The idea that choices can be improved by framing has made nudging politically attractive. It is not surprising that nudge-​type interventions—​quick, simple, cheap, and (perhaps most importantly) low-​r isk—​tend to be appealing for all parties. The insights

Introduction  13

of behavioral economics have become cornerstones of understanding consumer behavior, helping inform policymakers on how to nudge people to make better choices. Government and private organizations can provide much aid to consumers in the critical decisions that they face. Consumers often do not make decisions in their best interest and, therefore, they should be nudged toward better decisions by those who are more knowledgeable.The approach seeks to guide people’s behavior in particular directions without the use of force or mandates. Nudges can be viewed as manipulations of the circumstances of choice. An example of nudging is placing, at the entrance of a school cafeteria, salads and fruits before less healthy foods. The layout of food is then expected to influence people’s choice of consumption. Thus, placing low-​fat food first will promote healthy outcomes without restricting the freedom of choice of the others. Other interventions that have shown promise in some school cafeterias include putting salad bars in the middle of the food-​ purchasing area, giving students a choice of vegetables, closing the lid on the ice-​ cream freezer, and requiring that sweets and soft drinks be paid for by cash rather than lunch cards. Individuals who prefer less healthy food do not have to make extra effort or pay higher prices to consume it. Importantly, a nudge changes the framing but does not change the set of choices. Putting fruits at eye level would be a nudge, but banning junk food would not, as this would change the set of choices. Banning junk food is rather a traditional behavioral intervention where an agent uses a commitment device that restricts his or her choice set. Other examples are the so-​called opt-​in and opt-​out options, such as default rules for regulating organ donations. Most countries experience a chronic undersupply of organs for transplant. Willingness to donate organs, however, varies dramatically by country.These variations are due to how the organ donation choice is framed: ‘if you want to be an organ donor, please check here’ (opt-​in) and ‘if you don’t want to be an organ donor, please check here’ (opt-​out). Organ donation rates have been low in countries which require donors to opt-​in—​take action to be listed as a potential donor. One is then presumed to be willingness to donate only by explicitly giving permission. The default effect is then the difference in choice between the opt-​in condition versus that in the opt-​out condition. When an online experiment made the default donor status opt-​out rather than opt-​in, organ donation rates nearly doubled from 42% to 82% (Johnson and Goldstein, 2003). One is then presumed willing to be a donor unless one explicitly indicates an unwillingness to donate. Furthermore, a follow-​up survey indicated that it raised willingness to serve as an organ donor—​making it opt-​out did not simply force unaware people onto the donor rolls. Thus, the previous low consent rates are interpreted not as unwillingness to serve as donors but as the influence of the default assumption that people do not want to be donors. Defaults exert a considerable influence on decisions. In most decisions in life, people are required to opt-​in; the default is to do nothing. In an opt-​out arrangement, on the other hand, the default is to participate in an arrangement. If one does not wish to participate, one can decline, but one must formally choose to decline. Nudges can be administratively reversed. Low rates of participation in optional training might be addressed by switching from an opt-​in

14 Introduction

to an opt-​out system. If participation rates appear to be too high, the system can be reversed. Nudges requiring waiting periods before actions can be taken may be sufficient to induce appropriate caution. Default rules are supposed to influence people’s behavior effectively because of the status quo bias (i.e., the tendency to place higher values on options perceived as status quo).9 When faced with making decisions, people are prone to inertia and tend to make the same kinds of choices they have made before. People always decide in a choice context—​an arrangement of alternatives specified in particular ways. They are influenced greatly by the environment and can be manipulated. There is no setting or situation without ‘choice architecture’ (i.e., the environment in which choices are made). It would be impossible for a choice environment to be completely neutral about the way that alternatives are presented and arranged.The idea is to design choice architecture in ways that engage with behavioral mechanisms that economists have traditionally regarded as nonrational. These mechanisms are to be used to increase privately and socially beneficial behavior without adjusting prices or restricting choice. Behaviorally informed policy is based on the concept of ‘libertarian paternalism,’ which preserves freedom of choice but nonetheless steers people in a particular direction. The idea is that people often do not make decisions in their best interest, and therefore, they should be nudged toward better decisions by those who are more knowledgeable. People should be given choices (the libertarian idea), but the choice decision should be guided by experts (the paternalistic idea).The libertarian paternalism approach seems to be less intrusive than direct government regulation, and it would merely nudge individuals into better decisions without directly forcing or imposing additional costs on them. People are free to continue with their existing behavior if they wish. For a policy intervention to meet the requirements of libertarian paternalism, it has to be liberty preserving. Retaining freedom of choice becomes the best safeguard against any misguided policy interventions. Nudges—​ the applications of libertarian paternalism—​ are low-​ cost, choice-​ preserving approaches to regulatory issues, including disclosure requirements, default rules, simplification, and use of salience and social norms. They enable policymakers to influence people’s behavior without putting effort into cumbersome legislative and deliberative processes. The core idea in libertarian paternalism is that, with knowledge of the behavioral influences, the context or environment that people face can be redesigned such that their automatic decisions better align with their deliberative preferences.

1.3  Why social preferences matter? The standard conception of the economic agent, Homo economicus, driven by material self-​interest has provided powerful theoretical tools in the analysis of diverse problems. However, experimental support for this concept has been fragile. Results from experiments on public goods games, ultimatum games, trust games, and gift exchange games demonstrate that subjects in fact deviate from self-​interest

Introduction  15

in systemic ways (e.g., Isaac and Walker, 1988; Fehr et al., 1993). Aside from being concerned with their own payoffs (self-​interest), subjects appear to be concerned also with the payoffs of others (other-​interest). Preferences having this property are referred to as being ‘socially interdependent’ or other-​regarding. Self-​interest is not of the only human motivation. Many critics of neoclassical economics have focused on the construct of Homo economicus, arguing that humans often demonstrate altruism and mutual concern. The expanding behavioral economic literature supports the argument that Homo economicus is an inappropriate assumption about how humans take economic decisions. Especially, the ultimatum game by Güth et al. (1982) has played an important role in weakening the exclusive reliance on the self-​interest hypothesis. The ultimatum game involves two parties, the proposer and the responder, who have to agree on the division of a fixed sum of money. The proposer can make one proposal of how to divide the amount; on the other hand, the responder can accept or reject the proposed division. In case of acceptance, the proposal is implemented; in case of rejection, however, both receive nothing. In the subgame perfect equilibrium, the responder accepts any positive amount of money and therefore the proposer offers the responder the smallest possible amount of money. However, experimental subjects repeatedly violate this theoretical prediction.The results from ultimatum game experiments in developed countries show that proposers do not demand nearly this amount, generally demanding between 50% and 60% of the available surplus. People are self-​interested from a core hypothesis of neoclassical economics, but they are often socially minded from empirically adequate behavioral accounts. This problem also arises about Adam Smith’s conception of the human character. Are people self-​interested or even selfish as they seem to be depicted in The Wealth of Nations, or are they socially minded as being depicted in The Theory of Moral Sentiments? Is there a relationship between the two components, the self-​regarding behavior and the other-​regarding one? Adam Smith devoted his book, The Theory of Moral Sentiments, to explaining how individuals are transformed into moral beings. The moral sentiments are profoundly ‘endogenous.’ Even as an individual is shaped by society, the experience of each person creates a unique perspective from which that person can act on. The emergence or maturation of civic ethics is the subject of Smith’s Theory of Moral Sentiments. In the opening of his Theory of Moral Sentiments, [h]‌ow selfish soever man may be supposed, there are evidently some principles in his nature, which interest him in the fortune of others, and render their happiness necessary to him, though he derives nothing from it except the pleasure of seeing it. Smith, [1759] 1981, p. 9 Sympathy involves concern for another’s welfare. People care and wonder about the feelings that fill the hearts of others.We are not capable of seeing into the hearts

16 Introduction

of others. But by observing others, we enter in an empathic relation with them. Empathy denotes sharing and understanding another’s feelings. We can imagine what it would be like to be in the shoes of the parties who are affected and consider what to feel—​what any one of us would or should feel—​in their situation. Others’ unhappiness somehow intrudes into us: if others are unhappy, we are also unhappy. Thus, following Smith ([1759] 1981), people are endowed with an altruistic mechanism that makes them share the fortunes of others. Smith’s ([1759] 1981) enduring contribution to moral philosophy is his theory of conscience in the form of the ‘impartial spectator.’ The concept of the impartial spectator, also called the ‘man within the breast,’ means an imagined third party who allows an individual to objectively judge the ethical status of one’s actions.The impartial spectator is a judge who adjudicates based on the context of one’s social milieu. Its essence lies in the social experience of being a spectator, and it is by the imagination only that we can form any conception of what others’ sensations are. This capacity to reflect on ourselves derives from our capacity to see ourselves as others see us. Individuals judge others’ actions by using the concept of empathy—​ sharing of feelings.We develop an otherwise impossible sense of propriety about our moral judgments. Social approval is considered as one of the important motives of individual behavior. Due to the dependence on social values, the impartial spectator helps individuals improve their behavior to make them more socially approvable. The relationship between imagination, sympathy, and moral judgment especially concerns Smith’s moral philosophy. The use of imagination allows an individual to make judgments about an agent and about one’s own actions. Sympathy is a specific form of ‘fellow-​feeling,’ a sharing of another’s feeling or motive as a result of projecting into one’s perspective and seeing one’s situation in the same emotionally and motivationally way we imagine one does. Sympathy is described as a significant feature of human behavior: the imagined change of positions as a device facilitating interpersonal comparisons. In sympathizing, one individual no longer fully distinguishes one’s interests from another with whom one identifies (one ‘feels with’ another). Smith’s account of sympathy involves the reasoned evaluation of the emotional response: The compassion of the spectator must arise altogether from the consideration of what he himself would feel if he was reduced to the same unhappy situation, and, what perhaps is impossible, was able to regard it with present reason and judgment. ibid., p. 12 In his analysis of moral judgment, sympathy is characterized as an outcome of imagination simply because we do not access to other minds: We have no immediate experience of what other men feel, we can form no idea of the manner in which they are affected. It is by the imagination only that we can form any conception of what are his sensations. ibid., p. 9

Introduction  17

Impartiality regulates moral judgment, but it does so by disciplining the way in which we enter into the agent’s point of view, not by providing its own perspective. Moral judgment involves an impartial projection into the agent’s standpoint. We imaginatively project, not as ourselves, but impartially, as any one of us. We can view our own conduct from the externalized situations. The human concern for other individuals is derived from feelings of sympathy. Smith’s explanation of social justice relies on imagination and sympathy to create a general resentment against causing emotional or physical harm to other individuals. The agents in The Theory of Moral Sentiments are driven by an internal struggle between their passions and the impartial spectator (Ashraf et al., 2005). The essence of Smith’s ethical system is the balance that must be considered in judgment. The judgment is divided into two parts. One is the degree to which the impartial spectator is in sympathy with the sentiments’ response to the cause. The impartial spectator is a principle of moral self-​reflection that develops over the course of our experience of exchanging places with others through the practice of imaginative sympathy. The other part of judgment is an act’s merit or demerit. This refers to the passion leading to a behavior. In Smith’s analysis, proper balance is the key to moral sentiments; the standard of proper balance is the sympathy of an impartial spectator. To know this standard would be a first step toward virtue. Individual ‘self-​command’ (the propriety of behavior) is another key concept in Smith’s system of moral philosophy. Self-​command is the power to enforce upon oneself the judgment of the impartial spectator. According to Smith: The man who acts according to the rules of perfect prudence, of strict justice, and of proper benevolence [i.e., with a balance of sentiments as dictated by the impartial spectator], may be said to be perfectly virtuous. But the most perfect knowledge of those rules will not alone enable him to act in this manner: his own passions are very apt to mislead him; sometimes to drive him and sometimes to seduce him to violate all the rules which he himself, in all his sober and cool hours, approves of. The most perfect knowledge, if it is to be supported by the most perfect self-​command, will not always enable him to do his duty. [1759] 1981, p. 237 If an individual had the self-​command necessary to maintain proper balance, the individual could achieve Smith’s standard of perfect virtue. For Smith ([1759] 1981), ideal liberal society would be inhabited by such perfectly virtuous individuals. Altruism has been the research topic in many academic disciplines, including biology, psychology, philosophy, and economics. The term ‘altruism,’ however, has been used differently in order to fit the particular research contexts. According to evolutionary biology (Trivers, 1971), prosocial behavior might be sustained by means of ‘reciprocity.’ In Trivers (1971), the problem of sustaining prosocial behavior is conceptualized as a two-​ player prisoner’s dilemma game. The key aspects of the prisoner’s dilemma are (i) cooperation maximizes the total payoff

18 Introduction

to everyone involved in the interaction (i.e., mutual cooperation provides more benefits than mutual defection); however, (ii) any player will receive a higher personal payoff by defecting, so a sizable temptation to cheat exists. In one-​shot, anonymous encounters, cooperation cannot be sustained because both players have reason to defect regardless of what the other chooses. In this situation, the rational payoff maximization assumption predicts no cooperation. There are gains for both players that can be obtained from mutual choices of cooperative strategies, but a noncooperative choice is individually rational for each if they interact only once. However, if interactions are repeated, cooperation may be rational as predicted by the folk theorem: the players care about their future sufficiently (or do not discount too heavily) and have accurate information about the other’s choices. Then, the players can make their cooperation contingent on cooperation from the other party, and they make defection contingent on defection from the other. Reciprocators receive a large gain when interacting with other reciprocators. The idea of conditionality in theories of reciprocity is crucial. If individuals observe that others behave pro-​socially, they will do so as well. Furthermore, in Axelrod (1984), the strategy ‘tit-​for-​tat’ embodies the fundamental idea behind reciprocity. The partner who plays the strategy tit-​for-​tat always cooperates with his or her partner in the first round of the repeated prisoner’s dilemma game.Thereafter, he or she cooperates if the partner had cooperated in the previous round and defects if the partner had defected in the previous round. In Axelrod’s (1984) analysis, the strategy tit-​for-​tat can successfully invade simulated populations of partners engaging in prisoner’s dilemma games, winning out over alternative strategies. Throughout their lifetime, people depend on frequent and varied cooperation with others. In the short run, it would be advantageous to cheat. Cheating, however, may compromise one’s reputation. This would make people better able to resist the temptation to cheat their partners in the first round and would enable them to generate a reputation for being cooperative. As expressed in Trivers (1971) and Axelrod (1984), altruistic assistance would be offered if the player expects one’s future benefit. Beneficence is modeled as a non-​myopic, self-​interested strategy to ensure future cooperation. The motivation of the player has to be explicitly considered in defining whether one’s act is altruistic. Is the self-​interested motive of reputation maintenance sufficient to explain cooperative behavior? The debate revolves around the question of whether ordinary people behave in the way predicted by a crude view of human beings: humans have only self-​regarding preferences (e.g., Gintis et al., 2005).This view presents human beings as selfish utility maximizers; the motivation for behaving morally is self-​interest. Opponents of this view claim that human beings have preferences for the well-​being of others. Fehr and Fischbacher (2005, p. 16) describe as follows: Strongly reciprocal individuals reward and punish in anonymous one-​shot interactions. Yet, they reward and punish more in repeated interactions or when their reputation is at stake. This suggests that they are motivated by a combination of altruistic and selfish concerns. Their altruistic motives

Introduction  19

induce them to cooperate and punish in one-​shot interactions and their selfish motives induce them to increase rewarding and punishing in repeated interactions or when reputation building is possible. The problem of identifying ‘purely’ altruistic motivations in behavior is complicated by the fact that selfish agents might be induced to mimic altruistic behavior. Selfish agents may behave cooperatively in order to build up a useful reputation for altruism. In societies where it is common to enter into relationships with people with whom one is indirectly (or not even indirectly) acquainted, reputation is often one’s source of information about them at the outset. Social approval is thus considered as one of the important motives of individual behavior. People want to achieve the reputation of being fair; they are fair because they care about their reputation. People may not be genuinely fair; they have to be rewarded for a good reputation.They often follow norms of reciprocity and fairness, even when obedience is not in their immediate self-​interest and there is no obvious sanction. Indeed, when individuals reciprocate each other, there exists for them an incentive to acquire a reputation by keeping to their promises and carrying out actions whose long-​run benefits will outweigh the short-​run costs.This mechanism allows individuals to establish mutually beneficial relationships. Reputation is then spread in their group.10 Norms can be sustained if the pecuniary advantage from breaking them is not sufficient to offset the forgone reputation effect. This is related to ‘indirect reciprocity.’ Indirect reciprocity is not based on repeated interactions between the same individuals (Boyd and Richerson, 1989; Nowak and Sigmund, 1998). Individuals have never met before, and the likelihood of meeting again is negligible. In Alexander (1987), indirect reciprocity is arranged in the form of a chain: Person A helps Person B, Person B helps Person C, and so on. Person A is eventually helped by someone else who may not have been directly helped by him or her. Such chains of indirect reciprocity, according to Alexander (1987), serve to show how reciprocity can explain more generalized kinds of moral behavior. Altruistic actions can be sustained if people who support others receive support in turn. To achieve such indirect reciprocity, building up a positive reputation is needed. For example, if individuals only cooperate with those of a reputation signaling that they have cooperated before, free riding is inhibited. Thus, reputation is a key element in indirect reciprocity. The altruist in Becker’s (1976) model helps the other because the other’s utility function is embedded in the altruist’s utility function. Therefore, the altruist derives pleasure not because the other is assisted, but rather because the other’s pleasure is already part of the altruist’s utility. The donor would donate a resource if the vicarious enjoyment of watching the pleasure of others exceeds, at the margin, the donor’s satisfaction from consuming one’s own resource. Then, the reason behind the donation is considered to be the alleviation of the donor’s discomfort at the sight of a beggar.The improvement of the beggar’s income is the donor’s attempt to enhance one’s vicarious pleasure.

20 Introduction

Andreoni (1990) extends the altruism model with a ‘warm-​glow’ motive for charitable giving. Individuals may get a warm glow from volunteering because helping others increases either their perceived self-​esteem or their self-​efficacy. The concept of warm glow has advanced our understanding of important dimensions of charitable giving. The model of warm-​glow giving predicts that people will partly reduce their own contributions when others or the government increase their share to the public good. Smith ([1759] 1981) attempts to explain why and how an individual will feel sympathy for other less well-​off individuals. Sympathy merely permits an individual to ‘feel’ the plight of another. For Smith ([1759] 1981), sympathy expresses the genuine concern over the interests of others. Sympathy, as the spring of altruism, allows individuals to act not according to strategic calculation but rather according to the concern over the other-​interest. The donor, in Becker’s (1976) model, increases one’s own utility by imagining how the recipient is enjoying the donated resources, which only accounts for a single-​situation transaction. That is, the act is motivated ultimately by the self-​interest. On the other hand, for Smith, the concern over the other-​interest arises from particular ‘situation-​switching.’ That is, the sympathetic act involves putting one’s self in the other’s situation rather than judging the other’s pressure from one’s own situation. The motive to satisfy self-​interest and other-​interest, for Smith, stems from the same general tendency of humans to sympathize. Smith’s principle of sympathy thus entails continuity between the pursuits of self-​interest and other-​interest.

1.4  From economic incentives to social incentives Market failures provide the rationale for nonmarket remedies. However, the remedies may themselves fail, for reasons different from the accounting for market failures. A key question in economics is how to provide ‘incentives’ to individuals to adopt a desired behavior. Changes in incentives change motivations for undertaking actions. One solution is to implement economic incentives. An economic incentive is commonly characterized as an extrinsically motivated reward. An extrinsic reward designed to motivate a specific action is offered before an action occurs. For public policy, it is of interest to know whether economic incentives and social preferences are substitutes of any kind, in the sense that economic incentives crowd out social preferences. As demonstrated by Bowles (2016), the introduction of economic incentives can undermine intrinsic, morally shaped motivations. Economic incentives interact with intrinsic versus extrinsic motivations. In the case of environmental protection, for example, intrinsic motivations would correspond to a situation in which individuals seek to protect natural resources for environmental rather than for personal economic reasons. Extrinsic motivation, in contrast, depends on whether external rewards or sanctions are used to induce performance. Higher economic incentives activate extrinsic motivations. However, economic incentives influencing individual behavior may lead to outcomes that diverge substantially from what is socially preferable. Extrinsic motivations created by economic incentives provided

Introduction  21

to participants in environmental protection projects may crowd out intrinsic motivations to conserve natural resources. Experimental and field evidence has confirmed that economic incentives can crowd out intrinsic motivations. Such was the case in a daycare center in Haifa, Israel, where instituting fines for parents picking up children late actually increased the number of parents arriving after hours (Gneezy and Rustichini, 2000). The formalization of the rules regarding late pickup entailed a change in the nature of sanction, in this case, from social to material sanctions. Bowles (2016) argues that this type of behavioral failure occurs partly because such systems are based on faulty assumptions that people are only selfish and amoral. Ensuring a sufficient blood supply remains difficult. How are more people persuaded to become blood donors? What is the role of incentives in blood donation? One of the controversial problems in the field has been whether payment or other incentives could increase the blood supply. A standard economic framework focuses on blood donation with paid donors. Based on standard economic theory, the paid donation system could supply all needs if blood donation is considered as a private rather than a public good and is supplied by the market system. For an exchange to take place, an individual must perceive benefits equal to or greater than perceived costs. Therefore, economic incentives as external rewards should raise the perceived value of the exchange, to increase the likelihood of a personal blood donating. According to Titmuss (1970), however, monetary compensation for donating blood might crowd out the supply of blood donors. Blood donation is intrinsically (as opposed to extrinsically) motivated. Economic incentives might crowd out intrinsic motivation (e.g., by destroying the sense of ‘gift’ or by creating doubts about the true motives for which the donations are performed). Economic incentives do not attract sufficient donors to maintain a regular blood supply. Hence, paying blood donors would be economically inefficient. For an intrinsically motivated person, monetary rewards create doubt about his or her true motives for doing a good deed. Maintaining an adequate blood supply depends on unpaid donors giving blood. Why do people donate blood without payment? The act of blood donation can be described as a prosocial behavior. Prosocial behaviors are acts that most people in society or in a social group generally consider to be beneficial to others. Donors often cite prosocial reasons such as altruism, empathy, or social responsibility for their willingness to donate. If blood donation is motivated by pure altruism (i.e., helping another, at a personal cost, without personal gain), are people justified in using interventions that highlight benefits to the donor (i.e., highlighting personal emotional rewards)? Pure altruism has been seen as the key driver for giving blood. An altruistic motivated blood donor would respond by giving more blood in situations where others give less. However, pure altruism is apparently not enough to guarantee a steady supply of blood. Blood donation organizations are increasingly interested in the potential of incentives to enhance the effectiveness of recruitment and retention campaigns. In conventional economics, the private provision of a public good fails due to its characteristic of being non-​rival and non-​excludable. Selfish individuals have

22 Introduction

no incentive to contribute. They may free ride on its provision. In behavioral economics, on the other hand, concerns about social preferences can lead to voluntary contributions to a public good. People behave as if they were intrinsically motivated rather than stimulated by any monetary reward. Social incentives can be understood to be of nonmonetary value that donors receive conditional on donation. Such social incentives may operate through feelings of status, esteem, or pride. Individuals who offer help may do so in order to gain some personal gratification. Andreoni (1990) suggests that people derive utility from the act of giving, labeled the warm-​glow effect.This type of human behavior is referred to as impure altruism. Reputation can be also considered as a main mechanism for giving blood. In general, reputation is how others perceive a person’s type, nature, or value. Reputational motivation refers to when a person’s behavior is guided partly by how others perceive him or her. People would sometimes try to fulfill the expectations of others that they should join a group going to a blood donor event.The reputation-​seeking behavior would increase contributions to a public good relatively more in a public setting than in a private setting. People may undertake prosocial activities in order to improve their self-​image. Extrinsic, monetary incentives can crowd out prosocial behavior via a feedback loop to reputational concerns (Bénabou and Tirole, 2006). The reputational concern reflects the possibility that receiving monetary rewards for doing a good deed can lead others to interpret prosocial action as greediness rather than social responsibility, which makes prosocial behavior a less efficient signal of social type. People perform good deeds or abstain from selfish ones given social norms that attach honor to generous actions and shame to selfish ones. Various types of intrinsic motivations exist. Intrinsic motivations also may involve moral obligations. This motivation stems from the duty that one must follow a particular moral rule. Laffont (1975) used the words ‘Kantian economics’ in the title of the paper. Laffont (1975) asks why (at least in some countries) people do not leave their beer cans on beaches even though the impact on their own welfare from doing so is negligible. According to Laffont (1975, p. 431), “[e]‌very economic action takes place in the framework of a moral or ethics.” In Laffont (1975), all individuals are ‘Kantians.’ An individual then assumes that, according to a principle, the others will act as he or she does.There exist circumstances under which actions ought to be guided by moral rules, regardless of the utility consequences. Morality for Immanuel Kant, the 18th-​century German philosopher, is reason in action. For Kantian ethics, social preferences are not truly moral. Kant is well known for emphasizing universality as a criterion of conformity to reason. The categorical imperative, a notion put forward in the Groundwork of the Metaphysics of Morals (Kant, 1785), is not itself a moral rule—​it is an abstract formal principle.11 The categorical imperative has the form: “act only according to that maxim whereby you can, at the same time, will that it should become a universal law.” This formula represents the familiar ethical standard of universalization, which holds that any rule of action one adopts must be consistent with everyone else’s adopting it as well. Moral rules are seen as rules of duty in Kantian ethics. The first step of moral deliberation is the formulation of the rule that governs the specific action. Once

Introduction  23

the rule governing the action has been found, the agent then compares this formula with the requirements stipulated by the categorical imperative that the rule displays the characteristics of universality and necessity. Whether an action is morally good depends on how it relates to this principle. A commitment to the principle is deeply embedded in Kantian moral philosophy.

Notes 1 Other goods are partially non-​rivalrous, in the sense that the quality of the benefit provided to each person diminishes as the number of people to whom it is provided rises. These goods are said to be congestible. 2 Buchanan (1966) distinguishes between availability of a public good and accessibility to the flows of services deriving from that good. That is, the fact that all can consume a public good does not imply that all enjoy an equal flow of services measured along some physical dimension. For a given produced quantity, the deriving stream of services can be distributed in different ways among consumers. For example, the supply of a fire station to all citizens does not correspond to an equal stream of services from the public good, namely protection against fire. Then, the degree of protection can be different for each citizen. 3 Laws that restrict underage drinking, prohibit drink driving, or provide help to alcoholics can be considered as well-​motivated policies other than taxation. For example, Luca et al. (2015) provide evidence that alcohol prohibition laws in India are effective in reducing consumption and that prohibition significantly reduces violence against women. 4 According to Tomer (2007), behavioral economics is a multifaceted area of economic research that encompasses a range of methods and approaches. Methodological pluralism should be accepted by behavioral economists. 5 One such judgment error is optimism bias: people believe that their own probability of facing a bad outcome is lower than it actually is. For example, most people think that their chances of having an accident are significantly lower than the average person’s chances of experiencing this event. However, these beliefs cannot be correct; if everyone is below ‘average,’ the average would be lower. 6 In an example of System 1 and System 2, one group of respondents is (individually) asked to estimate the total number of murders in Detroit in a year, and another group is asked to estimate the total number of murders in Michigan in a year. Typically, the first group on average estimated a higher number of murders than the second. Then, System 1 thinking is in evidence. Detroit evokes a violent city, associated with many murders; Michigan evokes apple-​g rowing farmland. Without System 2 engagement, the fact that ‘Detroit is in Michigan’ does not come to mind for the second group of respondents. 7 White House Executive Order, available at​the-​press-​office/​ 2015/​09/​15/​ executive-​order-​using-​behavioral-​science-​insights-​better-​serve-​american. 8 See Lourenço et al. (2016). 9 Retirement saving in 401(k) saving plans was encouraged tremendously by setting the default to automatic enrollment. In most standard plans, employees must act (opt-​in) to enroll, but many more employees choose to participate when enrollment is automatic (the default). 10 Akerlof (1980) develops an economic model to show that disobedience to norms may involve a loss of reputation. In his model, a reputation function depends on the dummy variable indicating whether the individual obeys or disobeys the code of behavior. It is

24 Introduction

specified in such a way that the larger the number of code believers, the more reputation id lost by disobedience to the code. Akerlof (1980) defines a norm as a moral expectation shared by a number of people in a group. A group of individuals must embrace and follow the code of behavior to give it the status of a norm. Certain groups of individuals can maintain a strong reputation over time. 11 Rawls (1971) also appealed to Kant’s categorical imperatives in his formulation of the theory of justice. To act from the principles of justice is act from the categorical imperatives in the sense that they apply to us whatever our particular aims are. Rawls and Kant share the view that standards of justice are necessary to inhibit destructive social conflict.


2.1  Preferences and choices 2.1.1  Revealed preference theory Economics has been developed predominantly as a science of choice. According to Robbins (1932, p. 16), “[e]‌conomics is the science which studies human behavior as a relationship between ends and scare means which have alternative uses.” Robbins called this definition ‘analytical,’ because it “focuses attention on a particular aspect of behavior, the form imposed by the influence of scarcity” (ibid., p. 17). Robbins defined economics as a particular type of decision (choice under scarcity). For Robbins, ‘scarcity’ is the scarcity of given means for the attainment of given ends, and the ‘ends’ in choice theory is the satisfaction of the agent’s preferences. Robbins, exposed to the philosophy of positivism, excluded mental and ethical concepts from science. Positivism basically claims that science can be demarcated from nonscience according to whether or not the propositions in question predict observable events. Since observability is the demarcation criterion for science, Robbins claimed that interpersonal comparison of utility could be outside science. One of the goals of the ordinal revolution during the early 1930s was to move away from the hedonistic and cardinal notions of utility and move into the direction of the type of observable empirical evidence consistent with the positivist philosophical ideas of the day. Economists generally think about consumer behavior in terms of a particular version of rational choice theory: utility maximization subject to a budget constraint.1 They have understood and explained consumer choice by considering the utility function merely as a convenient mathematical representation of the preferences. In modern economics, the theoretical framework is ordinal utility theory. Parametric specification of the functional form permits empirical analyses and tests of its implications. Then, empirical findings depend both on actual DOI: 10.4324/​9780429344817-2

26  Preferences, utility, and welfare

observations of economic behavior and on the parametric choice of the utility function. Another approach to consumer choice is revealed preference theory. The conventional economic approach relies on the theory of revealed preferences, in which the concept of preferences is entirely defined in terms of choice and has no mental or psychological meaning. The theory of revealed preferences was first formulated by Samuelson (1938) in order to derive testable implications of rational consumption behavior without postulating a utility function.2 Samuelson’s (1938) revealed preference theory starts directly from the observed choices of the consumer: the preferences of individuals can be deduced from their observed choices. If a consumer is observed to have chosen a certain consumption bundle x, while another bundle y was also available, then the consumer reveals one’s preference for x over y. In this manner, choices say something about the underlying preferences of the consumer. Standard consumer theory thus allows to extrapolate public policy outcomes from the observation of private choices. Individuals have well-​defined preference rankings, and these rankings form the basis for welfare analysis. Samuelson (1938) introduced the weak axiom of revealed preference (WARP), which derives a testable implication of rational consumption behavior based on two different budget sets.WARP provides a test of the simplest form of the utility maximization hypothesis: if a bundle x is revealed preferred over a bundle y, then, at some other instance, y should not be revealed preferred over x. WARP requires the revealed preference relation to be asymmetric. Houthakker (1950), as an extension of Samuelson’s (1938) work, introduced the strong axiom of revealed preference (SARP), which derives a testable implication of rationality for any set of budget sets by exploiting transitivity. SARP states that the revealed preference relation is acyclic. In revealed preference theory, consistency of choice is required for the analysis of consumer choice. This approach consists of ‘nonparametric’ tests of the consistency of a set of decisions with some underlying utility function (Afriat, 1967; Varian, 1982). Samuelson (1938) and Houthakker (1950) assumed that one can observe the entire demand function of a consumer. They started with continuous real-​ valued demand functions in general form and not observed price-​quantity combinations. According to Afriat (1967), the revealed preference principle of Samuelson (1938), elaborated by Houthakker (1950), is absent by which the hypothesis can be accepted or rejected on the basis of observed choice of the consumer, supposed to be finite in number. Considering the more realistic setting where one only observes a finite set of choices, Afriat (1967) provided a slight relaxation of SARP (cyclical consistency): the necessary and sufficient conditions for the existence of a utility function whose imposed choice behavior is consistent with the data. Afriat’s (1967) theorem thus started with finite choice data and provided a way to actually construct a well-​behaved piecewise linear utility function that is consistent with the consumer choices. Afriat’s (1967) approach is more suitable as a basis for empirical analysis. Using the term generalized axiom of revealed preference (GARP) instead of cyclical consistency, Varian (1982) made Afriat’s (1967) approach more transparent. GARP is equivalent to assuming that the consumer behaves as if he or

Preferences, utility, and welfare  27

she maximizes some locally non-​satiated utility function. People’s behavior can be expected to conform to consistency axioms such as GARP only in specific environments and within a temporally limited scope. Samuelson’s (1938) goal was to provide more adequate foundations for the existing restrictions on individual demand functions by eliminating any reference to utility and the associated unobservable mental states. His goal was to move in the direction of the type of observable empirical evidence—​objective, not introspective. Following Samuelson’s (1938) axiomatic reasoning, the preference concept is totally defined on the basis of choice behavior. Consider a set of given and continuous demand function di defined for n goods, i = 1, …, n. The demand function specifies the quantity xi of good i consumed, given the set of prices p1, …, pn and the available money income M. Thus, we have: xi = di(p1, …, pn, M) for all i = 1, …, n. (2.1) All the functions di are assumed to be homogeneous of degree zero.That is, they are independent of the units in which prices are expressed: di ( λ p, λ M ) = di ( p, M ) for all λ > 0. Furthermore, the consumer’s budget constraint is satisfied: pi xi = M .

∑ i

Samuelson’s (1938) approach, which is nowadays known as the WARP, specifies consistency restrictions on the prices and the quantities of the goods which the consumer purchased at those prices. That is, for two bundles x0 and x1 and two sets of prices p0 and p1, if the consumer chose bundle x0 at prices p0 even though bundle x1 was affordable, then consistent behavior would imply that if bundle x1 were chosen at prices p1, it was because x0 was not affordable at p1. Formally: n

∑ i =1

pi0 xi1 ≤


∑ i =1

pi0 xi0 →


∑ i =1

pi1xi0 >


∑p x

1 1 i i

i =1

WARP thus provides a test for the consistency of the consumer’s choices based on potentially observable prices and quantities, and it is not necessary to introduce utility or preferences into the analysis. WARP can be expressed more generally in the choice-​ based approach, considering the actual choices made by the individual. The choice-​based approach is based on a ‘choice function’ C(.) specifying in a particular domain of non-​empty sets of alternatives X. For each subset S of X, a ‘choice set’ C(S) represents the chosen elements from S. Following our example of consumer theory, an alternative x ∈ X may correspond to a bundle of goods. A set S (⊂ X) can be understood as a particular set of all the affordable bundles for a consumer, given his or her wealth and market prices. Then, C(S) would be the bundle (or bundles) that consumer chooses to buy, among all bundles he or she can afford in S. Let us now consider two examples of choice structures, C1(.) and C2(.). Define the set of alternatives X = {x, y, z}, which contains only three elements. Consider two different sets (both

28  Preferences, utility, and welfare

of which are subsets of the set of alternatives X), S1 = {x, y} and S2 = {x, y, z}. In choice structure one, C1(.), the individual chooses only element x, regardless of which set is presented to him or her. That is, C1({x, y}) = {x} and C1({x, y, z}) = {x}. In choice structure two, C2(.), the consumer still chooses only alternative x when he or she is confronted with S1, that is, C2({x, y}) = {x}. When the set is enlarged to contain alternative z as well as it does in S2, his or her choice reverts to only alternative y, that is, C2({x, y, z}) = {y}. From these examples, we have to define consistency in choices. In the choice-​based approach, WARP is formally defined as follows: WARP : x, y ∈ S and x ∈C (S ) → x ∈C (S ' ), for any S ' with x, y ∈ S ' and y ∈C (S ' ). Consider a set S with x, y ∈ S, where element x is chosen, x ∈C (S ). Then, the choice structure satisfies WARP if, for any other set S ’, where alternatives x and y are still available, x, y ∈ S ' , and where alternative y is chosen, y ∈C (S ' ), alternative x is also chosen, x ∈C (S ' ). There are alternative ways of generating binary relations of preference from any choice function. First, the weak preference relation , for all x, y in X, is defined as follows: x  y if and only if for some S, x ∈C (S ) and y ∈ S. The preference relation  is thus defined by the choice function. The bundle x is ‘weakly preferred to’ or ‘at least as good as’ y if x is chosen when y is available. Strict preference () and indifference (~) are defined correspondingly. For all x, y in X: x  y if and only if x  y and not y  x. x ~ y if and only if both x  y and y  x. The next step in the development of revealed preference theory was Houthakker’s (1950) SARP. Houthakker’s (1950) (‘indirect’ revealed preference) approach extended the WARP condition from pairwise choices to chains of choices. In the choice-​based approach, SARP is formally defined as follows: SARP: for all alternatives x1, x2, …, xn and subsets S1, S2, …, Sn, if x1 , x 2 ,…, x n ∈ S1 x 2 ,…, x n ∈ S 2 … x n −1 , x n ∈ S n −1

Preferences, utility, and welfare  29

xn ∈ Sn and x1 ∈C (S1 ), x 2 ∈C (S 2 ),…, x n ∈C (S n ), then for all T with x1 , x 2 ,…, x n ∈T and all x i ∈C (T ), we must have x1 , x 2 ,…, x i −1 ∈C (T ) SARP is obviously stronger than WARP. SARP guarantees that preference relation  is complete, asymmetric, and transitive. As a consequence, it also guarantees that the existence of an underlying utility function representing the binary relation defined by . In the choice-​based approach, the construction of choice functions depends on the available data about finite sets of choices. For empirical studies, it is more suitable to construct a utility function consistent with the choices observed on the basis of a finite set of data regarding consumers’ choices. The centerpiece of revealed preference theory is the belief that people make consistent choices, with utility and utility maximization simply being convenient analytical tools to describe these choices. Revealed preference theory argues that economics should be about choices, should be concerned only with data reporting choices, and should show no interest in questions that go beyond observable choices. Standard welfare analysis almost always assumes that revealed preferences are preferences that represent the individual’s true interests.

2.1.2  Critique of revealed preference theory Amartya Sen is well known as a critic of rational choice theory in economics. For Sen, [m]‌uch of economic theory seems to be concerned with strong, silent men who never speak! One has to sneak in behind them to see what they are doing in the market, etc., and deduce from it what they prefer, what makes them better off, what they think is right, and so on. 1982, p. 9 Sen (1973) points out the paradoxical attitude of economists toward psychology. That is, economists do not ask people about the motivations for their actions but prefer to look solely at behavior. They are unable to say anything about how preferences are formed. Sen claims that choices made by rational agents may fail to reflect their personal welfare as represented by their preferences. In economics, the standard interpretation of preference is that of revealed preference: a person is defined to prefer some bundle x to another bundle y if he or she chooses x

30  Preferences, utility, and welfare

from a set of options that includes y. Revealed preference theory is grounded on the behaviorist postulate which equates observable choices to preferences. That is, when an action is observed, that action is interpreted as the result of some motive. For the interpretation of underlying preferences from observations of behavior, Sen (1973, p. 241, original emphasis) writes: If a collection of goods y could have been bought by a certain individual within his budget when he in fact was observed to buy another collection x, it is to be presumed that he has revealed a preference for x over y. The outside observer notices that this person chose x when y was available and infers that he preferred x to y. From the point of view of introspection of the person in question, the process runs from his preference to his choice, but from the point of view of the scientific observer the arrow runs in the opposite direction: choices are observed first and preferences are then presumed from these observations. In his criticism, Sen argues that several patterns of preferences can lead to identical patterns of observable choices, which implies that observing behavior does not provide enough information to infer preferences. The distinction between preference and choice is at the core of Sen’s critique of revealed preference theory. He argues against the idea that preferences can be entirely defined without a “peep into the head of the consumer” (Sen, 1973, p. 243). The concept of revealed preferences involves no reference to ‘mental states.’ Sen generally interprets preferences as mental states, to be interpreted either in terms of pleasure or in terms of desire. Preferences are the mental states that precede choices: to prefer x to y is to be in a state of mind in which one is disposed to choose x rather than y. Preferences are not what causes the consumer to choose particular goods. Consider someone choosing x over y. If both x and y were available to the consumer and the price is the same, people might conclude that the consumer preferred x to y. However, is merely choosing x over y enough to conclude that the consumer prefers x to y? Preferences cannot be revealed by choice alone. Conventional economics makes no assumptions about how choices are actually made. Preferences are only the constructs that summarize choices. We need to know the individual’s beliefs relevant to the perception of the choice situation. According to Hausman, one of the critics of revealed preference theory: Beliefs mediate the relationship between choices and the preferences with which economists are concerned. Economists can infer preferences from choices or choices from preferences only given premises concerning the agent’s beliefs. Different preferences can lead to the same action, depending on what the agent believes. Neither beliefs nor preferences can be identified from choice data without assumptions about the other. 2012, p. 30

Preferences, utility, and welfare  31

Three factors—​ preferences, beliefs, and choices—​ are involved. Preferences are mental states that cause our actions, such as choosing one good over another, along with our beliefs. Sen is critical toward the idea of rationality as internal consistency of choice: rationality of choice is identified with a set of predefined axioms of consistency. Following Sen (2002), rationality is broadly interpreted as the discipline of subjecting one’s choices to reasoned scrutiny. Beyond maximizing some quantity (which implies that choices exhibit some consistency), Sen (2002, p. 36) claims that: A person is not only an entity that can enjoy one’s own consumption, experience and appreciate one’s welfare, and have one’s goals, but also an entity that can examine one’s values and objectives and choose in the light of those values and objectives … We might or might not be much moved by moral concerns or by social reasons, but neither are we prohibited from entertaining these questions, in shaping our values and if necessary revising our objectives in that light. Sen (1993; 1997) argues that the revealed preference axioms will be violated by introducing the problem of ‘menu-​dependence.’The actual choice of the consumer depends on the content of the set of available alternatives. According to Sen (1993; 1997), many perfectly rational choices will violate the axioms of revealed preference when the choice situation, or the menu, changes. Such an argument seems to be close to that of the recent behavioral economics. People do not always exhibit the consistent behavior described by neoclassical economics and its revealed-​ choice foundation. In a behavioral setting, preferences are often incomplete and unstable. The change in the choice space can affect the relative preference for two different bundles even if both bundles are available before and after the change.The preference ordering of the person changes with the set of available alternatives. Furthermore, the change in the choice space may involve the introduction of a social or moral obligation. That is, what people choose reflects what they think they ought to do socially or morally, which does not reveal their true preferences. Rules, in some cases, may take the form of self-​imposed constraints that lead to a discrepancy between choice and true preferences, generating menu-​dependent preferences. According to Sen, the theory of revealed preferences cannot be defended on purely empirical grounds: “the number of actual choices that can be studied is extremely limited” (1973, p. 243). As we see in (Section 2.1), Samuelson’s (1938) analysis starts with a set of demand functions satisfying what would be chosen for any set of prices and any income level. In discussing the theory of revealed preferences, the class of budget sets that a consumer may face is an infinite class, and we cannot possibly observe the consumer’s choices from an infinite number of budget sets. Empirical applications of revealed preference theory are based on finite choice data. Observed consistency with respect to only the subset of choice data is

32  Preferences, utility, and welfare

not sufficient for consistency over all of the relevant parameters. Even if the postulate is justified for a given number of market-​choice situations, there is no guarantee that it will be justified in some future market-​choice situation.

2.1.3  Behavioral economics and revealed preference theory Some researchers insist that policy evaluations must maintain a strict adherence to the principle of revealed preferences. In the neoclassical economic approach, the terms ‘utility maximization’ and ‘choice’ are synonymous. Individuals in the approach have well-​defined preferences and make decisions to maximize those preferences. ‘Well-​defined’ preferences, in economics, satisfy two axioms: completeness and transitivity. Completeness means that between any two alternatives x and y, the agent must either strictly prefer x to y, strictly prefer y to x, or be indifferent between x and y.Transitivity, on the other hand, means that for any three alternatives x, y, and z, if x is preferred to y and y is preferred to z, then x must be preferred to z (the preferences involved may be either weak or strict). Completeness and transitivity serve to rule out internal contradictions and to impose a form of consistency over the entire set of preferences. Those preferences are supposed to accurately reflect the true costs and benefits of the available options.3 It has been observed that people often prefer immediate gratification. Under hyperbolic preferences, the discount rate from today’s perspective between today and tomorrow is larger than the discount rate between any two consecutive dates in the future: the further away is a payoff, the higher is individual’s patience. When preferences are hyperbolic, an individual will prefer a larger reward (x) at time t + 1 to a smaller reward (y) at time t, and yet prefer y to x when t and t + 1 have both come closer to the present. This can be characterized as a violation of either completeness, transitivity, or both. Concerning temptation and self-​control, Gul and Pesendorfer (2001; 2004) argue that observed behavioral anomalies in individual’s decision-​making should be explained by an extension of the preference domain. They propose defining the preference domain over both allocations and choice sets. In this case, there is no conflict between preference and choice, and the principle of revealed preferences can be applied to recover preferences from choice data under some identifying assumptions. If some choices feel tempting (when available) and this detracts from well-​being, an individual may prefer small choice sets to large ones. If the individual is sophisticated and can correctly forecast the effect of future temptations, he or she could prefer a commitment device that limits his or her future choice sets. As in the neoclassical economic framework, individuals maintain the same lifetime preference ranking at every moment in time. A desire to constrain future choices, in this setting, does not imply that preferences change over time. Other scholars, on the other hand, investigate the possibility to relax, modify, or depart from the ordinary revealed preference approach.The behavioral incoherence caused by framing renders the ordinary revealed preference approach infeasible. Preferences express individuals’ subjective judgments about their welfare. In conventional economic theory, revealed preferences are preferences that represent

Preferences, utility, and welfare  33

the economic agent’s true interests. A revealed preference approach provides a convenient welfare standard: what is good for an individual is what that individual would choose. Neoclassical welfare economics assumes that preferences depend only on the attributes of the items in the choice set and thus are unaffected by the framing of the choice. It is built on the notion that we can learn what is better for an individual by observing one’s choices. However, there are many cases in which the choices of consumers do not reveal true preferences, but rather reflect the combined influence of true preferences and decision-​ making errors. For instance, in many situations, individuals passively accept defaults that are chosen by others. Accepted outcomes vary with the type of default that is used. The default biases outcomes. Since variation in defaults generates large variation in outcomes, preferences revealed through passive choice are often inconsistent. If behavioral failures lead to biased choices, it is not clear that observed choices should be used to infer preferences. Selection does not imply that the option is at least weakly better than other options that were available. Following Bernheim (2016), neoclassical welfare economics is based on the following preference-​theoretic premises: Premise 1:  Each individual is the best judge of one’s own well-​being. Premise 2:  Each individual has coherent, stable preferences. Premise 3:  Each individual’s preferences guide one’s choices: when one chooses, one seeks to benefit oneself. Behavioral economics challenges the validity of these premises, and it arguably calls for a new welfare paradigm. Contrary to premise 1, behavioral economics shows that people are fallible (easily confused, not that smart, and often irrational) and they do not reliably exercise good judgment. Contrary to premise 2, preferences are not invariant to context. Behavioral economics shows that observed choices are highly context dependent and that framing greatly influences individual’s decision. Given these observations, welfare criteria are no longer based on the notion of allocation of resources. Therefore, we might separate between positive models describing choices and normative ones describing welfare. The evidence of time-​inconsistent behavior also motivates a relaxation of premise 2. As a modeling strategy, we can endow individuals with well-​behaved lifetime preferences that vary at different points in time. Policymakers will not be able to accurately predict a priori how decision-​makers will respond to potential policies using conventional economic theory. The evidence that preferences and choices diverge motivates a relaxation of premise 3. In neoclassical welfare economics, choice is a result of preferences that at once represent well-​being and behavior. In behavioral economics, however, individuals do not truly maximize a utility function that both measures their well-​being and guides all of their choices. Individuals often choose to do things that are not in their own best interest. True preferences actually exist inside their heads, but different versions of preferences may either coexist or replace each other. Then, true preferences are considered as those preferences that individuals would

34  Preferences, utility, and welfare

have acted on, had their reasoning not been impaired by psychological biases. True preferences preexist independently of the particular framing of the choice problem. Individuals therefore access true preferences imperfectly when they are called upon to make a choice. Unlike revealed preferences, true preferences may not be revealed through choice. Psychological factors, such as limited attention, inferior cognitive abilities, or lack of self-​control, can operate against the satisfaction of individual’s true preferences. The choice seems to unambiguously leave individuals worse off than they could be. Individuals attempt to optimize utility according to their true preferences, but they encounter cognitive or emotional conditions which trigger systematic mistakes. If individuals make mistakes, and if these mistakes are systematic, it might be possible to make them better off by restricting their choices and putting attractive bad choices off limits. Thus, it is feasible to improve welfare by restricting or altering individual choice. If the true preference set is reconstructed, the satisfaction of these preferences could be used as a basis for welfare-​enhancing policy proposals. Behavioral economists have questioned the satisfaction of revealed preferences as the normative foundation of neoclassical welfare economics. For a variety of reasons, individuals do not possess stable, context-​independent revealed preferences. People are systematically affected by factors that have little or no effect on their welfare, interests, or goals. Then, preferences revealed through choice cannot be used as a measure of welfare for policy purposes. We have no basis to rely on when determining what makes an individual better off. We have to tackle the problem of making welfare claims when consumer choices are inconsistent. Following Bernheim (2009, p. 315), knowledge of a choice correspondence may shed insufficient light on objectives, and hence on the mapping from the objects of choice to well-​being. One can attempt to identify welfare, partially or completely, by restricting the set of allowable unconventional rationalizations, but useful restrictions are difficult to justify. Those conceptual difficulties have led some to argue that economists should try to infer well-​being from self-​reported happiness and/​or neurobiological activity. Unfortunately, it is every bit as problematic to identify useful information concerning internal well-​being from such data as it is from choice. In order to reconstruct what the agent really prefers, we must first identify the subset of preferences whose satisfaction is taken to reliably increase one’s well-​ being. Bernheim and Rangel (2009) provide a choice-​theoretic welfare analysis that allows for the individual’s inconsistent behavior impaired by ‘ancillary conditions.’ Ancillary conditions are properties of the choice environment which may affect behavior, but which the planner treats as normatively irrelevant. Bernheim and Rangel’s (2009) approach is to propose a criterion that respects the individual’s revealed preferences over pairs of objects if those preferences are not affected by changes in ancillary conditions. Their analysis only requires observed choices to

Preferences, utility, and welfare  35

make statements about welfare. If the choice between objects is unaffected by changes in ancillary conditions, the underlying preference will be accepted as a true preference. This ‘purifies’ observed choices from individual’s choices treated as decision errors. Their purely choice-​based concept has the benefit of not requiring any conjecture about the decision process. Considering only the purified data, the revealed preferences are considered to be a reflection of welfare-​enhancing choice. They suggest a welfare preference relation that ranks an alternative as welfare superior to another only if the latter is never chosen when the former is available. An individual is weakly better off with x than y if there is no frame at which the individual chooses y but not x when both are available.

2.2  Utility 2.2.1  What is utility? The concept of ‘utility’ has been defined in at least two different ways. First, utility can be seen as the ultimate ‘good’ and, therefore, the natural aim of public policy. This is intimately related to ‘well-​being.’ Second, utility can be seen as a rationalization of individual behavior (choice-​based utility). Utility is directly related to individual behavior. Neoclassical economic theory generally explains consumer behavior in terms of a particular version of rational choice theory: utility maximization subject to a budget constraint. Thus, individual behavior follows from utility maximization, and the same measure of utility corresponds with well-​being, or the aim of public policy. In the field of economics, there are different ideas about what (choice-​based) utility is. One key divide is between ‘cardinal’ and ‘ordinal’ interpretations of utility. A cardinal utility function yields a real-​number value for each possible state of affairs. On the other hand, an ordinal utility function for an individual consists of a rank ordering of possible states of affairs for that individual. The standard framework of consumer choice has been ordinal utility theory since economists early in the 20th century. The arguments of the consumer’s utility function are assumed to be the quantities of the various goods consumed as well as the amount of leisure. The budget constraint requires the total outlay on consumer goods not to exceed the amount of labor income. There are ways to extend the models to take account of time and uncertainty, thereby to explain saving and portfolio behavior. Neoclassical economic analysis is built up from theories of individual behavior that have the origins in the 1870s, beginning with the work of William Stanley Jevons, Carl Menger, and Léon Walras.The ‘marginal revolution’ pioneered by these scholars focused economists’ attention on the ideas of utility and utility maximization. How should we interpret the concept of utility? For Jeremy Bentham and his followers, utility is a psychological (or physiological) magnitude which measures an individual’s inner happiness.Thus, under ideal conditions, utility could be treated as an observable quantity of pleasure with the same measurable properties as weight. Following Jevons, “[p]‌leasure and pain are undoubtedly the ultimate objects of the

36  Preferences, utility, and welfare

Calculus of Economics.To satisfy our wants to the utmost with the least effort … in other words, to maximize pleasure, is the problem of Economics” ([1871] 1965, p. 37, original emphasis). The early neoclassical economists understood utility in terms of conscious experience like pleasure or happiness. In an account of deliberation, individuals weight the pleasure and pain that would result from various actions and choose the one they perceive as leading to the greatest balance of pleasure over pain. According to Francis Ysidro Edgeworth, with his proposed ‘hedometer,’ utilities are real and measurable. Early choice theory, starting in the 1870s, employed a ‘cardinal’ and ‘hedonistic’ notion of utility. Utility is cardinal in the sense that differences in the valuation of various bundles of goods take on numerical values, and it is hedonistic in the sense that levels of utility are associated with the amount of pleasurable psychic feeling received from the bundle. Utility maximization could be viewed not only as a means for identifying which choices individuals would make but also as a reasonable model of the ‘process’ by which they made choices. The first-​generation neoclassical theory was clearly psychological. Psychology was exclusively identified with hedonistic psychology at that time. The cardinal and hedonistic notion of utility was obviously linked to utilitarian ethics, making interpersonal utility comparisons possible. In his Foundations of Economic Analysis, Samuelson argued as follows: “to a man like Edgeworth, steeped as he was in the Utilitarian tradition, individual utility—​nay social utility—​was as real as his morning jam” (1947 [1983], p. 206). Utilitarianism requires both cardinality and full interpersonal comparability. This point about utilitarianism is closely related to welfare economics, the normative branch of economic theory. If the goal is to maximize the sum of human pleasurable psychic feelings, it would be useful to add up the utility levels of different individuals. Following this utilitarian principle, social welfare is measured by summing utility numbers for all the members of society. This principle asserts that the best public policy is the one which gives the greatest social welfare to the individual members.4 However, both cardinality and interpersonal comparability pose measurement problems for economists. Even in the case of a single person, it is difficult to measure cardinal utilities reliably. Measurements that support interpersonal comparisons are more difficult to justify. William James, in his 1890 book The Principles of Psychology, criticized psychological hedonism as unscientific. According to James, psychological hedonists assume that behavior always aims at the goal of maximum pleasure and minimum pain, but behavior is often impulsive, not goal oriented. Even though pleasure and pain were well-​defined magnitudes, it would still be unrealistic to model humans as goal oriented. Hedonistic psychology was based entirely on the internal subjective feelings of the individuals. By the rise of experimental psychology in the early 20th century, the introspective method of inner observation was no longer counted as an empirical one. Psychology was then identified with hedonistic psychology exclusively. Psychologists were largely comfortable with talking about mental states. Psychological hedonism, based on introspection, became the main point of attack for critics of marginal utility theory. Therefore, it became necessary for marginalist

Preferences, utility, and welfare  37

economic theory to reform its foundations. This movement contributed to the replacement of the older notion of cardinal utility with the new notion of ordinal utility. According to the theory of ordinal utility, the fundamental assumption is that people have preferences. An individual’s preference ordering just represents his or her ranking of whatever options are available. Provided these preferences conform to some axioms, they can be represented by a utility function. The values taken by the function are called utilities. Of any pair of alternatives, the function assigns a greater utility to the one that is preferred. For the proponents of ordinal utility theory, questions of motivation and preference formation are outside the scope of economics. Utilities have nothing to do with psychological states. The internal mental states associated with utility are not necessary to understand preferences. Utility is then taken as an index or a measurement of preference satisfaction. Each individual has utility: an individual’s utility function is U : X → .5 The number U ( x ) is the utility that the individual obtains from x ∈ X . It is supposed that one’s utility function represents the preferences in the sense that for all x, y ∈ X , U (x ) ≥ U ( y ) ↔ x  y


That is, more utility is obtained with more preferred alternatives, and the same utility is obtained with alternatives that are indifferent to each other.The expression in (2.2) does not indicate whether utility or preference is the primitive concept. (2.2) is used to define utility from preference or, alternatively, preference is defined from utility using (2.2). In the former case, utility only has meaning as a representation of preference. The utility function preserves the order of preference, but it has no other meaning. The centerpiece of revealed preference theory is the belief that individuals make consistent choices, with utility and utility maximization simply being convenient analytical devices to describe these choices. Harsanyi (1955) presents a mental state account of well-​being. According to Harsanyi (1955), an individual’s utility is determined by a function of an alternative and the causal variables that determine the tastes and other subjective factors. The causal variables comprehensively describe things that influence an individual’s well-​ being other than the alternative being evaluated itself, such as his or her biological features and life history. Formally, for all x ∈ X , individual i’s utility is given by:

( )

U i ( x ) = v x ; hi ,


where hi denotes the values of the causal variables for individual i. If i’s tastes change, the utility obtained with x in (2.3) can differ before and after the tastes change. An alternative basis for making utility comparisons is provided by the concept of an extended preference. A social situation is a pair (x; i) consisting of being individual i complete with i’s characteristics, both objective and subjective, when the alternative is x. Formally, the problem of interpersonal comparisons can be expressed as follows. O is a set of outcomes, i.e., states of affairs, {x, y, z, …}. N is a set of

38  Preferences, utility, and welfare

individuals, {1, 2, …, N}. A ‘history’ (x; i) is a pairing of an individual i and an outcome x. The set H is the set of all histories. That is, H is O ×� N. The basic operation in interpersonal comparisons is imaginative empathy. We imagine ourselves to be in the shoes of another person. For Harsanyi, empathy is the capacity to assume someone else’s perspective. The spectator develops extended preferences by exercising this capacity. The spectator is asked to develop a preference as between (x; i) and (y; j) by considering, first, the outcome in which ‘I have all the attributes of i in x’ and, second, the outcome in which ‘I have all the attributes of j in y.’ The spectator does not know the referent for ‘I.’ The empathetic projection by the spectator is nothing other than the thought exercise of taking on the properties of each subject. The spectator ought to judge the utility of another individual’s position not in terms of one’s own attitudes but rather in terms of the attitudes of the individual actually holding the position. The empathy-​based conception of extended preferences might ask for the spectator to be placed under a kind of ‘veil of ignorance’ about one’s own attributes.6 The term ‘veil of ignorance’ is then used to denote any construction that translates outcomes into the bundles of histories therein and appeals to spectators’ preference (extended preferences) over these histories for some normative purpose. If the individual does not know what one’s own position is, one should give equal probabilities to the various positions that one can conceivably hold.

2.2.2  Expected utility theory Expected utility theory states that the decision-​ maker chooses among risky prospects by comparing their expected utility values. In this approach, preferences are defined over prospects, where a prospect is to be understood as a list of consequences with associated probabilities. It is assumed that all consequences and probabilities are known to individuals. In choosing among prospects, they are said to confront a situation of risk. In these expected utility models, people maximize the weighted sums obtained by adding the utility values of outcomes multiplied by their respective probabilities. The principle of maximizing expected monetary values antedates expected utility theory. The issue was why people would pay only a small dollar amount for a game of infinite mathematical expectation. A fair coin is flipped repeatedly until a head is produced. In the so-​called St. Petersburg game, people are asked how much they will pay for the following prospect: if tails comes out of the first toss of a fair coin, to receive nothing and stop the game, and in the complementary case, to receive $2 and stay in the game; if tails comes out of the second toss of the coin, to receive nothing and stop the game, and in the complementary case, to receive $4 and stay in the game; and so on ad infinitum.The payoff will then be $2k, where k is the number of the throw producing the first head. This means the game has many possible outcomes: $2, 4, …, 2k, with probabilities 1/​2, 1/​4, …, (1/​2)k, respectively.The expected monetary value of this prospect is infinite, since

∞ k=1

(1 / 2)k 2k = ∞. Most people would only be prepared to pay a relatively

Preferences, utility, and welfare  39

small amount of money to enter it. Individuals place subjective values, or utilities, on monetary outcomes. In neoclassical economics, preferences are axiomatically required to be ‘rational’ so that choices are rational. As long as an individual’s preferences over risky options obey certain axioms, the individual behaves as if he or she is maximizing the expected value of a utility function. It was not until von Neumann and Morgenstern (1947) that expected utility maximization was formally proved to be a rational decision criterion. The preferences of an individual among prospects are governed solely by the outcomes and their respective probabilities. For an individual, the outcomes of a risky prospect are defined as x1, …, xn. Associated with each outcome xk is the probability of its occurrence qk, k = 1, …, n. We denote a fixed set of pure consequences by x = (x1, …, xn) and denote the associated probability distribution by q = (q1, …, qn) such that


q k =1 k

= 1. Such a prospect is symbolized as (x, q). A compound pro-

spect [(x, q), p; (y, r), 1−p], 0 < p < 1, is one which offers a probability p of receiving the outcome x (= x1, …, xn) with probability q (= q1, …, qn and


q k =1 k

= 1) and

a probability 1− p of receiving the outcome y (= y1, …, yn) with probability r (= r1, …, rn and


r k =1 k

= 1). Preferences over prospects can be represented by a function

V(.) which assigns a real-​valued index to each prospect. The function V(.) is a representation of preference in the sense that V(x, q) > V(y, r) if and only if (x, q)  (y, r), where  represents the relation ‘is weakly preferred to.’ Then, an individual will choose the prospect (x, q) over the prospect (y, r) if and only if the value assigned to (x, q) by V(.) is no less than that assigned to (y, r). The expected utility hypothesis can be derived from three axioms: ordering, continuity, and independence. The ordering axiom requires both completeness and transitivity. Completeness entails that, for all prospects (x, q) and (y, r), either (x, q)  (y, r) or (y, r)  (x, q) or both. Transitivity requires that, for all prospects (x, q), (y, r), and (z, s), if (x, q)  (y, r) and (y, r)  (z, s), then (x, q)  (z, s). Continuity requires that, for all prospects (x, q), (y, r), (z, s) where (x, q)  (y, r) and (y, r)  (z, s), there exists some p such that [(x, q), p; (z, s), 1− p] ~ (y, r), where ~ represents the relation of indifference. Continuity implies that all prospects of intermediate preference are indifferent to some mixture between a pair of better and worse prospects. Independence requires that, for all prospects (x, q), (y, r), (z, s) and for all p, if (x, q)  (y, r), then [(x, q), p; (z, s), 1− p]  [(y, r), p; (z, s), 1− p]. The independence axiom states that a preference between two prospects should be independent of any common (or irrelevant) components, as if those common components are cancelled out. If all three axioms hold, preferences can be represented by: V (x, q ) =

∑q u(x ), k




where (x, q) is any prospect, and u(.) is a utility function defined on the set of consequences. This utility function is unique up to positive linear transformations,

40  Preferences, utility, and welfare

meaning that if the function u(.) represents an individual’s preferences then so will u*(.) if and only if u*(.) = au(.) + b for numbers a > 0 and b. Put differently, people can arbitrarily choose numbers for the origin and the scale of their utility function. In expected utility theory, the concept of risk aversion is important. If some gamble is less preferred than its expected monetary value for sure, the preference is said to be risk averse. A concave utility function implies risk-​averse behavior. That is, an individual with a concave utility function u(x) will always have a certain amount x to any risky prospect with expected value equal to x. Expected utility theory imposes more structure on individual rationality than does utility theory under certainty. However, expected utility theory has been criticized to be flawed as a theory of rationality. For Maurice Allais (1953), expected utility theory is not an adequate characterization of individual risk preferences.The well-​known paradox is due to Allais (1953), ‘Allais paradox.’ Consider the following pair of hypothetical problems: X1 vs. Y1 and X2 vs. Y2. In the first, the decision-​ maker has to choose between the two prospects: X1 = ($100, 1) or Y1 = ($500, 0.1; $100, 0.89; $0, 0.01). The first option gives 100 US dollars for sure; the second option gives $500 with a probability of 0.1; $100 with a probability of 0.89, otherwise nothing. As a second problem, the decision-​ maker has to choose between the two prospects: X2 = ($100, 0.11; $0, 0.89) or Y2 = ($500, 0.1; $0, 0.9). Allais (1953) expected that people faced with these choices might prefer X1 to Y1 in the first problem and Y2 to X2 in the second problem. Suppose we set u($0) = 0 and u($500) = 1. Then expected utility theory implies that X1 is preferred to Y1 if and only if: u($100) > 0.1 + 0.89u($100).(2.5) We can subtract 0.89u($100) from both sides of (2.5): 0.01u($100) > 0.1


But Y2 preferred to X2 implies that 0.01u($100) < 0.1, a direct contradiction of (2.6). The rule to maximize expected utility purports to yield rational decisions when options involve risk. However, it has been criticized for neglecting to take

Preferences, utility, and welfare  41

account of a decision-​maker’s attitude toward risk. With respect to the hypothesis of expected utility maximization, people are not very good at judging probabilities. They systematically violate just about every logical implication of decision theory. For Allais, utility is psychologically real. His treatment of risk is very natural. Suppose that an agent, having wealth of $a, is faced with a choice between $5 for sure (option A), and either $9 or $1 on the flip of a fair coin (option B). By applying the utility function u, option A yields u(a + 5) for sure, while option B yields u(a + 9) or u(a + 1) with a probability of 0.5 each. On the expected utility view, the agent chooses the option with the highest expected utility, that is, max[u(a + 5), 0.5u(a + 1) + 0.5u(a + 9)]. On Allais’ view, however, an agent’s choice among gambles might be influenced by the variance in utility that one receives. Since option A has zero variance and option B has a positive variance, the agent might prefer option A to B.The agent’s attitude toward risk can be reflected in one’s attention to the variance as well as the expectation of the utilities associated with each option. Many patterns of observed economic behavior under risk cannot be explained by standard expected utility theory. Experiments find that gains and losses are evaluated quite differently. Outcomes are interpreted as gains and losses relative to a reference point. We may think of the reference point as status quo wealth. By integrating certain psychological aspects of decision-​making, Kahneman and Tversky (1979) challenge standard expected utility theory. They present an alternative model of choice, prospect theory, based on experimental findings. In prospect theory, outcomes are evaluated via a utility function with the shape that is kinked at the reference point. The utility function is concave for gains and convex for losses, and it is steeper in the domain of losses. This involves loss aversion. That is, the disutility of giving up something is greater than the utility associated with acquiring something. As an important behavioral feature, people overweight probabilities of outcomes that are considered certain in relation to other outcomes. This is called the certainty effect: a risk-​averse preference for a sure gain over a probable gain with a larger expected value, and a risk-​seeking preference for a probable loss over a certain loss with a smaller expected value.This is generalized as the reflection effect, which specifies that people are risk averse when the outcome is positive and risk seeking when it is negative.

2.3  Welfare: personal and social The change in welfare economics, from the ‘old’ to the ‘new’ one, was associated with the move from cardinal to ordinal utility within consumer choice theory during 1930s. The old (hedonistic-​utilitarian) welfare economics was criticized by arguing against the possibility of making purely scientific interpersonal utility comparisons. The new welfare economics, associated originally with Vilfredo Pareto, has been formalized by economists (Bergson, 1938; Lange, 1942; Samuelson, 1947 [1983]). For Bergson (1938, p. 310), the object of his paper is “to state in a precise form the value judgments required for the derivation of the conditions of maximum economic welfare which have been advanced in the studies of the

42  Preferences, utility, and welfare

Cambridge economists, Pareto and Barone, and Mr. Lerner.” The idea of a welfare function is utilitarian in the sense that its goal is to measure individual wants and to construct an index of utility. In the old welfare economics, utilitarianism focused on trade-​offs between different members of society and thus had a definite moral content. With the ordinalist revolution and the positivist twist on utilitarianism, the moral content of welfare theory was abandoned. Public policies affect a lot of people in different ways. Any policy decision therefore involves a judgment of its desirability.Welfare principles have to be based on the prevailing values of the community being discussed. However, Bergson did not discuss where his social welfare function might come from. It is simply a mathematical statement, the explanation of which is left open pending discussion later on. The economic analysis of public policy is concerned with evaluating the desirability of policy effects. It is then necessary to evaluate different states of the world in terms of their distributional results. The purpose of welfare economics is to obtain a social ordering over alternative possible states of the world, thus promoting normative principles.7 However, figuring out what policy changes are desirable when everyone’s interests are taken into account is a very complicated problem. Judging whether a change is good or bad requires comparing one person’s welfare to another’s. A Pareto-​optimal allocation of resources is one of the key principles of the new welfare economics. If it is possible to make one person better off without making someone else worse off, there exists a potential Pareto improvement. Thus, a Pareto-​optimal allocation is one from which there does not exist such a potential Pareto improvement. One procedure to evaluate economic outcomes is to use a social welfare function to determine whether social welfare increases or not. The social welfare function represents some ethical judgment about the appropriate distribution of welfare across people affected by a policy change. In welfare economics, a social welfare function assigns a numerical index of social welfare (W) to each of the relevant states based on the values of relevant society.The social welfare function is completely general, not specifying the variables entering the function. Let h denote an index for the set of individual consumers, where h = 1, …, H. For the set of individual consumers, social welfare is described as a function of the allocation x: W = W (x )



= W x1 , , x H , where xh is the consumption bundle of individual h. The social welfare function is nearly symmetric with respect to the consumption of all individuals (that everyone counts for approximately the same). The new welfare economics presents a general framework for welfare analysis that is not dependent on any specific set of ethical commitments, while at the same time accommodating various ethical views. As Bergson (1938, p. 328) explains, “[f]‌or only if the welfare principles are based upon prevailing values, can they be relevant to the activity of the community in question.”

Preferences, utility, and welfare  43

The problem facing the policymakers is then to choose xh for h = 1, …, H so as to solve the following problem:


max W x1 ,… , x H x


such that H




h =1

and F ( X ) ≤ 0, where X is the aggregate consumption bundle, and F(X) is the transformation function available to society. In this problem, it is assumed that the policymakers can choose the allocation directly. An alternative hypothesis is to assume that they use certain limited instruments (such as taxes, quotas, etc.) to determine the allocation. In the new welfare economics, an ‘individualistic’ social welfare function, based on individual valuations, is specified as a more restrictive functional form. In its individualistic form, a social welfare function is constructed in two stages: the individual utilities are first determined and then these utilities are aggregated into the value of social welfare. The social welfare function gives the welfare of the whole society as a function of the utilities of individuals, just as the utility function gives the welfare of the individual as a function of the quantities of goods that he or she consumes. It represents a value judgment of how society should rank different distributions of utility across people. A just allocation that maximizes an individualistic social welfare function can be achieved by the use of lump-​sum taxes and transfers, then relying on the market to reach a competitive equilibrium. Given the utility function of individual h, Uh(xh), the social welfare function is given by:

( ( )

( ))

W = W U 1 x1 , ,U H x H with

∂W (⋅) ∂U h

> 0 for all h,

where each utility function Uh(.) is assumed to be concave, and social welfare is a concave function W(.) of individual utilities. The marginal change in social welfare from a change in individual h’s consumption is: ∂W ∂x h


∂W ∂U h ∂U h ∂x h


which depends on the W and Uh functions and the level of xh. Adding up all the changes in social welfare from the changes in each individual’s consumption caused

44  Preferences, utility, and welfare

by a policy gives the welfare effects of that policy. The change in social welfare for an increase in individual h’s consumption decreases as xh increases. Dollars to low-​ income people count more highly for social welfare than dollars to high-​income people. Each Uh is a cardinal, interpersonally comparable utility value yielded by some procedure for individual h. The social welfare function necessarily makes the implicit or explicit interpersonal utility comparisons. Without assumptions concerning interpersonal comparisons of utility, it is impossible to identify the social optimal allocation (Samuelson, 1947 [1983]). To support the social utility function, we allow full interpersonal comparability and cardinal utility information. Its purpose is to make interpersonal welfare judgments. A utility function can be deduced from observing the choices an individual actually makes. However, a social welfare function cannot be observed, and it must be specified according to a particular ethical judgment. Restrictions are imposed on the social welfare function, embodying normative judgments. Although there is no agreement on the appropriate W(.) function, it is possible to specify either more restrictive functional forms or explicit ones for the W(.) as well as each of the Uh(.)s. For example, if one takes W(.) to be additively separable and takes each Uh(.) to be a cardinal indicator of each individual’s utility, one would have the traditional utilitarian social welfare function.8 The social welfare function would be useful for making interpersonal utility comparison explicit and seeing the implications of different value judgments. Economics has insisted, since Robbins’ (1932) description of the subject, on a strict separation between ‘positive’ and ‘normative.’ The traditional interpretation of the distinction between positive and normative in economics has been that positive is about ‘what is’ and normative is about ‘what ought to be in order to be moral.’ According to Robbins (1932), economics cannot pronounce on the validity of ultimate judgments of value. Most economists do not want to impose their own values. However, many policy choices involve value judgments. To answer important questions of income distribution and taxation requires some criteria of justice. The idea of a social welfare function is part of normative economics. According to Samuelson (1938, p. 261), “[a]‌t the outset, it is understood of course that every discussion of welfare economics implies certain ethical assumptions. I do not propose, however, to discuss the philosophical grounds for holding or rejecting different ethical precepts or assumptions.” The new welfare economics claims that many policies can be shown to be good or bad by avoiding the issues that arguably make it so difficult. In order to make welfare judgments, it is only necessary that economists agree on the definition of an ordinal function involving as variables the quantities of goods consumed by all individuals. Ethical judgments have to be imposed, but the emphasis is on what is called Pareto optimality. The concept of Pareto optimality itself involves value judgments.9 Economists have often argued that a policy is welfare improving if the winners can compensate the losers. This is called the potential Pareto improvement criterion. Pointing out that Pareto optimality involves ethical values only draws attention to the fact that ethical values are involved, and the door is open to consideration of other possible ethical values.

Preferences, utility, and welfare  45

Much discussion in economics has drawn on the two fundamental theorems of welfare economics. First, a perfectly competitive equilibrium is Pareto optimal.The first fundamental theorem of welfare economics shows that competitive markets are an efficient way to satisfy given wants with scarce means. However, a competitive equilibrium may fail to develop (e.g., economies of scale) or it may not be still Pareto optimal (possibly, because of externalities). Even though the first fundamental theorem guarantees that competitive markets will lead to a Pareto-​optimal outcome, this outcome need not be a just one. Second, any Pareto-​optimal outcome can be achieved by lump-​sum taxes and transfers and the use of markets to achieve a competitive equilibrium. The second fundamental theorem of welfare economics has formed much of the basis of the discussion of the relation of markets to justice. The goal of justice can be expressed as maximizing an individualistic social welfare function that depends on each individual’s utility (with each individual’s utility function depending only on the goods received). Welfare economics can specify an objective function (e.g., equal outcomes, or the maximin criterion) and aggregate individual utilities with appropriate weights. At the maximum, it must be Pareto optimal; otherwise we could increase the value of the function by making someone better off without making someone else worse off. From the second theorem, we know that we can achieve this Pareto-​optimal point with appropriate lump-​sum transfers and competitive markets. To judge well-​being in terms of utility is what traditional welfare economics tends to do. In traditional welfare economics, policies have been evaluated by examining the responses to economic incentives, with values, habits, and norms taken as given and beyond analysis. Sen (1982) is largely responsible for reminding us that there is more to life than achieving utility. The utility approach provides an incomplete account of well-​being that cannot be used as a reliable proxy for human development. Many types of nonutility information have intrinsic importance for the assessment of well-​being. For example, consider two persons who are both undernourished. For the first person, the undernourishment is the result of material deprivation. The second person, on the contrary, is wealthy but freely chooses to fast for religious reasons. It is intuitive to say that their situations are not the same in terms of well-​being. Sen, therefore, introduces the notion of ‘capabilities’ to capture the real opportunities of persons. The capability of an individual can be considered as the possibility of achieving various combinations of ‘functionings.’ Functionings include not only doings but also states of being.10 Thus, physical and mental states (such as being in good health, being happy, etc.), as well as physical and mental activities, count as functionings. For example, being happy is just one of many aspects of being that are relevant to an overall evaluation of well-​being.11 The quality of one’s life depends not only on one’s functionings but also on one’s opportunity to function. Although functionings are important, it is capability, the extent to which a person is able to function in a particular way, that is the basis upon which Sen recommends evaluation. Sen’s particular concern is to advocate the capability approach as an alternative to traditional welfare economics. Capabilities are opportunities to function. A person’s state of being is understood

46  Preferences, utility, and welfare

as a vector of functionings. Functionings need to be understood as including both desirable functionings (e.g., states of pleasure) and undesirable functionings (e.g., states of pain). The set of feasible vectors for any person is the person’s capability set. The chosen combination of functionings is one of the elements of that set. Functionings are distinguished from commodities: commodities are objects which a person might use, while a functioning is an aspect of living itself. Sen’s capability approach provides a conceptual framework that allows us to go beyond the usual framework of individual resource allocation to consider how resources can best be used to achieve individual freedom.The well-​being of a person must be thoroughly dependent on the functionings achieved. Capabilities, defined as the opportunity set from which people can choose, are a reflection of the real freedom of individuals. Sen’s theory of functionings and capabilities does not prescribe any particular functionings and capabilities. Different capabilities are likely to be important in different contexts.

2.4  Well-​being and happiness Well-​being can be broadly considered as people’s positive evaluations of their lives. Economic measures of income have ignored large areas of well-​being and are poor measures of well-​being. However, increasing recognition of these distortions has had relatively little practical impact: “GNP per capita continues to be regarded as the quintessential indicator of a country’s living standard” (Dasgupta, 2001, p. 53). Gross national product (GNP) statistics, as is well known, measure current economic activity, but they ignore wealth variation, household production of services, destruction of the natural environment, and many determinants of well-​being such as personal safety, health, and longevity. Economists and psychologists operate with different accounts of well-​being. Higher GNP does not mean higher social welfare. If people care about other aspects of their lives, such as their health, the environment, leisure, and so on, and if some of these variables are negatively correlated with GNP per capita, then higher GNP might accompany an unchanging level of happiness. In fact, GNP increases when rising crime, pollution, or health hazards trigger defensive expenditures. This insight is related to the idea of replacing GNP per capita in favor of broader measures of welfare. Taking a psychological perspective, we could investigate what is the best measure of well-​being for describing the emotion of feeling well. Subjective well-​being is often used by psychologists as an umbrella term for how people think and feel about their lives. It takes an individual’s well-​being to be one’s overall assessment of one’s life. Economic indicators should be supplemented with well-​being indicators. In conventional economic theory, other things being equal, more choices mean a higher quality of life because people with choices can select courses of action that maximize their well-​being. Since income correlates with number of choices, greater income is equivalent to higher well-​being. Therefore, social welfare has been related to the value of total income. Monetary measures still predominate in many contexts (money increases well-​being). Countries have grown wealthier over

Preferences, utility, and welfare  47

time, and they have made major improvements in other indicators, such as morbidity, morality, and literacy rates. However, as nations become wealthier and basic needs are largely met, economic indicators increasingly miss their target. More money does not make people happier. Happiness is based, not on one’s absolute income, but on a comparison to others, which in turn depends on one’s society and is adjusted as average incomes change. Whereas economists adopt some kind of preference satisfaction account, psychologists rely on some kind of happiness account. Happiness is seemingly one of the most important elements in everyone’s life. One definition of happiness is the degree to which a person judges the overall quality of his or her life as favorable. Different measures of welfare and different value judgments affect public policy. For psychologists, it is natural that happiness should be studied in part by asking people how they feel. Many economists doubt that subjective welfare can be used for interpersonal comparisons because, although ordinal preferences can be retrieved from observed choices, cardinal and interpersonally comparable measurement seems impossible. Economists have not traditionally trusted subjective responses. For them, utility should only be evaluated by means of preferences revealed in choice data. But it is nonetheless important to think of happiness surveys as complements to rather than substitutes for income-​based measures of progress. Happiness surveys can provide novel information and suggest new analytical approaches.These studies are directly relevant to policy analysis. The measurement of well-​being is an essential input for social evaluation. Economists should examine surveys of happiness to uncover the effects of factors such as unemployment, income inequality, commuting, and smoking. The use of responses to survey questions asking individuals their current level of satisfaction with life is at the heart of happiness research. Everyone has one’s own ideas about happiness and the good life; a person can be asked how satisfied one is with one’s own life. Unlike the more widely used economic measures (typically based on data about income, market transactions, and the like), subjective measures are based on answers to satisfaction questions: with subjective questions on well-​ being, people are asked about where on a scale of, for example, 0–​10, they are in terms of life satisfaction or happiness. As an example, “we would like to ask you about your satisfaction with your life in general. Please answer according to the following scale: 0 means ‘completely dissatisfied,’ 10 means ‘completely satisfied.’ How satisfied are you with your life, all things considered?”The subjective question is a straightforward strategy to ask people about their well-​being. Individual satisfaction does not depend solely on material consumption. The pleasant life is characterized by positive moods, positive emotions, and pleasures. Happiness studies quantify the relative importance of material and nonmaterial factors. Using the responses to similar subjective questions, researchers can examine the relationship between reported happiness and individuals’ circumstances, such as own income or income of the others, occupational status, family situation, health, inflation, unemployment rate, or the focus of their survey. Answers to satisfaction questions are not only determined by monetary income or material consumption. The direct

48  Preferences, utility, and welfare

measurement of well-​being captures aspects of welfare beyond real income. The subjective measures are typically presented as alternatives to, or improvements over, traditional economic measures for purposes of the evaluation of public policy. Some researchers treat such self-​reported well-​being data as measuring a simple notion of flow utility (i.e., welfare experienced contemporaneously or in the immediate recent past); other researchers treat such data as measuring lifetime utility (i.e., the expected present discounted value of flow utility from birth).12 A satisfaction question such as “How satisfied are you with life, all things considered?” is ambiguous on one aspect: it could be about all the dimensions of the current slice of life (‘these days’) or about the whole life span (‘your life as a whole’) as viewed from the time of the survey. If it is the former, there will be a strong correlation between current emotions and the evaluation of the current situation (or recent life events). A good event may boost one’s satisfaction with one’s recent life even though the effect on one’s satisfaction with the whole life span is more modest. The term ‘subjective well-​being’ is used in psychology for an individual’s evaluation of the extent to which one experiences positive and negative affect, happiness, or satisfaction with life. Reported subjective well-​being is taken as a proxy measure for individual welfare. The measures of subjective well-​being allow for the analysis of the relationship between individual well-​being and economic conditions (such as income, unemployment, inflation, and inequality). However, many economists are very skeptical of the use of subjective well-​being data, because the theoretical status of happiness is unclear in economics. First of all, it is important to distinguish utility and happiness. While utility is a reflection of people’s choices, happiness is a reflection of people’s feelings. Lifetime utility is the standard welfare measure in economics at the individual level. It is often considered as a discounted sum of flow utility over time. As a welfare measure, lifetime utility is based on trusting an individual’s judgments as reflected in choices. People’s choices show that they value a wide range of goods that are not traded, or only partially traded, in markets. On the other hand, psychologists reliably measure happiness in the sense of how people feel at a given time. Measuring happiness as part of a large-​scale survey presents an extra issue in that the survey itself may present a significant slice of a day. Happiness is not the same thing as flow utility. There is evidence that utility has gone up, but happiness has not. According to Easterlin (1995), real per capita gross domestic product (GDP) and real consumption expenditure in the United States has risen dramatically, but the percentage of people saying they are very happy has been falling slightly. Well-​being is a multifaceted experience. It is at least possible to distinguish two of its components: emotions and cognitions. Emotion is considered as a mental feeling or affection (e.g., desire, hope, pain, etc.) as distinct from cognitions or volitions. Positive and negative emotions come in a permanent flow when individuals are awake. For the cognitive component, on the other hand, individuals take some distance to formulate a judgment over their life. One can see subjective well-​being as reflecting a cognitive judgment about the extent to which one is leading a good life. The relevant preferences about life dimensions should

Preferences, utility, and welfare  49

involve proper deliberation, and they are not always revealed in actual behavior. Traditionally, economists assume that it is sufficient to pay attention to decisions. This is because people’s choices should reveal their preferences. In conventional economic theory, consumer preferences are stable over time. However, if behavioral failures lead to biased choices, it is not clear that observed choices should be used to infer preferences. Studies from behavioral economics indicate that individuals are not capable of maximizing their well-​being, because of the structure of decision-​ making situations. In recent years, economists have started to measure well-​being directly, through subjective well-​being indicators. Some researchers have introduced a distinction between ‘experienced utility’ and ‘decision utility’ (Kahneman et al., 1997).While decision utility is the perceived utility on which decisions are based, experienced utility is the real after-​decision utility. That is, decision utility is linked to prospective choices, and, on the other hand, experienced utility is interpreted in terms of the hedonic experience of an outcome. In its original interpretation, utility is interpreted in hedonistic terms, as a measure of pleasure and pain. Experience means hedonic one. Experienced utility describes the pleasure and pain elicited by a chosen stimulus while anticipating, experiencing, or remembering it. It is measured directly. In Jeremy Bentham’s view, the value that a person assigns to pleasure or pain depends on its duration.13 The hedonic experience is the integral of experienced utility function values over a period of time. According to Kahneman et al. (1997), experienced utility is a hedonic quality expressed as a function of time that represents an individual’s experience from a single activity. Activity is the source of hedonic experience, and experienced utility is defined in terms of this activity, not any preference relations between activities. Experienced utility usually refers to outcomes that extend over time. Every instant of an experience that influences the pleasure or pain provides us with ‘instant utility.’ Instant utility is the basic unit of experience used to catalog the moment-​by-​moment pleasure and pain evoked by an experience. The integral of all moments of instant utility provided by an experience is its total utility. People’s past affective experiences often guide their decisions about the future. A first approximation to understanding people’s decisions and choices is purely hedonistic. That is, people seek to repeat in the future what they have liked in the past and avoid further experiences with what they have disliked. However, memory is fallible and introduces systematic biases into evaluations. Memory-​based evaluations of a past event are disproportionally influenced by the event’s ‘peak and end’ experience and insensitive to the event’s duration. Presumably, the peak–​end bias occurs because the experiences are well recalled at the time of evaluation. Economists abandoned experienced utility early in the 20th century, in favor of a new interpretation (a person’s decision utility is revealed by his or her choices). Kahneman et al. (1997) provide a defense of experienced utility with applications to economics. Instant utility is evaluated with self-​report scales both within and between persons. To measure experienced utility is associated with the measurement of happiness/​life satisfaction/​subjective well-​being.

50  Preferences, utility, and welfare

Notes 1 More generally, decision-​makers are assumed to act as if they either maximize or minimize an objective function over some feasible set. 2 For Samuelson (1938, p. 71), his aim was to “develop the theory of consumer’s behaviour freed from any vestigial traces of the utility concept.” That is, the main objective was to make consumer theory totally independent from any psychological assumption and to give a pure behaviorist interpretation to standard concepts in consumer theory such as indifference curves. 3 In situations involving uncertainty, individuals have well-​formed beliefs about how uncertainty will resolve itself.That is, when new information becomes available, individuals update their beliefs using Bayes’ law. 4 Utilitarianism, defined traditionally and most generally, is the doctrine that the morality of an action depends solely on its consequences. The general principle underlying this doctrine is that the act is right (wrong) if the consequences are good (bad). 5  (respectively,  + ,  ++ ) is the set of all (respectively, all nonnegative, all positive) real numbers. The n-​fold Cartesian products of these sets are the Euclidean n-​space n and its nonnegative and positive orthants n+ and n++ . 6 Rawls (1971) argues that rational actors behind a ‘veil of ignorance’ about their individual identities would not choose utilitarianism. Rawls (1971) specifies the conditions of the initial situation and the character of the actors in that situation in more detail. Rawls’ original position is a state of mind in which individuals choose principles of conduct from behind a veil of ignorance. In the original position, one is assumed to lack knowledge of the conditions of one’s life in society. According to Rawls, agents behind the veil of ignorance will choose among principles that determine their life chances on the basis of which will best secure the basic means (primary goods). 7 Gul and Pesendorfer (2007) distinguish three types of welfare analysis: welfare I, II, and III. As a tool for analyzing economic institutions and models, welfare I identifies Pareto-​ efficient outcomes and relates them to equilibrium outcomes. By positing an objective function for a policymaker, welfare II evaluates different policies or institutions according to that objective function. Welfare III corresponds to normative economics. Welfare III asks ‘which of these two options should the policymaker choose?’ while welfare II asks ‘which of these two options might the policymaker choose?’ 8 According to Jeremy Bentham, the widely accepted founding father of British utilitarianism, the societal objective ought to be to achieve the greatest happiness for the greatest number. All experiences can be measured on a single scale of pain and pleasure. They can be compared, and they guide us on what we ought to do and what we shall do. 9 In his famous theorem, Arrow (1950) established that there is no aggregation procedure from individual utility to social welfare that satisfies the Pareto criterion. Arrow’s theorem is a formal statement of the obvious truth that there are unavoidable conflicts of interest or dilemmas in society. 10 Sen distinguishes four different aspects of the relationship between a good (for example, a bike) and an individual: a ‘good’ is the item (the bike); ‘utility’ is the pleasure derived from that item (pleasure from having the bike); ‘characteristics’ are qualities of goods (transport); and ‘functioning’ relates to the individual’s use of the good (moving). 11 If person A is happier than person B, but person B is better nourished than person A, then A scores more highly on one functioning but B scores more highly on another.

Preferences, utility, and welfare  51

12 Thus, psychologists draw a distinction between the well-​being from life as a whole (‘context-​free’) and the well-​being associated with a single area of life (‘contextspecific’). 13 Bentham ([1781] 1988) contended that pain and pleasure are feelings that are experienced on a continuum and that they can be compared. Pain and pleasure guide us on what we ought to do and on what we shall do.


3.1  Taxation Taxation affects the allocation of resources, the distribution of income, and the welfare of individual households. Taxes may be levied on income, wealth, or expenditure; on rents; or on profits. Commodities may be charged at different rates or uniformly; income may be taxed proportionately or progressively. Taxes affect work incentives, the amount of saving, and the level and pattern of investment. The process of taxation may be divided into several distinct phases: design of the tax codes, information collection of taxpayers, the assessment of liabilities, and tax collection. Some taxes distort the allocation of resources and lead to inefficiencies. The level and structure of taxes determine the distribution of after-​tax income among different groups. The main purpose of taxation is to raise resources to finance government expenditure. Should additional revenue be raised by introducing a value-​added tax, increasing income tax rates, or enforcing the existing income tax more effectively? A central issue in public economics is the appropriate design of a tax system. The problem of tax design can be seen as one of finding a way of raising the resources which is administratively and politically feasible and which promotes equality and efficiency as far as possible. Normative arguments involve a balancing of the economic criteria of efficiency and equity. From an efficiency point of view, an ideal tax system is consistent with a Pareto-​ optimal allocation of resources. Pigou (1920) included negative external effects as disturbances in the market price mechanism. Taxes are suggested as environmental policy instruments to adjust or correct distorted price signals. The aim of these instruments is to internalize the costs to those responsible for the external negative effects in order to compensate those negatively affected. Since Pigou (1920), economists have argued that indirect taxes can be used to improve the efficiency of DOI: 10.4324/​9780429344817-3

Economic incentives in public economics  53

the market allocation of resources in the presence of negative externalities. Pigou’s basic insight can be described simply in the modern terminology of private and social cost. If a firm’s production causes pollution, the marginal social cost of production will diverge from the marginal private cost.Then the firm will produce too much. The implied policy is to levy a tax equal to the difference between marginal social and marginal private costs.The internal cost for the firm could be made equal to the cost for the caused damage in monetary terms. Hence, the tax would induce the firm to produce the socially efficient output. What are desirable characteristics of tax systems? Tax systems have been discussed in terms of at least three different criteria:1 (i) the need for taxes to be fair; (ii) the need to minimize administrative costs; and (iii) the need to minimize disincentive effects. First, one could evaluate any tax policy proposal in terms of fairness or justice. Fairness is a key concept in redistribution issues. Even if the economy functions well, it may involve a high degree of inequality in the distribution and wealth. The reduction of excessive inequality is a potential justification for government intervention. Particular requirements of fairness could shed light on the design of the optimal income tax schedule. Second, one could argue that a good tax system minimizes the resource cost involved in assessing, collecting, and paying the taxes. This may be a rather dominant concern for tax administrators although they typically emphasize the costs incurred by the tax collectors and tend to neglect those borne by the taxpayers. Third, the optimal tax system is the one which minimizes the aggregate deadweight loss for any given tax revenue or public expenditure. Thus, a tax system is usually viewed as balancing the various desirable attributes of taxation. Economic analysis is used to combine these criteria into one, deriving the relative weights that should be applied to each criterion. This is done by using the concepts of individual utility and social welfare. Optimal taxation theory is a normative approach based on standard tools of public economics, as applied when the first-​best allocation of resources and distribution of income cannot be achieved. Optimal taxation theory combines economic criteria into a single objective and finds the best tax system (which minimizes the excess burden of taxation) subject to various constraints on tax instruments and information available to the government. Economists have found it very difficult to model the relationship between tax rates and administrative costs. Therefore, administrative costs have been ignored in their analysis. The work on optimal taxation is concerned with tax structures that take account of both the requirements of economic efficiency and the need to be fair between one taxpayer and another. A tax system that is economically efficient may not be considered fair and vice versa; the purpose of the optimal taxation theory is to find the best balance between the two.

54  Economic incentives in public economics

In practice, governments can impose income tax systems more complicated than the linear income tax.The literature on optimal taxation has obtained useful insights into the likely shape of the optimal tax (Mirrlees, 1971; Tuomala, 1990; Atkinson, 1995). In the literature, household incomes come from variable labor supplies, with different households able to generate different amounts of income per unit of labor supplied. Governments use income tax systems with multiple marginal tax rates. In general, nonlinear tax systems allow the marginal tax rate to change continuously with the level of income. A nonlinear income tax constitutes a reasonably realistic description of the income tax instrument that many countries have (or potentially have) at their disposal. Mirrlees (1971) considers the problem of designing an optimal nonlinear income tax, in which the marginal tax can vary as income rises. In Mirrlees (1971), the only disincentive effect of taxation is on the number of hours supplied by each worker. In his model, differences between the wages of different workers are produced by differences in their fixed productivities. It is possible to consider the effect of changing the marginal tax rate over some range without changing the marginal tax rate at any other incomes. People with income within the range will experience an income effect from the higher taxes which would tend to increase their labor supply. On the other hand, they will also experience a substitution effect which tends to reduce their labor supply because of the reduced reward for additional work. The fact that these two effects go in opposite directions means that the effect of an income tax increase on labor supply could be in either direction, depending on which effect is stronger. People with incomes below the range of increase will not be affected by this change. Furthermore, people with incomes above the range will not experience a substitution effect (since their marginal tax rate has not changed). However, they will pay more tax because of the higher rate applied to some of their income, which will have an income effect.2 The optimal tax approach provides a rigorous framework to address key issues of taxation, operating with an explicit objective function and constraints defined by a fully specified economic model. It has focused mostly on social objectives defined in terms of welfarist (typically, utilitarian) social welfare functions, based on interpersonal comparisons of utility. Social welfare is seen as an indicator of the well-​ being of society. However, it is not necessarily seen as simply the sum of individual utilities. The social marginal utility of an individual’s income may reflect various personal characteristics and ethical values embodied in the social welfare function, including fairness requirements potentially. The problem facing the government is that individuals have different income-​ earning ability levels that the tax authorities cannot observe. If the government could observe these ability levels, it could levy non-​distortionary individual-​specific taxes on ability. In fact, as a proxy for taxing ability, the government taxes income. But it must recognize that the income tax gives individuals the incentives to change how hard they work to escape taxation. We consider a general case of nonlinear taxation.There exists a continuum population of individuals. Each individual has the same preference ordering represented by a utility function u = u(c, l) over consumption c and hours worked l, where ux

Economic incentives in public economics  55

> 0 and ul < 0. Individuals differ only in their income-​earning ability or the wage rate, n. Gross income is given by nl.There is a distribution of n on the interval (n , n ) represented by the density function f(n). Let T denote the nonlinear tax schedule set by the government. The ability to choose an arbitrary income tax function T(.) offers the government the opportunity to impose individual-​ specific lump-​ sum taxes. The government cannot observe individuals’ productivities. But the government can observe income and design a nonlinear tax schedule T(nl) on the basis of that. For individual n, c = nl -​T(nl), which is the budget constraint. We can think of individual n as choosing x(n) and l(n) to maximize the utility function u(c, l, n). We define: v(n) = u(c(n), l(n), n). Then, v(n) -​u(c(n), l(n), n) = 0 < v(m) -​u(c(n), l(n), m), i.e., m = n minimizes v(m) -​ u(c(n), l(n), m). Consequently, we obtain the incentive compatibility constraints (or self-​selection constraints): v′(n) = un(c(n), l(n), n). (3.1) In optimal tax theory, the aim of public policy is expressed as maximizing the following social evaluation criterion: n

W = v (n ) f (n ) dn. n


Furthermore, the revenue requirement is set: n

∫ T (nl (n )) f (n ) = A, n


where A is a constant. The objective of the government is assumed to be the maximization of social welfare (3.2), subject to the incentive compatibility constraints (3.1) and the revenue requirement (3.3). A social welfare indicates the value that society places on the welfare of different individuals. It is a weighted sum of the utilities of all individuals in society. Individuals may differ in their earning ability. To choose its tax instruments, the government takes into account all relevant behavioral responses of individuals. Tax bases should be taxed only if doing so raises social welfare. Proportional tax rates reduce administrative costs because they eliminate the necessity of separate measurement of the tax bases. Proportional tax rates also reduce

56  Economic incentives in public economics

the distortions from changes in the relative prices of commodities. Commodity tax rates should be largely proportional. On efficiency grounds, commodity tax rates should be chosen to achieve equal proportional reductions in the (compensated) demands for all commodities.Therefore, commodities with more inelastic demands should be taxed more heavily in order to reduce the excess burden of taxation. On equity grounds, however, if these goods are consumed predominantly by those with lower income, they should be taxed at lower rates. Furthermore, income taxes, on equity grounds, should be higher on those with greater income; on efficiency grounds, marginal tax rates should be lower the more responsive individuals are in their labor decisions. Thus, the optimal tax rules involve balancing these efficiency and equity considerations. The theory of optimal taxation supposes that policy is made by a benevolent one who respects individual preferences as well as some social preference for equality. As long as the government attaches a larger weight to the individuals with a lower welfare, maximization of social welfare will produce a more equal welfare distribution. Many results depend on the particular forms of individual utility function and social welfare function.To practitioners of public finance, the theory of optimal taxation seems highly technical and hence of little policy relevance. Public economists recognize that optimal tax theory has its limitations, but they go on to say that it is nevertheless a powerful tool, and the conclusions of optimal tax theory will inform the way in which we discuss policy. Redistribution is inherently a political matter. The policy instrument for redistribution is a system of taxes and transfers related to pre-​tax income. Redistribution through an income tax usually entails distortions of incentives. However, the resulting efficiency loss has to be weighed against potential improvements in the fairness of the distribution of resources. When resources are redistributed, moving people from one outcome to another, some members of society will be made better off, while others will be worse off. It is even possible for an inefficient outcome to be socially preferable to an efficient one from society’s point of view. Evaluating the trade-​offs between efficient and equitable outcomes requires society’s concern for the poor relative to the rich. The trade-​offs between efficient and equitable outcomes are evaluated by a social welfare function.The social welfare function may take various forms, depending on society’s taste for equity. If a society is indifferent to how well-​being is distributed among its members, the social welfare function is simply the unweighted aggregation of individual well-​being. Alternatively, if a society particularly values improvements to the well-​being of certain members, for example, the disadvantaged, an additive social welfare function will include weights that vary in magnitude across types of people. Whatever the form, through the social welfare function, the economic framework integrates society’s preferences for equity into the government’s objective. Tax systems have to achieve a range of public policy objectives in a complex and changing environment. They exist not only to raise revenue but are also used in support of a range of objectives in different circumstances.The aims of policymakers are often multidimensional, and their priorities can change over time.

Economic incentives in public economics  57

3.2  Sin taxes Governments may use fiscal policies or regulations to influence the prices of products with potential adverse health effects to change their level of consumption. Tax policy can be used to reduce the prevalence of unhealthy behavior. ‘Sin taxes’ are taxes levied on products or services that are perceived as undesirable or harmful. A definite advantage of the sin tax tool is its effectiveness in changing the consumption habits of the population. If the goal of public policy is to internalize the externalities associated with smoking, it is clear that cigarettes should be taxed. Conventional economic models make different predictions for the effect of excise taxes on people’s utility: the increases in tobacco taxes reduce smoking but make individuals worse off. People suffer a loss when a normal good is taxed, but they would experience increased utility when the tax helps to overcome a bad habit.Thus, sin taxes might be advocated to encourage individuals to improve their health. Sin taxes that discourage the consumption of harmful goods have the potential to increase social welfare. Traditionally, sin taxes have been limited to alcohol and tobacco. However, the spectrum of sin taxes has expanded in the past decade. Governments have two interests in the benefits of sin taxes: reducing unfavorable behavior and generating more tax revenues. These revenues may or may not be used for health-​related spending. If negative individual behavior creates a public cost, it must be determined whether sin taxes reduce negative behavior and whether the revenue they generate can be used to improve public health. Obesity is a major health problem.3 Understanding its causes is key to creating effective policy solutions. Decreases in physical activity are considered as a cause of the obesity epidemic. Weight increases can be attributed to insufficient physical activity as to excessive caloric intake. People gain weight when the calories consumed are greater than calories expended. Therefore, obesity is a result of an imbalance between calories consumed and calories expended; the average calories expended has not changed significantly, but the average calories consumed have risen markedly. Obesity is a significant risk factor for noncommunicable diseases—​ which include heart disease, diabetes, high blood pressure, and stroke. The question arises how the increases in body weight, causing harm to people’s health, are to be evaluated. Obesity has been a key component of nutrition policies for many governments over the last decade. Interventions and regulations for obesity aim to change individual behavior around food choices and reduce nutrient intake. They encourage healthier eating by educating consumers on appropriate food and energy intake. Policies designed to reduce obesity have focused on reducing calorie intake by increasing the availability of nutrition information, hoping that such information will lead consumers to make healthier food choices. This implies that people make sub-​optimal food decisions because they lack perfect information. Information provision such as calorie labeling can be viewed as a traditional economic intervention. Such information provision can be justified if food companies have an incentive to conceal information and consumers do not have a large

58  Economic incentives in public economics

enough incentive to unilaterally seek it out. This is the rationale for the various public information campaigns regarding food and nutrition. The prevalence of obesity is not only a major public health concern but is also an economic problem.4 Obesity is in part due to the decisions that people actually make with regard to food or exercise. There has been increasing recognition of the role of sugar-​sweetened beverages, including carbonated sodas, sweetened teas, and energy drinks, in causing obesity. Sugar-​sweetened beverages represent the largest source of added sugar and excess calories in the diet and have been linked to obesity, diabetes, and cardiovascular diseases. Because these calories come in liquid form, they do not make people feel satiated. This excess calorie consumption causes weight gain. Weight gain is thought to independently affect diabetes and cardiovascular disease.The diseases caused by sugar-​sweetened beverages impose several kinds of costs. First, they can reduce income, through missed work hours, lower productivity, and/​ or discrimination in the labor market. Second, they can impose nonfinancial costs—​ early death as the extreme case. Third, they can increase health-​care costs. Sugar-​ sweetened beverages are, on the surface, cheap indulgences. In contrast with prices of fruit and vegetables, the low price of these beverages has fueled their widespread overconsumption along with their mass marketing. People have reacted to the low price of these beverages by consuming more frequently. Even with perfect information about the nutrient content of food and the health consequences of obesity, some fraction of the population will choose to engage in a lifestyle that leads to weight gain because the costs of not doing so are too high. Obesity imposes both externalities in the form of health-​care costs borne by the population and internalities in the form of health consequences to individuals themselves. One common policy recommendation for reducing sugar intake and obesity is a tax on sugar-​sweetened beverages, sometimes called ‘soda taxes.’ If some costs from the adverse health effects are externalized through health insurance, externalities are important in order to rationalize soda taxes. The economic rationale for soda taxes is then built on the principle of externality-​correcting taxes. If a good has harmful effects which are not considered by its consumers, people will consume too much of it in an unregulated market. Therefore, a tax can raise welfare by reducing consumption toward the efficient level at which marginal social cost equals marginal social benefit. The externality reduction accrues to the government’s budget, due to reduced medical expenditures on treatments for conditions such as diabetes and heart disease. However, it is unclear whether individuals are changing unhealthy behavior due to government intervention. Information-​provision policies such as nutrition labels might be presented in a format that people have difficulty in evaluating. Many people have difficulty in understanding numeric nutrition information and/​or lack the motivation to make use of it. Consumers may not only lack health information, but they may also misperceive or misuse the health information they have. It is increasingly clear that policies, guided by a rational view of human behavior, have a relatively small impact on people’s food choices. Conventional economic theory

Economic incentives in public economics  59

suggests that people will pick their most preferred option, regardless of how the options are presented. People are, in reality, highly prone to sticking with the default option even when superior options are available. It points to the importance of how options are presented to people in the first place. Conventional economic models assume that people can easily discard unwanted goods. However, resisting tempting food is not costless in terms of mental exertions. People face some costs to enforcing their desired actions with regard to eating. At the heart of many health problems is the tendency to overemphasize immediate benefits relative to delayed benefits. Decisions are made between the present and the future. Although the present temptation (e.g., chocolate mousse on the dessert tray) is so salient, the future health implications are remote and incremental. People might face self-​control problems when exposed to the temptation of immediate gratification from food and beverage consumption. Self-​control problems arise when preferences are inconsistent across time or context. People consume more calories than what they consider good for themselves when they think about their diet. Obesity provides information about people’s evaluation of their situation after they have decided about their consumption. People think that they would be better off if they would consume less and care more about their future well-​being. If they have time-​inconsistent preferences, they will become heavier and less happy. When consumers have time-​inconsistent preferences, they consume too much of goods whose current consumption causes negative health effects in the future. It is important to know whether and to what extent people face a self-​control problem when tempted by the abundance of food. Market-​based mechanisms to correct the distortion caused by self-​control problems are likely to be ineffective, and consumers may value sin taxes as a commitment device. Using the optimal taxation framework, O’Donoghue and Rabin (2006) examine sin taxes, in which individuals are heterogeneous in self-​control problems. They find that a sin tax can be Pareto improving as it achieves the goals of redistribution and correcting the self-​control problems. Tobacco smoking has adverse health consequences, and public policymakers have attempted to reduce it. Different interventions have been introduced to combat smoking.5 For example, most people would be aware that ‘smoking can be hazardous to your health’ in the words of the warning on packs of cigarettes. The taxes imposed on cigarettes are also a traditional economic intervention. Discussions about tobacco taxes typically focus on reducing tobacco consumption and improving public health. Since a rise in the tax rate increases the price of cigarettes, fewer cigarettes will be consumed. Cigarette consumption can be fairly price sensitive. This is equally consistent with two different models of why people smoke. According to economic theory, habits are a type of addictive behavior. The instantaneous utility from consuming addictive goods such as cigarettes depends on past consumption of the same good in specific ways: U(t) = U(a(t), c(t), S(t)) = v(a(t), S(t)) + u(c(t)),

60  Economic incentives in public economics

where a(t) and c(t) are the levels of consumption of the addictive and ordinary goods at period t, respectively. Furthermore, S(t), the stock of past consumption of c(t) (or the stock of ‘addictive capital’), evolves according to: S(t +1) = (1 -​d) (S(t) + a(t)), where 0 < d < 1 is the depreciation rate of the stock. Under the rational addiction model of Becker and Murphy (1988), agents choose an over-​life consumption path to maximize expected utility. Time consistency is a key assumption of the rational addiction model. Addiction is shown to be rational behavior: consumers optimally make smoking decisions recognizing the addictive properties of cigarettes and all costs which include current and future cigarette prices, health costs, and costs of quitting. The Becker–​Murphy model suggests that potential addiction arises whenever consumption of a good displays adjacent complementarity: someone is addicted to a good only when past consumption of the good raises the marginal utility of present consumption.6 Furthermore, it suggests that adjacent complementarity is closely related to the principle of ‘reinforcement’: greater current consumption raises future consumption. The consumption of addictive goods at different times is complement. An increase in either past or expected future prices decreases current consumption. The price-​theoretic framework yields the prediction that addicts will respond much more to permanent than temporary price changes. Cigarette taxation will reduce smoking in the rational addiction analysis. However, a rise in taxes will make smokers worse off since the price of cigarette becomes more expensive. The alternative to the rational addiction model is one in which smokers demonstrate time-​inconsistent behavior. Behavioral economics emphasizes that habitual behavior can be explained by having time-​ inconsistent preferences. Though incentives are already in place for smokers, cigarettes offer an even stronger case. While the incentives are in the future (the medical aspects of not smoking), some people spend a large portion of their immediate income on cigarettes. At the heart of public health problems is the human tendency to overemphasize immediate benefits relative to delayed benefits. Smokers, in this framework, apply a higher discount rate to decisions closer to the present than to the future. As a result, smokers sometimes make decisions which are not in their best long-​run interest. Time consistency means that the future actions required to maximize the current present value of utility remain optimal in the periods when the actions are to be taken. In a I-​period model, a time-​consistent, exponential discounter makes decisions at period t, according to the discounted utility function: I −t

∑δ (t ) U (t + i ) , i

i =0

where 0 < δ < 1.Time consistency requires that the future should be discounted at a fixed rate, independent of when the costs and benefits of the actions actually occur.

Economic incentives in public economics  61

As the alternative of this type of discounting, a time-​inconsistent, quasi-​hyperbolic discounter makes decisions at t, according to the discounted utility function: U (t ) + β

I −t

∑δ (t ) U (t + i ) , i

i =1

where 0 < β < 1. In this form of time inconsistency, the discount factor between consecutive future periods (δ ) is larger than between the current period and the next one (βδ). People appear to have higher discount rates over payoffs in the near future than in the distant future. Therefore, an individual is impatient when faced with a choice between today and tomorrow, but he or she would become patient in the future.The individual often favors short-​term gains that entail long-​term losses. For a quasi-​hyperbolic discounter, a conflict between the current self and the future one is created, which is the essence of the self-​control problem. Under the behavioral model of Gruber and Köszegi (2001), time-​inconsistent smokers have self-​control problems. Smokers may have taken up smoking with the intent of quitting sometime in the future. However, when the future arrives, they find that quitting is more difficult than they expected. Smokers again lack the self-​control to stick to their long-​term goals. Taxes will also reduce smoking in the behavioral model of time inconsistency. Furthermore, the higher taxes provide a commitment device that help smokers deal with their self-​control problem. Then, the reduction in smoking will make smokers better off. Thus, for a higher tax on cigarettes, there are two models that have opposite welfare implications. In the rational addition model of Becker and Murphy (1988), the welfare of smokers drops as cigarettes become more expensive. On the other hand, in the behavioral model of Gruber and Köszegi (2001), if smokers have self-​ control problems, they always want to quit in the future but not in the present. Then, a cigarette tax can raise the welfare of these types of smokers by providing them with a commitment device that allows them to do something that they would not otherwise choose. The sin tax solution has conflicting arguments for and against its application. Opponents perceive sin tax as a paternalistic solution; it restricts the freedom of consumption of society and the free market. In addition, fiscal policies aimed at raising commodity prices raise public concerns as they may impose an unfair financial burden on low-​income households. Sin taxes are regressive in nature. The policies are seemed to be unjust because low-​income households are forced to bear relatively higher costs related to the introduction of such taxes on some goods than high-​income households. On the other hand, health effects of sin taxes are progressive in nature, therefore health inequalities may decrease. Harmful goods are consumed disproportionately by poor consumers. The biases that cause overconsumption are themselves regressive. Therefore, sin taxes might bring greater benefits on the poor than on the rich. Many smokers report repeated unsuccessful efforts to quit smoking. It is often difficult to wait for a delayed reward when an immediately gratifying alternative

62  Economic incentives in public economics

is available. Although people initially decide to take a far-​sighted course of action (e.g., quitting smoking), they subsequently succumb to temptation. One element common to addictions is the addict’s inability to escape when he or she wants to. The immediacy of reward associated with smoking clearly has some importance in understanding why quitting is difficult. People who become addicted may get greater pleasure from their substances, and they may also discount future rewards more sharply. Smokers attempting to quit smoking may use practices such as rationing of cigarette purchases in order to encourage self-​control. Smoking regulations limit smokers’ choices and are consequently welfare reducing. However, smokers may perceive such restrictions as beneficial. Smokers who have tried to quit in the past but failed and plan to try again would have higher costs of quitting and, consequently, value restrictions more highly than smokers planning to quit for the first time. In recent years, approaches from behavioral economics are used to encourage smoking cessation, healthy eating, energy conservation, retirement savings, charitable giving, and other choices having individual or social benefits.

3.3  Imperfect information and contracts The dissemination of knowledge is crucial in society.According to Hayek (1948), the knowledge that an individual can possess is dispersed, and each individual can just possess a little piece of all the knowledge available in society. Consumer preferences are continually changing, new modes of production are to be discovered, and tacit knowledge particular to the time and place are dispersed among people. Hayek’s (1948) idea about the division of knowledge presents a major challenge to economics. Relevant knowledge is never possessed in its totality by any one person. In economic models, we ought to keep carefully apart what the observing economist knows and what the agents whose actions are under examination are supposed to know. Much of economic importance depends on what agents know. Each agent’s information structure can be represented by an information partition. If an agent cannot distinguish one state from another, those states are grouped into the same information set of one’s partition. Individuals learn the information dispersed in their environments. The economic problem in society is, in fact, a problem of the utilization of knowledge not given to anyone in its totality. Tax administrations deal with information gathering. Preventing tax evasion would be easy if the administration had complete information about all agents. However, agents cannot be completely observed in the public sector. Information varies in quality.The agents who potentially engage in tax evasion are differentiated according to their abilities to avoid detection. The administration does not have full information about the types of the agents. The cost of gathering information depends on how accessible the information is. The field of information economics has been especially influential and insightful in modern economics. Seminal contributions in the field are the informative role of prices in market economies (Stigler, 1961); the prevalence of asymmetric information and adverse selection (Akerlof, 1970); the transmission of private information

Economic incentives in public economics  63

through signaling activity (Spence, 1973); and screening as identifying the qualities of goods, individuals, and other items (Stiglitz, 1975). Asymmetric information is a common feature of market interactions. Sellers of a product often know more about its quality than buyers. Akerlof (1970) develops the concept of asymmetric information with the example case of used-​vehicle market. His model describes how asymmetric information may lead to market failure. The buyer sees the average of the whole market and measures the value of a class of goods, while the seller has more intimate knowledge of a specific item. According to Akerlof (1970), due to this asymmetric information, the seller has an incentive to sell goods of less than the average market quality. If it is difficult for buyers to assess the quality of the product, and if quality is costly to produce, sellers of high-​quality products will not be able to command higher prices for higher quality. As a result, high-​quality products will withdraw from the market. The average quality of goods in the market will then reduce. High-​quality sellers leave the market until only low-​quality goods remain for sale. The process of the bad cars, commonly known as ‘lemons,’ starting to dominate the used-​vehicle market is called adverse selection. Akerlof (1970) shows that, in a market with asymmetric information between buyers and sellers, adverse selection is likely to result. Most cars traded in the used-​ vehicle market become lemons.7 Spence (1973) demonstrates how agents in a market can use signaling to counteract the effects of adverse selection. Signaling, in the asymmetric-​information context, refers to observable actions taken by agents to convince the opposite party of the quality of their products. Spence (1973) develops the idea of signaling, using the job market as an example. The employer is not sure of the productive capabilities of an individual before hiring. Therefore, hiring employees is considered as an investment decision under uncertainty. Potential employees, in Spence’s (1973) model, confront offered wage schedules based on their signals. Signaling costs are the total costs of changing a signal. In Spence (1973), signaling costs are assumed to be negatively correlated with productive capability, which means that, in the job market example, getting a qualification or a degree is easier for more productive people. The employees in the market select the signals they want to transmit in order to choose a suitable wage schedule. A ‘signaling equilibrium’ is generated in Spence’s (1973) model when the employers’ beliefs are confirmed by the signaling generated through the offered wage schedule. Signaling equilibrium is stable if the potential employees in the market differentiate themselves from each other by signaling and thus reduce the information asymmetry between themselves and the employer. Stiglitz (1975) develops the concept of screening as devices to identify the qualities of goods, individuals, and other items. In Stiglitz’s (1975) model, each individual has one characteristic, higher productivity or lower productivity. Stiglitz (1975) introduces a screening process: which productivity can be screened by paying some cost. It is assumed that the cost is more than the difference of wages justified by higher productivity and the mean wage, and that the cost is also less than the difference of the wage justified by higher productivity and that of lower

64  Economic incentives in public economics

productivity. There are two possible equilibria for the model: non-​screening equilibrium and full-​screening equilibrium. In the non-​screening equilibrium, each worker receives a wage proportional to the mean productivity of workers. Since the cost of screening is more than the gain available, the workers of higher productivity do not have an incentive for screening. In the full-​screening equilibrium, on the other hand, all workers receive wages of the low-​productivity group if not screened. Then, the workers of higher productivity have an incentive for screening and will pay the cost of the screening process.The workers of lower productivity do not benefit from the screening process and thus will not pay for it. Economic theories of government regulation begin with the premise that markets sometimes fail. Typically, in public economics, four market failures could justify government intervention in markets—​imperfect competition (or natural monopoly), externalities, public goods, and asymmetric information. Agency theory has its roots in the information economics literature. As such, government regulation is placed into an explicit decision-​making setting. Agency theory models are constructed based on the idea that it is important to examine incentive problems and their resolution in an economic environment where the potential incentive problem actually exists. In the simplest setting, the organization is reduced to the principal and the agent: the principal–​agent model. The principal constructs incentives, while the agent makes decisions on the principal’s behalf. In the principal–​agent model, the efforts undertaken by the agent are not observable by the principal. The standard models of moral hazard analyze the design of incentives to exert the appropriate level of effort by linking the agent’s compensation to his performance. The government does not have all of the information that it needs to achieve its goals. Since the government cannot observe each individual’s ability, it cannot make an individual’s tax payment contingent on their ability type. It can only make it contingent on their income, which can be manipulated by the taxpayer. Some individual characteristics that are relevant for public policy are known only privately to agents. The problem of the government is to choose a policy that best achieves its objectives given the resource and information constraints it faces. A policy is then designed to induce each agent to reveal one’s characteristics truthfully. The disincentive to reveal relevant information can lead to differential information on the part of the state and its citizens. Consider a simple model with two identical workers who differ only in their productivity. Intuitively, high-​productivity workers can earn a given income with less effort than can low-​productivity ones. In this case, the tax system is obviously progressive: high-​productivity types pay positive taxes, while low-​productivity ones receive a subsidy. While the government knows the common utility function, it may only know the underlying distribution of workers’ abilities. Before imposing the tax-​transfer scheme, the government does not know which of the workers is more productive. As the government attempts to redistribute income from better off to worse off using income as the indicator, at some point before income equality is reached, high-​productivity workers will be tempted to mimic low-​productivity ones. The government cannot use differential

Economic incentives in public economics  65

lump-​sum taxes based on ability; high-​productivity workers will have an incentive to be low-​productivity ones to raise their utility.The goal is then to design schemes so that both high-​and low-​productivity workers work appropriately hard. If there is to be public intervention, it should be incentive based rather than command and control. This is a consensus view among public economists. Economic incentives are offers that define a benefit or an extrinsic bonus, which depends on fulfilling an intentionally designed request to alter the status quo. An individual would be led to make choices if there were the incentives to do so. Individuals will perform best when the incentive system links rewards as closely as possible to performance. The theory of incentives studies the design of rules or mechanisms that provide individual agents with economic incentives to adopt an action compatible with overall efficiency or to reveal truthfully private information that is socially relevant. An economic allocation is said to be informationally feasible when the private information necessary to implement it is willingly supplied by the owners of such information. In a mechanism, the planner endows each agent with a message set from which the agent is invited to select a message. The planner, at the same time, commits oneself to an outcome function, a rule which translates the messages of the agents into a feasible social state. Any outcome associated with the equilibrium point of such a game is informationally feasible. If we ask people for their valuation of public goods, they may understate if they have to pay (they may exaggerate if they do not have to pay). Information asymmetry is then essential. The question of how to provide public goods efficiently is a central focus of public economics, largely due to the difficulty in observing individual preferences. The preference revelation problem involves eliciting truthful information from the agents about their types. Incentive compatibility essentially refers to offering the right amount of incentives to induce truth revelation by the agents. The discovery of an incentive-​compatible mechanism for preference revelation for public goods is considered to be an important advance in economic analysis (Myerson, 1982). Truthful revelation of valuations allows the government to implement a Pareto-​ optimal level of the public good. Extrinsic motivation depends on whether economic incentives are used to induce performance. When only extrinsic motivations are in play, higher economic incentives affect targeted actions almost exactly as conventional economic theory predicts. However, conventional economic incentives and social preferences may be either complements or substitutes (Bowles and Hwang, 2008; Bowles, 2016). It is unavoidable that economic incentives adopted by a principal provide some information about the principal’s preferences. Explicit incentives may convey distrust toward the agent. The use of explicit incentives may reveal that the principal would like to profit unfairly at the expense of the agent, which may compromise the agent’s preexisting predispositions of reciprocity toward the principal. Any sort of performance-​based contract is subject to moral hazard problems: a decision taken by the agent after the contract is signed affects the utility of the principal. Moral hazard is present when one party in a transaction takes actions that the other cannot observe. As an organization expands, it will be difficult to correlate

66  Economic incentives in public economics

the work of a particular worker with the value of the organization’s output.Various devices to overcome these problems have been examined, and some of which rely on self-​interested, extrinsic motivation by all parties. However, incentive schemes may rely crucially on intrinsic motivation. Akerlof (1982) argues that employers might pay high wages for workers to induce high effort as part of a ‘gift’ exchange. Workers and employers view part of the exchange between them as an exchange of gift—​a gift of work effort above what can be enforced and a gift of pay above what the market demands. Such an efficiency wage explains the existence of a wage that persists above the market clearing equilibrium as a means of providing workers with an incentive to perform well on the job. Thus, paying a high fixed wage often increases profits because it creates intrinsic motivation for the worker to help the employer. It is profitable for the principal if reciprocal agents respond sufficiently strongly to the principal’s kindness.These lines of analysis imply that an agent might be willing to exert effort for reasons other than explicit extrinsic incentives.Worker performance may be related to workers’ perception of the fairness of distribution. Then, the workers can be induced to exert discretionary effort by paying them a fair wage.

3.4  Tax compliance This section focuses on ‘tax compliance’ problems. Improving tax compliance is a major goal for governments. An individual who is motivated extrinsically expects to receive a benefit or avoid a punishment from an external source, whereas an individual who is motivated intrinsically is prompted to act for reasons of personal morality or internal feelings of satisfaction. Whether increased audits and penalties are the best ways to deal with noncompliance depends on the reasons why taxpayers comply or fail to comply. Economics-​based theories of tax compliance suggest that compliance decision is an economic decision, such that taxpayers weigh economic gains from evasion with possible sanctions from having their evasion detected and identified by the tax authority. Individuals who do not like paying taxes would take a variety of actions to reduce their tax liabilities. Tax evasion consists of illegal and intentional actions taken by individuals and firms to reduce their legally due tax obligations. One possibility is to enforce people to pay their taxes following a deterrence policy. The deterrence hypothesis asserts that people respond significantly to the deterring incentives created by the criminal justice system. Then, increasing the resources which are devoted to the punishment of criminals may be the best policy prescription for reducing the amount of crime. However, prosecuting and punishing offenders can be costly. Therefore, policymakers have to balance these costs against the advantages of reducing crime when making policy decisions. Although higher penalty and audit rates generate benefits, they also entail costs both to the government that must use resources in its efforts and to the individuals who suffer a loss in utility from greater enforcement. From these considerations, the government

Economic incentives in public economics  67

should not expand its enforcement actions to the point where an additional dollar of enforcement costs yields an additional dollar of benefits: the former involves a resource cost to the economy, while the latter is simply a transfer from the private to the public sector. Tax evasion is the deliberate act of breaking the law in order to reduce taxes. In economics-​based theories, the tax evasion decision is a function of detection likelihood, the size of the penalty, and the individual’s degree of risk aversion. Taxpayers are then described as rational utility maximizers who are solely concerned with advancing their private economic interests. The economic theory of tax compliance focuses only on extrinsic motivations related to deterrence, and tax policy predicts high compliance under high audit probabilities or penalties. In contrast, psychology-​based theories of compliance assume that psychological factors are also important to taxpayers. Individuals may comply due to a wide range of intrinsic motivations including moral sentiments, guilt, reciprocity, and social norms. Then, taxpayers may comply even where the risk of audit is low. Some taxpayers’ behaviors may follow the economics-​based theories, while others may follow the psychology-​ based ones, and mixtures are also possible. The standard view of tax compliance in economic theory is that taxes are a ‘burden.’ Most of the literature has relied on the standard economic model of tax compliance developed by Allingham and Sandmo (1972). Based on the economics-​ of-​crime model by Becker (1968), the standard approach of tax evasion uses a simple expected utility maximization problem, in which tax evasion pays off but is risky because of penalties when detected.Tax evasion levels are the optimal decision of risk-​averse individuals who maximize their expected utility. A taxpayer’s decision whether to evade tax is based on the expected utility after considering the probability of audit, the size of fine, tax rates, and income. Thus, tax evasion is part of an optimal portfolio choice: the individual who chooses to evade taxes makes a risky bet that he or she will not be caught and convicted. Tax compliance can be seen as a ‘gamble’ with the risk of punishment (though individuals may not see evasion this way). Even if the individual is risk averse, he or she will accept the gamble when its expected value is sufficiently large. Individuals focus exclusively on the monetary incentives of the evasion gamble, and they pay taxes solely because they fear detection and punishment. We illustrate the expected utility approach by considering a simple version of the standard model. It is supposed that the true tax base is known to the taxpayer but is costly observable by the tax authority. Let y denote exogenous true income. An individual must choose how much of y to report to the tax authority and how much to underreport. The individual pays taxes at a linear tax rate T on R of income that is reported. However, the individual may be audited with a fixed probability of detection p. If audited, all underreported income, y -​R, is discovered. When evasion is detected, the taxpayer is forced to pay the evaded tax plus a penalty. Let f be the proportional penalty assessed on detected evasion. The individual must pay a penalty at the rate f on unreported taxes T (y -​R).

68  Economic incentives in public economics

If underreporting is not caught, income yN is income less taxes paid on reported income: yN = y -​TR. (3.4) If underreporting is caught, income yA is: yA = y(1 –​T) –​f T (y –​R). (3.5) Each dollar of taxable income understatement offers a payoff of T with probability (1 -​p), along with a penalty of fT with probability p.The expected payoff per dollar of evaded income thus becomes (1 –​p)T –​pfT. Then, the tax rate T has no effect on the terms of the tax evasion gamble: as T rises, the reward from a successful understatement of a dollar rises, but the cost of a detected understatement rises proportionately. A risk-​averse individual chooses a report (R), and thus an amount of unreported income y -​R, in order to maximize expected utility: maxR (1 –​p)u(yN) + pu(yA), (3.6) where u(.) is a von Neumann-​Morgenstern utility function which represents the individual’s preferences toward risk. This optimization generates a standard first-​ order condition for an interior solution. The first-​order condition for optimal evasion in (3.6) is as follows: u ' ( yN ) u (yA ) '


fp 1− p


This model predicts that the risk-​averse individual will do some evasion if p(1 + f) < 1.Thus, an increase in either the probability of detection p or the penalty rate f increases reported income. Note that T does not appear in (3.7), other than via an income effect in (3.4) and (3.5). Thus, in the economics-​of-​crime approach, tax compliance has been analyzed as the individual decision-​making between paying and evading taxes. A taxpayer decides how much income to self-​report when facing a probability of detection and a penalty for cheating. Compliance, in this approach, depends on enforcement. Given the structure of the tax system and enforcement process, taxpayers are faced with opportunities to reduce their tax payments. Under low audit probabilities and low penalties, the expected return to evasion is high and the conventional economic model predicts substantial noncompliance. The government can encourage greater tax compliance by increasing the audit and the penalty rates. However, this prediction is in stark contrast with the observation that there is a high level of voluntary compliance in modern tax systems despite low audit rates and fairly modest penalties.Tax compliance is not simply the result of opportunities to evade tax. Taxpayers do not always behave as the rational, self-​interested ones

Economic incentives in public economics  69

portrayed by the neoclassical economic models: the taxpaying decision does not rest solely on the financial decisions. The economics-​of-​crime approach significantly overpredicts noncompliance, given the very low probability of audit. The puzzle of tax compliance is not why people decide to evade their tax obligation but why so many people pay their taxes to begin with. Compliance cannot be explained entirely by purely economic considerations. We have to shift away from a focus on enforcement-​driven tax compliance to an effort to uncover the roots of voluntary tax compliance. The theory of tax compliance might be extended to incorporate ethics. Individuals are often motivated by many other factors that are broadly classified as ethics (morality, altruism, fairness, or the like). It would be impossible to understand an individual’s compliance decisions without considering the ethical dimensions and their implications for behavior.Tax compliance, to some extent, can be attributed to ‘tax morale’ of taxpayers (Luttmer and Singhal, 2014). Individuals may have intrinsic motivations to pay taxes or feel guilt (or shame) for failure to comply. Tax morale is considered as the intrinsic motivation to pay taxes that arises from the moral obligation to pay taxes as a contribution to society. Tax morale affects actual tax compliance behavior. Governments can improve tax morale by increasing voluntary compliance with tax laws and by creating a social norm of compliance. Tax morale is usually perceived as being part of the meta-​preferences of taxpayers, but it is treated as a black box without discussing or even considering how it might arise or how it might be maintained. For example, we can discuss perceived trust in state functioning or public finance. According to Frey (2003), tax morale decreases when taxpayers have little trust in authorities and are treated with no respect. Taxpayers who trust their government and believe that tax revenue is spent properly would be more willing to pay taxes. In contrast, tax morale would decrease as people believe that tax money is spent redundantly and mistrust their government. There are obvious limits to a government’s ability to increase tax compliance via the traditional policies of greater audits and fines. In fact, greater enforcement could reduce compliance if the enforcement crowds out the intrinsic motivation that leads individuals to pay their taxes. An individual will comply as long as he or she believes that compliance is the right thing to do. If others behave according to some socially accepted mode of behavior, then the individual will behave appropriately. The behavior of paying tax follows directly from attitudes, particularly the intrinsic motivation to pay. Tax authorities should pursue policies that reflect a perceived importance of nonpecuniary factors in tax compliance decisions. The conventional economic model misses important aspects of the real-​world reporting environment. Observed compliance levels can be explained by accounting for psychological or cultural aspects of the reporting decision. Different countries have varying degrees of tax morale which influences tax compliance (Alm and Torgler, 2006). These variations are difficult to explain solely in terms of the financial constraints that people face. For example, social norms can affect the strength of intrinsic motivations to pay taxes. A social norm represents a pattern of behavior that is judged in a similar way by others and that is sustained in part by social

70  Economic incentives in public economics

approval or disapproval. The social norms of tax compliance differ across countries, and these differences would affect compliance. When others behave accordingly to some socially acceptable mode, the individual will behave appropriately as well; if others do not behave so, the individual will respond in kind. In order to reduce tax evasion, administrations should persuade taxpayers by emphasizing that tax compliance is practiced by the great majority of people.

3.5  The environment The problems of externalities and the associated market failures have long been discussed in traditional economics. If markets are complete and perfectly competitive, the competitive equilibrium implements an efficient outcome. The only role for government is then redistribution. Market failures create scope for government action to improve the allocation of resources. Public economics has traditionally focused on the idea that market failure is the critical source of economic inefficiency. When markets fail, public economics offers a set of policy tools that can increase social welfare.8 The Pareto criterion is then employed to judge policy interventions. In designing an optimal tax system, we must consider how taxes interact with market imperfections. Optimal taxation theory has characterized the pattern of optimal tax policy and suggested avenues for tax reform. Taxes on activities creating externalities could reduce the economic inefficiency caused by externalities. Optimal tax models aimed to correct externalities suggest that the optimal tax balances the marginal social damage from the externality with the marginal social benefit of the activity that generates the externality. But the optimal tax does not necessarily eliminate the activity that generates the externality. Externalities arise when the actions of one or more agents give rise to uncompensated economic implications for others. From the Pigouvian perspective, the emitter causes the externality. Taxing an externality-​producing activity results in efficient outcomes.9 The externality problem is then solved by making the emitter liable for economic damages and imposing an emissions tax. On the other hand, according to Coase (1960), there is no reason to blame one party as opposed to the other. Externalities are then ‘reciprocal.’ The Coase theorem posits that allocating legal rights to one party or another will not affect outcomes if transaction costs are sufficiently low. With low transaction costs, the parties would be expected to bargain to the efficient outcome under either legal regime. For Coase (1960), efficiency requires determining which party could change behavior most cheaply. Public policy is directed at the satisfaction of individual preferences. However, public policy that tries to correct a market failure may generate a new set of distortions. Observed choices do not necessarily reveal true preferences. It would be difficult for policymakers to distinguish between individuals making mistakes and individuals having unusual preferences. If choices and preferences diverge, policymakers might not be able to make judgments about whether a particular public policy improves social welfare. How is it possible to make judgments about

Economic incentives in public economics  71

individual and social welfare when the information we have about preferences is often inconsistent? Public policy which corrects for market failures may be complicated if individuals are not fully rational. A new category of market failures stems from individuals failing to optimize rationally in their decision-​making. Market-​based tools which depend on shaping economic incentives are premised on agents optimizing in response. In many contexts, standard economic analysis is insufficient to identify the sources of impediments to market efficiency since psychological and cognitive biases of individuals are assumed to be the main driver for ‘behavioral’ market inefficiencies. Bounded rationality can result in actual behavior departing from the neoclassical economic prediction in systematic ways. A traditional market failure (externality) interacts with a behavioral market failure (consumer mistakes in choices). Preferences differ between contexts. Individuals show preference reversals where the relative valuation of items depends on the means by which the value is elicited. Such preferences would obviously complicate the task of environmental valuation. The environment is affected by existing production and consumption patterns. To counter potential environmental problems, behavioral changes are needed. Economists see pollution as the consequence of an absence of prices for certain scarce environmental resources (such as clean air and water). The introduction of surrogate prices has been prescribed in the form of unit taxes. Pigouvian taxes increase the relative cost and the prices of activities and products harmful for the environment. These taxes provide economic incentives to reduce the pressures on the environment. It is well known that, among the array of pollution control instruments, economists prefer those which provide economic incentives through prices rather than through command and control. The main advantage of market-​based instruments (such as emission taxes, subsides on abatement, and tradable permits) is their cost efficiency. Market forces spontaneously work in a cost-​effective way to reduce the quantity of emissions. Pigouvian taxes generate revenue that can be potentially used to promote further environmental protection. They provide a flexible and cost-​effective means for reinforcing the polluter-​pays principle and for reaching environmental policy objectives. The idea that market failure can be reversed through market-​like incentives rests on the presumption of rational behavior—​people facing economic incentives will act with purpose and make consistent choices that take into account the consequences of their choices. However, relying on rational choice theory to guide environmental policy makes sense only if people make consistent and systemic choices. Conventional economics assumes that consumers have stable, consistent, and context-​independent preferences over the consumption options available to them. Thus, it is only through changing prices of goods, not through changing preferences, that people’s choices change. Preference change at the individual level has not played a substantial role in public policy. However, according to Bowles (1998), preferences coevolve with economic institutions, and various mechanisms of cultural transmission give rise to preference changes. The processes

72  Economic incentives in public economics

of preference formation are often not subject to control by the agents themselves. If climate policies shift preferences toward low-​carbon consumption, the costs of mitigating climate change would be lower. Policy measures can have an influence on preferences. As policies that change preferences, awareness-​raising campaigns are considered. These not only provide information but also aim to change people’s habits. Policy induces stronger preferences for low-​carbon consumption goods. People may not be aware that some of their preferences will be formed by the policy choices. Energy efficiency is about providing the same products and services with less energy use. The focus on energy efficiency is motivated by a desire to reduce emissions of carbon dioxide and other pollutants. For example, for market-​based tools like a gas tax to work properly, consumers must respond by making cost-​ justified investments in cars with higher fuel economy. Financial incentives (such as subsidies, tax incentives, or low-​cost loans for the purchase of energy-​efficient durables) are used to encourage consumers to invest in energy-​efficient products. However, we frequently observe that consumers do not choose the most energy-​ efficient appliance, even if this appliance is also the most cost-​efficient choice for them. Many consumers are unwilling to buy pricey fuel-​efficient products even when the reduction in fuel costs would eventually more than cover the increased up-​front cost. Individuals behave as if they undervalue the product attribute of energy efficiency in a variety of choice situations. Privately cost-​effective technologies often go unadopted. The low adoption of energy-​efficient appliances despite apparently large cost savings is closely related to what Jaffe and Stavins (1994) call the ‘energy paradox.’ More broadly, the energy paradox is defined to describe the slower than socially optimal rate of diffusion of energy-​efficient products. Even when energy-​ saving measures are demonstrably cost-​effective, many consumers remain reluctant to introduce these things into their lives and homes. Many consumers are concerned about climate change and understand the importance of saving energy, but this concern does not reliably translate into taking practical steps to reduce household energy consumption. The underlying issues for determining the existence of an energy paradox are whether individual decision-​making is modeled correctly and whether all relevant costs are accounted for. Energy-​related practices are often influenced by certain behavioral biases which are actually predictable from the perspective of behavioral economics. Systematic behavioral biases in consumer decision-​making may explain the energy paradox. Consumers may have heterogeneous information or behavioral biases. Policymakers often use a consumer protection rationale for energy efficiency policy, suggesting that imperfect information and behavioral biases could explain why consumers don’t take up privately profitable energy efficiency investments. In the case of imperfect information, information provision is likely to be the most appropriate policy. However, it is not clear that information provision alone would necessarily be effective. It will not be successful if consumers do not respond to it. Investing in energy efficiency may be

Economic incentives in public economics  73

risky due to the irreversibility of the investment and fluctuating energy prices. If energy prices fall, the return on the investment declines. Therefore, consumers may be better off delaying the investment until the uncertainties are resolved. People would deviate from the predictions of expected utility theory when they make choices involving risk. Optimal policy will depend on the degree of deviation from expected utility theory. Consider reference-​dependent preferences, in which outcomes are evaluated based on gains or losses relative to a reference point rather than on absolute levels. Then, gains and losses may be treated asymmetrically: losses count relatively more than equivalent gains. Individual’s utility from any outcome depends on the outcome’s relationship to a particular reference point. In many cases, individuals exhibit loss aversion, which means that the decline in utility from a relative loss is much larger than the increase in utility from an equivalent relative gain. Losses are then weighted more heavily than equalized gains, particularly as the stakes rise. Prospect theory by Kahneman and Tversky (1979) explains these behaviors. It implies that consumers are acting sub-​optimally in a welfare-​relevant sense. Consumers’ choices over energy efficiency investments involve uncertainty, and they may be explained by prospect theory. For most consumers, energy efficiency is invisible. The performance of energy-​using durable goods could be the greatest source of uncertainty for consumers making choices among them. Even when energy-​using devices are clearly labeled with energy efficiency ratings, there can be considerable uncertainty about the performance an individual will obtain. Generally, the energy efficiency realized in actual use will differ from the rated efficiency in ways that are difficult to predict. When selecting the energy efficiency of a purchased durable good, individuals behave as if they undervalue future energy savings.This downward bias is a consequence of substantial uncertainty about the value of future energy savings. Loss aversion, combined with uncertainty, is thus offered as an explanation for the tendency of markets to undervalue energy efficiency relative to its expected value. People typically focus on the risks, costs, or losses associated with adopting a new behavior and tend to discount equivalent gains and benefits. When faced with making a decision, people perceive the disutility of losing something as far greater than the utility of gaining something. This suggests that the way consumers make decisions about energy efficiency leads to a slower diffusion of energy-​efficient products than would be expected. People who are more loss averse are less likely to undertake energy efficiency investments like buying a fuel-​efficient car or energy-​efficient light bulbs. The term ‘behavioral market failures’ is used to parallel the familiar idea of traditional market failures. Behavioral market failure means that a person fails to behave as predicted by rational choice theory. Public policy addresses both the market failure generated by externalities and the behavioral market failure generated by loss aversion. People’s choices over energy efficiency are explained by prospect theory. Behavioral market failures can affect thinking about public policy. When the behavioral market failure is likely, policies that change the choice setting can change economic behavior.

74  Economic incentives in public economics

Sellers may have better information than buyers but are unable to convey that information credibly to the market, which leads to a market failure from asymmetric information. If consumers are inattentive to cost-​effective efficiency, information provision may be effective. However, if information provision is ineffective, changing the choice setting may be a better approach. Rather than emphasizing the benefits of saving energy, focusing on the costs and losses associated with energy-​ wasting practices may be more effective to improve consumers’ energy efficiency. Thus, loss-​framed messages will have a greater behavioral impact than gain-​framed messages. This may make the information more salient and motivating. ‘Green nudges,’ a part of the well-​known nudge agenda popularized by Thaler and Sunstein (2008), aim at encouraging pro-​environmental behavior. Nudges are designed to correct individuals’ behavior. A nudge has to be a simple intervention. Nudge interventions are cheaper to implement than traditional ones. Nudges are not costly, which requires less information than a tax regarding an implementation point of view. To nudge someone is to deliberately intervene in a given choice architecture, without changing economic incentives or the option set itself. The traditional form of intervention is paternalistic. But a nudge is a particular case of ‘libertarian paternalism,’ in which the term ‘libertarian’ means that the freedom of choices faced by the individual is preserved. Especially, green nudges can be used to encourage people to voluntarily contribute to a public good, namely, environmental protection. Eco-​labels are an important subset of green nudges; providing consumers with additional information necessarily improves their choices. Compared to other standard forms of regulation like Pigouvian taxes, nudges are inexpensive and thus they are easy to implement politically. The tax is paid on the difference between an individual’s contribution (i.e., actions made to reduce the level of pollution) and the socially optimal one. The nudge implementation, on the other hand, does not require the regulator to have knowledge about individual contributions: individuals should react by themselves when receiving the nudge (as long as they contribute less than the announced value). The effectiveness of green nudging is largely due to individuals’ awareness and cognitive efforts in processing environmental consequences of individual behavior. Priming individuals by providing information on appropriate behavior is a common way to promote pro-​ environmental behavior. The environment is often associated with strong moral feelings (e.g., guilt, shame, pride), which may affect people’s beliefs and attitudes toward green consumption and policies. People prefer to see themselves as ‘green’ rather than ‘greedy.’ Nudges can use social influences to affect individual consumption, typically by providing feedback information about self-​consumption and others’ consumption. Reputation effects can be important for the purchase of green products. Buying green products may enable people to signal prosocial behavior to others. Pro-​ environmental behavior can be motivated by encouraging people to cultivate a positive self-​image as pro-​environmental consumers. Social norms are shared perceptions of ideal forms of behavior to which individuals try to conform. Perceiving what others

Economic incentives in public economics  75

do has a powerful influence on individuals’ behavior. Awareness and acceptance of a social norm is likely to modify behavior accordingly. If individual behavior is compared with conservation practices of others, people would be persuaded to reduce emissions and energy consumption. Social norms could provide consumers with information about their energy consumption relative to others along with suggestions for reducing energy use. Social cognitive factors therefore deserve greater attention in environmental policy design.

Notes 1 In his book The Wealth of Nations, published in 1776, Adam Smith laid down four canons or maxims of taxation. (1) Canon of equality: Equality here means justice. Tax payments should be proportional to how much a person benefits from living in society.There should be proportionality across levels of income and sources of income such as rent, profit, and wages. (2) Canon of certainty:Tax liabilities ought to be clear and certain, rather than arbitrary. (3) Canon of convenience: Every tax ought to be levied at a time or in a manner that is most likely convenient for the taxpayer. (4) Canon of economy: Collection costs of taxes should be minimized as far as possible since they add nothing to the national product. If the collection costs of a tax are more than the total revenue yielded by it, it is not worthwhile to levy it. The second and third canons are self-​evidently desirable, and they have not been widely discussed in the economic literature. 2 There are the following three effects of the tax increase on tax payments and welfare: (1) the tax payments of people with the increased marginal rate will probably fall, (2) the tax payments of people with income above the range of increase will rise, and (3) the utility levels of both types affected by the tax increase will fall. 3 The most commonly used criterion to classify obesity is body mass index (BMI): BMI is defined as a person’s weight in kilograms divided by the square of height in meters (kg/​ m2). BMI can be also used to identify individuals at an increased risk of morbidity. Based on the WHO classification, a BMI of 18.5–​25 is considered normal weight, a BMI greater than or equal to 25 overweight, and a BMI greater than or equal to 30 obese. Further, individuals with a BMI greater than 35 are classified as severely obese, and those with a BMI greater than 49 are classified as morbidly obese. 4 Obesity and its economic costs are borne on three levels. First, at an individual level, obesity imposes costs by limiting personal opportunity in many ways. Second, in the workplace, costs are born by employers due to underperformance, absences, and higher insurance premia. Third, obesity affects expenditures by local, state, and national governments, in which programs cover some of the private and workplace costs of illness and unemployment. 5 The graphic warning labels on cigarette packs seem to be designed more to elicit powerful negative emotions. Bans on the display of cigarettes in stores make it more awkward to purchase them and also remove cues that might otherwise induce craving in smokers who are trying to quit. 6 In Becker and Murphy’s (1988) analysis, the existence of multiple stable equilibria is shown, where agents either abstain or consume substantially. 7 As an example of adverse selection in developing countries, Akerlof presents credit markets in India in the 1960s, where local moneylenders charged interest rates that were twice as high as ones in large cities. Without knowing the local borrowers’ creditworthiness, a

76  Economic incentives in public economics

middleman risks attracting those with poor repayment prospects and becomes liable to heavy losses. 8 Conventional economic theory also argues the converse: when markets have not failed, public sector intervention should be limited. 9 Automobiles are a critical part of the pollution problem. Gasoline consumption maps neatly into greenhouse gas emissions. A Pigouvian tax on emissions, feasible in the form of a gasoline tax, can restore market efficiency.


4.1  Bounded rationality 4.1.1  Herbert Simon and bounded rationality Traditional policy analysis fundamentally assumes that consumers’ choices identify their true preferences. However, in practice, consumers’ choices may not maximize their own welfare. It is necessary to extend traditional policy analysis to allow for the possibility of consumer mistakes. Rationality has been given a specific meaning in neoclassical economic theory. In Herbert Simon’s words: “the rational man of economics is a maximizer who will settle for nothing less than the best” (1978, p. 2). While the neoclassical economic approach established a close connection between rationality and utility (or profit) maximization, Simon scrutinized the implications of departures of actual behavior from the neoclassical assumptions.1 Following Simon, the term ‘rational’ is applied not to substantive outcomes, but to the process from which decisions have evolved. ‘Bounded rationality’ is an alternative conception of rationality that incorporates the cognitive processes of decision-​makers more realistically. Simon’s creativity was spread across distinct areas of empirical sciences—​social, behavioral, and computer sciences. In the 1950s, four path-​breaking lines of his work were published. (i) Bounded rationality: In 1957, Simon published a collection of his papers under the title Models of Man: Social and Rational.2 The volume includes his 1955 paper titled “A behavioral model of rational choice” in the Quarterly Journal of Economics. (ii) Cognitive psychology: In 1958, Allen Newell, Clifford Shaw, and Simon published their paper, “Elements of a theory of human problem solving,” in the Psychological Review. DOI: 10.4324/​9780429344817-4

78  Behavioral economics and public policy

The standard theory of problem-​ solving, initially outlined by their work, focuses on how humans respond when confronted with an unfamiliar task. Though Simon was not the progenitor of the “cognitive revolution,” his contributions to it were enormous.3 (iii) Artificial intelligence: Beginning in the mid-​1950s, Newell, Shaw, and Simon’s research on the logic theory of machine, their chess playing program, and the general problem-​ solver played a major role in the early development of artificial intelligence. In 1957, they published the joint paper, “Empirical explorations of the logic theory of machine.” (iv) Behavioral organization theory: An organizational theory must specify how decisions are made in organizations. In 1958, James March and Simon published their book Organizations. Their work linked organization studies and the newly developed behavioral decision theory. In Simon’s (1955) paper “A behavioral model of rational choice,” it is argued that if economists are interested in understanding actual decision behavior, the research would need to focus on the perceptual, cognitive, and learning factors that cause human decision behavior to deviate from that predicted by the normative ‘economic man’ model. Economic man is, according to Simon (1955, p. 99), assumed to have knowledge of the relevant aspects of his environment which, if not absolutely complete, is at least impressively clear and voluminous. He is assumed also to have a well-​organized and stable system of preferences, and a skill in computation that enables him to calculate, for the alternative courses of action that are available to him, which of these will permit him to reach the highest attainable point on his preference scale. Simon (1955) focuses on organisms, including economic man, in complex environments. His original statement on the notion of bounded rationality emphasizes the access to information and the computational capacities that are actually possessed by organisms. Our understanding of economic activity is strongly influenced by the notion of bounded rationality and by direct analogies and tools from biological and computational sciences. For Simon (1985, p. 297), “bounded rationality is not irrationality.” Bounded rationality, taking into account the cognitive limitations of the decision-​maker, is a weakened form of rationality with regard to the maximizing behavior assumed by expected utility theory. An adequate theory of bounded rationality should describe the real processes that individuals use to make actual decisions. A choice is a selection of one, among numerous possible alternatives, to be carried out.‘Global rationality,’ or the rationality of neoclassical economic theory, requires knowledge of all possible alternatives. However, for actual behavior, just a few of these alternatives are considered.The concept of bounded rationality implies that humans are, in practice,

Behavioral economics and public policy  79

incapable of excreting global rationality. Bounded rationality is an alternative conception of rationality that models the cognitive processes of decision-​makers more realistically. Individuals are presumed to attempt to act rationally but to be bounded in their ability to achieve rationally.Therefore, they have to employ alternative cognitive strategies to make decisions. Simon (1976) coins the terms ‘substantive rationality’ and ‘procedural rationality’ to describe the difference between the question of what decision to make and the question of how to make it. Behavior is substantively rational when it is adequate to the realization of given ends, subject to given conditions and constraints. Global rationality is understood as substantive one; it is only concerned with ‘what the decision-​maker chooses.’ The concept of procedural rationality, on the other hand, focuses on ‘what procedures the decision-​maker uses.’ Procedural rationality is “the effectiveness, in light of human cognitive powers and limitations, of the procedures used to choose actions” (Simon, 1978, p. 9). Procedural rationality requires logic of discovery; it can be seen as the process of finding reasonable solutions, given limited information and computational capacities. This interpretation emphasizes the distinction between the real world and the individual’s perception of (or reasoning about) the world. The procedural aspect characterizes the presence of a search process. The choice conditions are therefore the subject of a search process. Each decision-​maker is goal oriented. That is, “the decision-​maker wishes to attain goals and uses his or her mind as well as possible to that end” (Simon, 1997, p. 293). Faced with several strategies, the decision-​maker looks for alternatives that satisfy goals, rather than trying to find the best imaginable solution. If the strategy selected is conductive to the achievement of the given end, the decision-​maker has behaved rationally. However, the decision-​maker often fails to accomplish the intention, because of the interaction between aspects of his or her cognitive architecture and the essential complexity of the environment he or she faces.That is, the decision-​maker must operate under the following two constraints. First, because of cognitive limitations (or limitations of knowledge), it is impossible for the behavior to reach any high degree of rationality. Second, since the decision-​maker lives in an ‘environment of givens,’ his or her behavior is adapted within the limit set by these givens. A wide variety of cognitive tasks require the processing of information distributed across the internal mind and the external environment. It is the interwoven processing of internal and external information that generates a person’s intelligent behavior. Human thought is then subject to improvement. Simon (1959) distinguishes the ‘objective’ information which exists ‘out there’ in the agent’s environment from the information which ‘enters’ his or her faculties and is processed by them. For Simon, people are information-​processing organisms.The information which first enters the individual’s faculties is selected from the total set of objective information. This selection process follows from the individual’s ‘perception’ and ‘attention.’ The information, selected by the individual’s perception and attention, first enters one’s faculties. The information entering one’s faculties is not necessarily representative of the objective information which is out there. The perceived world of the decision-​maker is thus different from the real world.

80  Behavioral economics and public policy

From the set of information which first enters the individual’s faculties, a subset of information will be selected contingent on the problem-​solving strategy which has been chosen.The individual eventually ends up processing the information through both rounds of selection: perception and attention in the first round and choice of problem-​solving path in the second. We must pay attention to the procedures that individuals use to choose their actions. There are problems of too much information in the real world. Human beings must use simplifications in their strategies to deal with complexity. Simon (1957) hypothesizes that economic agents perform limited searches, accepting the first satisfactory decision. The agents do not have complete and accurate knowledge about alternatives. Alternatives are generally not fixed in advance but are generated. Therefore, the agents have to search for information.4 They search until they find alternatives that are satisfactory enough, rather than until the marginal expected cost of search equals the marginal expected return. Rather than assuming that economic agents are globally aware of all the possibilities and comparatively compute them, Simon (1957) emphasizes the search for possibilities and the localness and limits of rationality. For boundedly rational agents, information gathering is costly.The number of possibilities and combinations is so large that search costs are not known in advance. Humans are not knowledgeable about all circumstances; there is always a trade-​off between allocating time and resources to gathering further information and proceeding to act on the basis of current information. Starting from their internal representation of the environment, economic agents cope with a problem by adopting simplifying strategies for its solution. March and Simon (1958) propose an alternative model of rationality with the following simplifying features: (i) Optimizing is replaced by satisficing—​the requirement that satisfactory levels of the criterion variables be attained. (ii) Alternatives of action and consequences of action are discovered sequentially through search processes. (iii) Repertoires of action programs are developed by organizations and individuals, and these serve as alternatives of choice in recurrent situations. (iv) Each specific action program deals with a restricted range of situations or consequences. (v) Each action program can be executed in semi-​independence of the others—​ they are only loosely coupled together. When overloaded with information or complexity, agents are often incapable of systematically processing all available information to maximize utility. Humans then choose not necessarily the best option or solution to a problem, but rather the first available option or solution that suffices. Thus, they replace the goal of optimizing with the goal of ‘satisficing.’ Satisficing is often explained as a procedure of decision-​ making by which the agents search for a solution until they find a satisfactory choice,

Behavioral economics and public policy  81

given the information available and their ability to compute the consequences.5 The search for alternatives ends when a certain criterion is reached (a ‘stopping rule’). According to March and Simon (1958, p. 182), the “criteria of satisfaction is closely related to the psychological notion of ‘aspiration levels.’ ” An empirically grounded theory of bounded rationality is described not in terms of optima, but in terms of values of targets which are satisfactory. In the situations, an agent must choose an option that meets some predetermined aspiration level. Satisficing addresses when an option is good enough in the sense that its payoff exceeds an aspiration level. Determination of an aspiration level is based on experience-​derived expectations of possible consequences. Rejecting an option that does not meet or exceed the aspiration level derives its justification from an observation that the option is rejected in favor of an unknown alternative that produces better consequences. The search for alternatives is compatible with limited computational resources, and it terminates when an option is identified that exceeds the aspiration level. Satisficing is a way “of simplifying the choice problem, to bring it within the powers of human computation” (Simon, 1957, p. 204). The satisficing approach can be described as a decision-​making strategy of the following form: 1. Set an aspiration level such that any option which reaches or surpasses it is good enough. 2. Begin to search and evaluate the options on offer. 3. Choose the first option which is good enough, given the aspiration level. Optimizing is thus replaced by satisficing. People form aspirations, search for alternatives satisficing them, and choose the first option that is good enough. The aspiration levels of goals are not rigid but are revised in an upward or downward direction according to whether performance exceeds or falls short of aspiration. For example, sales in a given period more or less than the target level fixed for that period will lead to higher or smaller targets, respectively, for the next period. An agent is serially evaluating alternatives as to the likelihood of satisfying his or her preferences. Alternatives can be discovered sequentially through search processes. In the search process, an effort is made to satisfy the aspiration level. The agent continues searching for alternatives if no satisfactory result is found yet. The satisficer stops searching for alternatives by choosing the first option to reach or exceed one’s aspiration level. Thus, we can summarize the following four principles about decision-​making. (i) The principle of bounded rationality: The capacity of the human mind for formulating and solving complex problems rationally is bounded. (ii) The principle of satisficing: An individual establishes one’s goal as an aspiration level.

82  Behavioral economics and public policy

(iii) The principle of search: An individual sequentially searches for alternatives and selects one that meets the aspiration level. (iv) The principle of adaptive behavior: An individual continually adjusts one’s behavior to changing environments. The behavioral theory of the firm places an explicit emphasis on the actual process of organizational decision-​making (March and Simon, 1958; Cyert and March, 1963).6 Empirical accounts of behavior in organizations suggest that behavior is often cognitive and calculative. In spite of their best efforts to deal with the complexity and unpredictability of the external world, individuals are limited in their ability to plan for the future and to accurately predict for the various contingencies that may arise. Organizations can be viewed as collections of individuals with multiple goals who operate in a defined structure of authority. In decision-​making, these multiple goals act as constraints or limits on a course of action. Cyert and March (1963) consider the firm as a site of decision-​making involving conflict, uncertainty, search, learning, and adaptation over time. Their theory of the firm is embedded in the form of flowcharts, algorithms, and computer programs. As an alternative conceptualization of the firm, their model emphasizes the interplay between cognition and action or between thinking and experiencing. An evidence-​based theory of the firm is developed in which the firm adapts to its environment through learning and experimentation. The boundedly rational managers think along the lines of simplified representations of their decision problem. The thought processes lead to strategic directions that guide local searches. Suboptimal decision-​making can be interpreted in terms of the manager’s level of information and one’s ability to use it, coupled with a willingness to accept adequate, rather than maximal, profit.

4.1.2  Routine Facing bounded rationality, economic agents simplify the structure of their decisions. Nelson and Winter (1982) develop a sophisticated evolutionary theory of economic change, building on Simon’s view of bounded rationality. Nelson and Winter (1982) focus on the firm as a decision-​making and change-​generating unit. Instead of maximizing the behavior of firms, Nelson and Winter (1982) use the concept of decision rules; instead of equilibrium, they see tendencies.The notion of ‘routine’ is a key concept in the foundation of their evolutionary theory. Routines refer to the regular and predictable aspects of firm behavior. Firms arise to create and coordinate the capabilities needed to exploit new opportunities in changing environments. Their boundaries are determined primarily by the relative costs of developing needed capabilities internally. The concept of routine-​based, boundedly rational decision-​makers pervades Nelson and Winter’s (1982) theory. Nelson and Winter (1982, p. 42) stress “the observed role of simple decision rules as immediate determinants of behavior and operation of satisficing principle in the search process

Behavioral economics and public policy  83

for new rules.” Routines are patterned, typically in the form of sequences of individuals’ actions. The contents are organization specific. Individual firms have a kind of genetic endowment in terms of technical routines.7 Searching for new routines is itself also a routine. Nelson and Winer (1982) distinguish among three classes of routines: (i) routines, called operating characteristics, that govern organizational decisions and behavior given a particular stock of resources; (ii) routines that augment or diminish the stock of resources in response to changes in the state of the organization or the environment; and (iii) search routines, including the organization’s own R&D and its investigation of what other firms are doing, that can modify various aspects of the operating characteristics. Routines are stores of empirical knowledge and, as such, conjectural in nature. Routines refer to formal as well as tacitly understood rules of behavior. When a particular routine is no longer deemed to be satisfactory because of changing market conditions, this triggers a search for a new routine (for example, through increased investment in R&D). Search activity generates the variety which is the basis of evolutionary change. Organizations will present a broad array of routines, some long established and some recently innovated. Drawing from Schumpeterian insights, Nelson and Winter (1982, p. 275) argue that “a central aspect of dynamic competition is that some firms deliberately strive to be leaders in technological innovations, while others attempt to keep up by imitating the success of leaders.” Evolutionary processes involve the generation of novelty, focusing attention on the importance of creativity and innovation. Schumpeter ([1912] 1934) emphasizes the entrepreneur’s role as a major innovator who shifts the economy to new equilibrium states by upsetting the routine operation of the market process (‘creative destruction’ in Schumpeter’s term).The leader-​type entrepreneur often stimulates a cluster of imitating activities by other followers. In Nelson and Winer’s (1982) evolutionary theory, there are three basic concepts: routine, search, and selection environment. “Routines in general play the role of genes in our evolutionary theory. Search routines stochastically generate mutations” (Nelson and Winter, 1982, p. 400). Firms compete for market shares on the basis of their specific routines that they built up and improved on in the past. Some routines will be simple habits, easily changed when better techniques become known, while others will be taken-​for-​g ranted elements that resist change. The nature of realized variations in organizational forms is random. For organizations, a changing environment may lead to the modification of their routines or the replacement of them. Nelson and Winter (1982) examine populations of firms with differing routines by addressing the interplay between changing external environments and changing routines.8 Some organizations whose routines are relatively better fit for coping with environmental conditions will thrive; other organizations will either imitate the more successful routines or decline.Variations in organizational activities may be conditioned by the nature of the routines, but there is no way to know which of the variations will prove better adapted to the environment.

84  Behavioral economics and public policy

4.1.3  Selective rationality and X-​efficiency Leibenstein (1966) proposes the ‘X-​efficiency’ framework underlying the notion that economic agents may not achieve allocative efficiency in their productive decisions and behavior. Although conventional microeconomic theory focuses on the problem of the efficient use of resources, the phenomenon of X-​efficiency has been unfairly neglected and considered as less important. X-​efficiency measures the extent to which the firm fails to realize its productive potential: it is the degree of inefficiency in the use of resources within the firm. The difference between maximal effectiveness of utilization and actual utilization is considered as the degree of X-​inefficiency. Individuals choose how rational they want to be in different situations, which means that they demonstrate a ‘selective rationality’ (degree of maximization deviation). Following Leibenstein (1978, p. 22), “[i]‌ndividuals often make trade-​offs between how they would like to behave without any constraints and how they should behave concerning constraints they are facing.” Individuals are subject to ‘inert areas’ within which their behavior is unresponsive to changes in external constraints. For Leibenstein, the firm is an organization of different individuals who have no consensus on their objectives. An organization can be inefficient because its outputs lie inside the efficiency frontier. Each firm’s managers are not completely in command of the full spectrum of the decision-​making process. The worker’s effort is a discretionary variable and not a constantly given value. Management must cope with each worker’s motivation preferences. X-​efficiency theory is based on the following postulates (Leibenstein, 1987): (i) Relaxing maximizing behavior: It is assumed that some forms of decision-​making, such as habits, conventions, moral imperatives, standard procedures, or emulation, can be, and frequently are, of a non-​maximizing nature. (ii) Inertia: Functional relations are surrounded by inner areas, within which changes in certain values of the independent variables do not result in changes of the dependent variables. (iii) Incomplete contracts: The employment contract is incomplete, in that the payment side is fairly well specified but the effort side remains mostly unspecified. There is a tension between the employee and the employer over how hard the former should work. (iv) Discretion: Every employee has effort discretion and a free-​rider incentive to move to his or her minimum effort level (even though he or she might want others to work effectively); the effort convention is thus a coordinated persistence of nonoptimal conventions, and it helps to explain the existence and persistence of inefficient behavior. There is no need to suppose that individuals actually use inputs as effectively as possible. Individuals supply different amounts of effort, where effort is a

Behavioral economics and public policy  85

multidimensional variable under different circumstances. For Leibenstein (1979), the ‘micro-​ micro’ problem is the study of what goes on inside a ‘black box.’ Leibenstein’s approach aims to explain organizations’ external activities through the study of their internal processes. In conventional economics, effort is assumed to be maximized in its quantity and quality dimensions, irrespective of institutional conditions. According to Leibenstein (1979), the main argument against the maximization postulate is an empirical one—​people frequently do not maximize. Economic agents are not all the same and therefore are characterized by different objective functions. Many firms remain X-​inefficient even when most of their members know the solutions to perform X-​efficiently. X-​inefficiency is the failure of a productive unit to fully utilize the resources it commands (and hence attain the familiar production possibility frontier).The introduction of effort variability allows for the existence of the X-​inefficient economy.9 Leibenstein’s analysis presents a reasonable vision of human behavior within the organizational context, where X-​ inefficiency can persist stubbornly.

4.1.4  Ecological rationality Human thought is adaptive and subject to improvement. Along with the notion of bounded rationality, Simon (1956) points out the importance of considering the environment in which economic agents operate. Simon (1956, p. 130) proposes to integrate environmental factors into the consideration of rationality: “we might hope to discover, by careful examination of some of the fundamental structural characteristics of the environment, some further clues as to the nature of the approximating mechanisms used in decision making.” Human beings are simple systems whose behavior reflects the complexity of the environment they live in. Simon (1969, p. 52) summarizes his approach by using the following metaphor:“[a]‌n ant, viewed as a behaving system, is quite simple. The apparent complexity of its behavior over time is largely a reflection of the complexity of the environment in which it finds itself.” Human behavior may be considered as a function of both cognition and the environment. Newell and Simon (1972) deal with the ‘task environment,’ defined as an environment coupled with a goal, problem, or task. Human behavior is explained by the nature of the task environment. That is, given enough time, human thought takes on the shape of the tasks facing it. They describe the relationship between the external environment and bounded rationality as follows: [j]‌ust as a scissors cannot cut without two blades, a theory of thinking and problem solving cannot predict behavior unless it encompasses both an analysis of the structure of task environments and an analysis of the limits of rational adaptation to task requirements Newell and Simon, 1972, p. 55 If these two blades are not closely matched, then decision-​making will be ineffectual. Human rationality cannot be understood merely by considering the

86  Behavioral economics and public policy

mental mechanisms that underlie human behavior. Instead, we should elucidate the relationship between the mental mechanisms and the environment in which they work. What their metaphor suggests is that we should expand the bounds of bounded rationality outside the head and into the environment.This does not mean replacing the mind with the environment. The mind is an adaptive system which closely fits the environment. An important aspect of any adaptive system is that the system’s behavior is adapted to the requirements of specific tasks. The mind has “responded to the shaping forces of an environment to which it must adapt in order to survive” (Simon, 1990, p. 2). This ‘ecological’ perspective on bounded rationality has guided an important line of research on judgment or decision-​making that is closely allied with Simon’s conception of bounded rationality. Gigerenzer, Todd, and the ABC (Center for Adaptive Behavior and Cognition) Research Group (1999) define ‘ecological rationality’ as the property of a heuristic. Heuristics require less effort than a rational, information-​based, and calculated choice. Heuristics are used to simplify complex tasks. Instead of searching for the universal tool that can solve all tasks, humans take a repertoire of specified heuristics that can solve specific tasks in specific environments. The heuristics individuals use are adapted to the structure of the environment. Ecological rationality claims that the rationality of a particular decision depends on the environment in which it takes place. The ecological rationality perspective is intended to study the mind in its environmental context. Ecological rationality implies a two-​way relationship between simple heuristics and their environment.The success of simple heuristics is enabled by their fit to environmental structure.The more closely a heuristic reflects important aspects of a particular environment, the more likely it is to succeed in the environment. Ecological rationality takes seriously Simon’s idea that the structure of the agent’s real environment has a decisive influence on his or her cognitive architecture. Simple heuristics can succeed by exploiting the structure of information in an environment. Making use of patterns of structured information in the environment is what allows effective heuristics to be simple. The mechanism takes the environmental structure into account in guiding what pieces of information to search for. It does not require seeking all available cues in any order. The framework of ecological rationality is provided to study the adaptivity of decision-​making. This framework assumes that people possess an ‘adaptive toolbox’ (Gigerenzer and Selten, 2001): the heuristics in the adaptive toolbox can yield accurate decisions in the face of limited time, knowledge, and computational power. None of the available strategies is an all-​purpose tool that can be successfully applied to every situation. Decision-​making mechanisms may not perform well outside the contact in which they are involved. To behave adaptively in the face of environmental challenges, agents must be able to make inferences that are fast and frugal. Fast and frugal (which are matched to particular environments) allow individuals to be ecologically rational.10 Sometimes heuristics can be more accurate than complex arithmetical calculus. Different environments can have different heuristics that exploit their particular informational structure to make adaptive decisions.

Behavioral economics and public policy  87

Vernon Smith (2008) introduces two concepts of rationality—​constructivist rationality and ecological rationality.11 Constructivist rationality involves conscious use of knowledge and reason to make decisions to solve specific problems. Neoclassical economic theory characterizes impersonal social systems by competitive equilibrium among fully informed individuals. The design may include many choices that are explainable by deliberate reasoning. However, humans interact in complex ways, and their actions arise in the environment beyond the reach of conscious reason. Human interactions shape the environment, and the environment provides a feedback mechanism that shapes human actions. For Smith (2008), ecological rationality is an evolutionary-​oriented notion of rationality. Ecological rationality, according to Smith (2008, p. 2), refers to emergent order in the form of the practices, norms, and evolving institutional rules governing action by individuals that are part of our cultural and biological heritage and are created by human interactions, but not by conscious human design. Through a process of social evolution, institutional arrangements emerge from human interactions that enable individuals to better coordinate their behavior. This evolutionary process takes place despite their imperfect knowledge of the structure of their environment. When people decide what to do in a given situation, they rely on rules or routines, instead of deliberately and analytically processing all available information. In fact, people follow rules because they know or believe that those rules will produce acceptable actions. Cheap, adequate solutions are often preferred to costly, perfect ones. In Simon’s theory of bounded rationality, simple procedures facilitate decision-​making when the environment is too complex relative to mental and computational capabilities. According to Simon (1993, p. 157), “[i]‌n large measure, we do what we do because we have learned from those who surround us, not from our experience, what is good for us and what is not.” Individuals have a tendency to act on advice and respect social norms. Then, people are socially ‘docile.’ Docility refers to the capacity for being instructed, or the tendency to accept and believe instructions received through social channels. Docile people often make choices under social advice to do so. Social influences can be seen in the way people deal with uncertainty. People prefer to adopt behavioral patterns resembling those adopted in similar situations in the past. Beliefs are formed on the basis of information received from qualified sources. Certain kinds of common understandings among individuals are required to produce the common associations of ideas that allow conventions to emerge and to reproduce themselves. It is the transmission mechanism of socially learned behavior and beliefs that can be received by docile people. In any case, following rules may be a simple and practical way of behaving in a complex social environment. Based on the knowledge of rules used in similar situations, social interaction will lead individuals to behave in a similar manner.

88  Behavioral economics and public policy

4.2  Perception and entrepreneurship 4.2.1 Uncertainty It is recognized that entrepreneurship is fundamentally the driving force of the market. However, there is little consensus about how the entrepreneurial role should be incorporated into economic analysis. In conventional economic models, entrepreneurs are simply people with low risk aversion (Milgrom and Roberts, 1992). Specifically, conventional economic models define entrepreneurship as self-​ employment. Kihlstrom and Laffont (1979) describe entrepreneurship in the occupational choice between employment and self-​employment. In their model, the random utility derived from running a business is governed by a probability distribution with known parameters. Risk-​averse agents remaining in the firm are designated as laborers. Unlike risk-​averse laborers, entrepreneurs bear the risks associated with production and contribute managerial skills. The entrepreneur functions in an equilibrium system incorporating firm formation. Individuals may differ not only in risk aversion but also in their access to information. However, an entrepreneur, in neoclassical economic theory, has access to information required initially to perceive all alternative opportunities and all the possible consequences of acting on an opportunity. In fact, the exercise of judgment should involve private information available only to a few. Publicly available information is not sufficient for taking an important business decision. An entrepreneur takes a decision based on information that is not available to others. Profit opportunities are sometimes known to particular entrepreneurs and not known to others due to information dispersion in markets. An important contribution of Frank Knight (1921), Risk, Uncertainty, and Profit, is the distinction between risk and uncertainty. Whereas risk has a probability distribution and allows insurance, uncertainty does not have a probability distribution. Risk is calculable a priori. Experience can teach us what percentage of bottles is going to burst in a champagne factory; these damages can be included in our cost calculations. Uncertainty is unpredictable and non-​ergodic. It is therefore uninsurable. The major difference between risk and uncertainty is whether it is possible to make ex ante calculations of the incidence of an event. Risk is understood to denote a situation in which known probabilities might be attached to the occurrence of future events. However, if the likelihood of its occurrence is unknown, a future event is uncertain. Knight (1921) regards uncertainty, rather than risk, as the relevant issue for business decisions. When making important decisions, people rarely know either what options are available or their possible consequences. People differ in their ability to select relevant similarities, and the ability is restricted to some particular field of knowledge.Then, “the existence of a problem of knowledge depends on the future being different from the past, while the possibility of the solution of the problem depends on the future being like the past” (Knight, 1921, p. 313). The case of the decision-​maker being uncertain about probabilities of outcomes, referred to as ambiguity, represents a limitation to applying the expected utility framework to individual decision-​making. Under conditions of ambiguity, maximization of

Behavioral economics and public policy  89

expected utility is impossible since this would imply knowing the probabilities of various outcomes. Uncertainty can be linked to a form of intuitive estimation in business situations that resist logical and statistical probability assessment. The expected returns to entrepreneurship tend to be low on average but exhibit a high variance. Entrepreneurs exist in running businesses despite low absolute returns. With a subjective basis for processing, optimistic beliefs may help individuals to overcome uncertainty and may, therefore, influence their entrepreneurial activity. Foss and Klein (2012) argue that entrepreneurial decision-​making is about judgment in the face of (Knightian) uncertainty. Judgment cannot be traded or contracted on; the entrepreneur is required to start a firm in order to take economic actions based on one’s own judgment. Judgment refers to decision-​making when the range of possible future outcomes is generally unknown. Entrepreneurial decision-​making applies to situations in which there is no obvious rule that can be applied. Under conditions of uncertainty, judgment is the (largely tacit) ability to make decisions that turn out to be reasonable or successful ex post (Langlois, 2005). Market participants have heterogeneous entrepreneurial judgments about future events. Uncertainty is thus consistent with subjectivism of expectations.

4.2.2  Schumpeterian entrepreneurship Joseph Schumpeter is, without any doubt, the best-​known contributor to the entrepreneurship field in economics. Schumpeter ([1912] 1934) considers the role of entrepreneurship in the economic development process. According to Schumpeter ([1912] 1934, p. 66), entrepreneurial activities consist of the following five items: (1) the introduction of a new good or of a new quality of a good; (2) the introduction of a new method of production; (3) the opening of a new market; (4) the conquest of a new source of supply of raw materials or half-​manufactured goods; and (5) the carrying out of the new organization of any industry, like the creation of a monopoly position or the breaking up of a monopoly position. For Schumpeter ([1912] 1934), an entrepreneur is a leader and ‘leads’ the means of production into new channels. Entrepreneurship as a form of leadership is to be distinguished sharply from routine. Schumpeter ([1912] 1934) emphasizes the entrepreneur’s role as a major innovator who shifts the economy to new equilibrium states by upsetting the routine operation of the market process (‘creative destruction’ in Schumpeter’s term). Experiential and localized knowledge explains firm heterogeneity and various aspects of competitive advantage, particularly differential innovation performance. The entrepreneur makes a new combination of already existing resources. The number of possible combinations is nearly infinite, and the entrepreneur intuitively selects a few that are possible. Thus, entrepreneurship tends to be an exceptional occurrence of massive importance. The entrepreneur’s leadership qualities enable him or her to see the right way to act. Others will follow in his or her wake. Nelson and Winter (1982, p. 275), drawing from Schumpeterian insights, argue that “a central aspect of dynamic competition is that some firms deliberately strive to be

90  Behavioral economics and public policy

leaders in technological innovations, while others attempt to keep up by imitating the success of the leaders.” The leader-​type entrepreneur often stimulates a cluster of innovating and imitating activity by other following entrepreneurs. Schumpeterian entrepreneurs are described as “leaders that vigorously rise above the masses, personalities that carry the rules of their behavior within themselves” (Schumpeter, [1912] 1934, p. 100). It is individuals who carry out entrepreneurial activities, no matter how they are defined. Their characteristics (personality, background, skills, etc.) matter. Following Leibenstein (1987, p. 120), there are entrepreneurs of unusual talents where modes of operation cannot be captured, described, and taught to others. But in my view, such people are at one end of the talent distribution found among entrepreneurs, which ranges from ordinary capacities, to the ability to slightly modify an existing firm, to the borderline of the great entrepreneurial innovators such as Henry Ford.12 Entrepreneurial activities can be directed toward the discovery of opportunities to earn profit. These activities are led by a group of people with a particular knowledge.

4.2.3  Kirznerian entrepreneurship By contrast to Schumpeterian heroic innovator-​entrepreneur aspect, Kirznerian entrepreneurship focuses on the ordinary profit-​ seeking endeavors of market participants to discover and exploit opportunities for improvement.13 That is, “[f]‌or Schumpeter the essence of entrepreneurship is the ability to break away from routine, to destroy existing structures, to move the system away from the even, circular flow of equilibrium” (Kirzner, 1973, p. 127). Schumpeterian entrepreneurship emphasizes the ability to create new combinations beyond the current production possibility frontier. On the other hand, Kirznerian entrepreneurship involves two phenomena—​the presence of profit opportunities and individuals who pursue those opportunities. Kirznerian entrepreneur is “routine-​resisting” (Kirzner, 1997, p. 71) only at the stage of perceiving profit opportunities. Schumpeterian and Kirznerian entrepreneurs don’t know the extent of their ignorance. Creation and discovery cannot be anticipated. However, Kirznerian discovery view of entrepreneurship can be distinguished from Schumpeterian creation perspective. Kirznerian entrepreneur reacts to exogenously generated changes in the marketplace. Kirznerian entrepreneur might be described as a follower, in Schumpeterian sense, of the leading, originating innovators. Abstracting from the social dimension of entrepreneurial behavior (which is a feature of the Schumpeterian entrepreneur), Kirzner focuses on the pure entrepreneurial function. Kirzner’s focus on the entrepreneur is inspired by the objective of enabling us to see the ‘inside’ workings of the capitalist system, while Schumpeter is concerned to enable us to see, ‘from the outside,’ what constitutes the essence of capitalism (Kirzner, 1999).

Behavioral economics and public policy  91

For Kirzner, the role of the entrepreneur is to achieve some kind of adjustment necessary to move markets toward the equilibrium states.14 In the equilibrium states, economic agents have exploited all the opportunities known to them. Kirznerian entrepreneurs become deliberate coordinators of resources in markets exhibiting disequilibria. Disequilibrium arises from the everyday mistakes market participants make in their investment, production, and distribution decisions. In markets, knowledge is neither complete nor perfect. Markets are constantly in disequilibrium states; for instance, the entrepreneurs could sell for high prices that which they can buy for low prices. Disequilibrium creates market niches, which becomes evident as underpriced products, unused capacity, unmet customer needs, and so on. Thus, it is disequilibrium that gives scope to the entrepreneurial function. For Kirzner (1981, p. 55), “the innovative role assigned by Schumpeter to his entrepreneur finds its place naturally within the broader Misesian theory.” There are three essential aspects of the broader Misesian theory of entrepreneurship, according to Kirzner (1981): (1) the recognition of the entrepreneurial element in each individual market participant; (2) the insight into entrepreneurship as the driving force behind the equilibrating tendencies within the price system; and (3) appreciation for the entrepreneurial basis for the social efficiency achieved by the market economy. Then, action is the present choice between future alternatives that must, in the face of the foggy uncertainty of the future, now be identified in the very act of choice. It is this aspect of human action that renders it, for Mises, essentially entrepreneurial. Kirzner, 1992, p. 128 What is missing in Schumpeterian account is an appreciation of the fact that the ordinary everyday profit-​motivated endeavors of market participants to discover and exploit opportunities for improvement keep the market process working. Kirzner emphasizes the quality of perception for recognizing a profit opportunity.The crucial role of entrepreneurs lies in “their alertness to hitherto unnoticed opportunities” (Kirzner, 1973, p. 39). Kirzner’s concept of ‘alertness’ is a singular property of individuals.15 Alertness is “the ability to notice without search opportunities that have hitherto been overlooked” (Kirzner, 1979, p. 48). An opportunity exists only if it is perceived by the entrepreneur. The entrepreneur has to overcome mental and behavioral habits and become liberated from the dictation of routine in order to exploit opportunities successfully. As Kirzner suggests, “[t]‌he successful businessman-​entrepreneur ‘sees’ what other market participants have not yet seen” (Kirzner, 1992, p. 129, original emphasis). People possess different information, which allows them to see particular opportunities that others cannot see. Entrepreneurs perceive external events and formulate business plans according to their knowledge. Knowledge shapes the interpretation of the external environment. Experiences from everyday life are accumulated into a stock of knowledge that can be used to interpret incoming events. The interpretation framework helps an entrepreneur to identify problems.

92  Behavioral economics and public policy

The discovery process for Kirznerian entrepreneur can be described as follows: (1) Opportunities contain knowledge beyond the control of the entrepreneur. (2) Opportunities are discovered by the entrepreneur on the basis of informational advantages and through other complementary resources. (3) Discovered opportunities contain subjective judgment. (4) The entrepreneurial act is carried out under uncertainty. There are market opportunities that are not being exploited simply because they have not been noticed. Disequilibrium exists because economic agents are ignorant and are often ignorant of their ignorance. Lack of awareness regarding the opportunities renders discovery a crucial phenomenon. One can be alert by exposing oneself to potential market opportunities. Discovery takes place as entrepreneurs seek to seize the opportunities afforded by market frictions which are typically not known in advance. The differential awareness becomes more crucial when the opportunity is less straightforward. Some individuals, but not others, recognize and exploit entrepreneurial opportunities. Individuals with more experience see a specific opportunity as more desirable than it is for others, making it more likely that they will exploit it. Competition is a discovery procedure fueled by Kirzneian entrepreneurship relative to hitherto ignored market opportunities. The entrepreneur enjoys the first mover’s reward. Thus, the function of Kirznerian entrepreneur is the act of opportunity recognition in market disequilibrium. Ignorance is understood in relation to knowledge. According to Kirzner (1979), ignorance is associated with two types of knowledge: deliberated acquisition and nondeliberated acquisition. The former can be gained and learnt by deliberated search, while the latter can only be spontaneously absorbed from everyday life experiences. Experiences from everyday life are accumulated into a stock of knowledge. This sort of knowledge is largely “the result of learning experiences that occurred entirely without having been planned nor are they deliberately searched for” (Kirzner, 1979, p. 142). According to Kirzner (1997, p. 71), “[a]‌n opportunity for pure profit cannot, by its nature, be the object of systematic search. Systematic search can be undertaken for a piece of missing information, …”. Thus, Kirzner (1997) distinguishes discovery from search as follows: What distinguishes discovery (relevant to hitherto unknown profit opportunities) from successful search (relevant to the deliberate production of information which one knew one had lacked) is that the former (unlike the latter) involves that surprise which accompanies the realization that one had overlooked something in fact readily available. p. 72, original emphasis It is impossible for individuals to search for something that they do not know about. This implies that the searchers must already possess a sense of direction to

Behavioral economics and public policy  93

guide them. To the extent that the knowledge structures that were developed in previous organizations shape the perceptions, the entrepreneurs tend to repeat their learned behaviors.This myopic learning process increases the pressures that eventually become codified as institutional practices and may lead to industrial homogeneity. One source of heterogeneity could emerge when entrepreneurs arrive from outside the industry. These entrepreneurs lack the experiences of incumbents and, therefore, may act differently than entrants who have had substantial experiences in the industry. Searching already involves some knowledge of what one is searching for in the industry. Thus, ignorance can be classified into two categories: lack of information and lack of awareness. For instance, the former is attributable to scarce knowledge in an organization. Then, one can be informed about the missing information through searching. Search is a bounded attempt to find new information related to a specific set of criteria. The entrepreneurs search beyond their own experiences. On the other hand, the latter corresponds to unnoticed opportunities in the market. Then, one does not know what one does not know (i.e., a total lack of knowledge). If organizations devote resources to actively searching for unnoticed opportunities, any search will be difficult due to the total lack of knowledge. Information about profit opportunities may be absent in advance, but it is revealed by alertness. According to Kirzner (1997), without knowing what to look for, the entrepreneur is ready to make discoveries. In such a state of ‘sheer ignorance,’ discovery includes some element of surprise. Surprise is the signal to sentient individuals that they have been inadvertently ignorant. If people are surprised by something, it is almost certainly the case that they did not know it (or that they thought they knew something that has suddenly turned out to be false). Alertness presupposes that discovery cannot be anticipated. For Kirzner (1979), opportunities exist in the minds of entrepreneurs and “just around the corner” to be discovered by an alert entrepreneur.

4.3  Errors and biases Since the early 1970s, two psychologists, Daniel Kahneman and Amos Tversky, published a series of articles testing the conventional economic theory of choice under uncertainty.Their main contributions can be divided into three areas: heuristics and biases, framing effects, and prospect theory. They deviate from what would be predicted by conventional economic theory. These deviations are neither random nor trivial but represent systematic patterns of cognitive biases. People rely on simplifying strategies, or heuristics, to solve statistical problems. Heuristics describe how people make judgments and decisions based on approximate ‘rules of thumb’ or mental shortcuts. Heuristics simplify the assessment of complex probabilistic hypotheses and require less effort compared to a rational, calculated choice. Indeed, heuristics often work through a process of ‘attribute substitution,’ in which people answer a hard question by substituting an easier one. People, for instance, resolve a question of probability not by investigating statistics,

94  Behavioral economics and public policy

but by asking whether a relevant incident comes easily to mind. While heuristics are frequently useful shortcuts, they also lead people into biased estimates and predictions. The presence of an error of judgment is demonstrated by comparing intuitive inferences and probability judgments to the rules of statistics and the laws of probability: In making predictions and judgements under uncertainty, people do not appear to follow the calculus of chance or the statistical theory of prediction. Instead, they rely on a limited number of heuristics which sometimes yield reasonable judgements and sometimes lead to severe and systematic errors. Kahneman and Tversky, 1973, p. 237 According to Thaler (1991, p. 4), Kahneman and Tversky have shown that “mental illusions should be considered the rule rather than the exception.” People are often insensitive to sample size. According to Tversky and Kahneman (1971), people ignore sample size and apply a ‘law of small numbers’ to make predictions in situations where the ‘law of large numbers’ is appropriate. An individual’s erroneous behavior is the result of false beliefs, for which the individual cannot really be blamed. People have much more faith in small samples than they should. Then, the “deviations of subjective from objective probability seem reliable, systematic, and difficult to eliminate” (Kahneman and Tversky, 1972, p. 431). Anomalies arise from the way humans process information to form beliefs.Tversky and Kahneman (1974) describe three heuristics that are employed to assess probabilities and to predict values. First, representativeness is the tendency for people to assess the likelihood of an event based on the similarity of that occurrence to their stereotypes of similar occurrences. For example, people ascribe characteristics to groups or subgroups based on their experiences with, or perceptions of, members of a group. When the experiences with members of a group are not representative of that group, the person might incorrectly ascribe the characteristic to the entire group. Second, availability is the tendency to assess the frequency or likely causes of an event by the degree to which instances or occurrences of that event are readily ‘available’ in memory. That is, if people can readily think of examples of events, they will inflate their probability estimates of the likelihood of their occurrence. An example is that the subjective probability of traffic accidents rises temporarily when one sees a car overturned by the side of the road. With regard to self-​control, lower perceived availability of past successes hinders self-​control. Finally, anchoring and adjustment refer to the tendency to assess quantities by starting from an initial value and adjusting to yield a final decision. People are overly influenced by anchors, even arbitrary ones. They make estimates by anchoring on an arbitrary reference value in the decision context, such as a phone number, social security number, or other random number, and indicate higher willingness to pay when those random numbers are higher. In an experimental study (Kahneman and Tversky, 1974), people being asked to estimate the number of African countries in the United Nations were given an arbitrary number between 0 and 100 (starting point) before

Behavioral economics and public policy  95

their evaluation. Their estimate was reported as being biased toward the assigned arbitrary starting point. According to Tversky and Kahneman (1983, p. 293), perhaps the simplest and the most basic qualitative law of probability is the conjunction rule: the probability of a conjunction, P(A&B), cannot exceed the probabilities of its constituents, P(A) and P(B), because the extension (or the possibility set) of the conjunction is included in the extension of its constituents. To test whether decision-​makers abide by the conjunction rule, they asked subjects to rank the likelihoods of certain conclusions that can be drawn from hypothetical personality sketches of fictitious individuals. In one version of the experiment, subjects were given the following personality sketch and asked to identify which of the two alternatives was more probable: Linda is 31 years old, single, outspoken, and very bright. She majored in philosophy. As a student, she was deeply concerned with issues of discrimination and social justice and also participated in antinuclear demonstrations. (1) Linda is a bank teller. (2) Linda is a bank teller and is active in the feminist movement. In this well-​known Linda problem, Tversky and Kahneman (1983) report that 85% of respondents indicated (2) more likely than (1). A majority of respondents (85%) declared the conjunction (bank teller and feminist) to be more probable than its constituent (bank teller), thereby violating the conjunction rule. If respondents engage their System 2, they can get the right answer. System 1 rather easily tells a story for (2); the description of Linda brings to mind a picture that she is active in the feminist movement. Fast-​thinking respondents see the feminist bank teller as a more likely story than merely a bank teller. Tversky and Kahneman (1983) conclude that the conjunction fallacy (assigning higher probability to the conjunction than its constituents) is prevalent in situations where likelihood judgments are mediated by intuitive heuristics such as representativeness and availability. Being a nonfeminist is relatively more associated with being a bank teller than being a feminist. People then compare the likelihoods not of bank teller versus feminist bank teller, but rather of the representative nonfeminist bank teller versus feminist bank teller. Thus, Linda is more likely to be the latter.

4.4  Gains and losses A framing effect is said to be present when different ways of describing the same choice problem change the choices that people make, even though the underlying information and choice options remain essentially the same. It is seen as a violation of the axiom of rational choice known as ‘descriptive invariance.’ People’s

96  Behavioral economics and public policy

preferences are affected by whether the decision problems are framed in terms of gains or losses (Tversky and Kahneman, 1981). Conventional economics assumes that people place the same value on gains and losses of equal amount. However, people respond differently to gains versus losses when they make choices.16 They tend to value losses more than gains. An important factor in economic decisions is ‘loss aversion.’ For example, in a manufacturing company that awards incentive bonuses based on work teams’ output, Hossain and List (2012) find that how one frames an incentive payment affects how much people respond to it. In their natural field experiment, some teams are randomly assigned to a gain frame and others to a loss frame. Gain-​ framed teams are then told that they will earn a bonus for every week in which they meet a performance benchmark, up to a maximum annual payment. On the other hand, loss-​framed teams are told that they will receive the maximum payment minus a deduction for every week in which they do not meet the benchmark. In the experiment, teams presented with the loss framing produce greater output, on average, than teams presented with the gain framing. Tversky and Kahneman (1981) discuss framing effects involving the different responses to gains and losses. People have a heightened tendency to focus on avoiding losses. Such impulsive avoidance may be triggered by automatic fear responses in the brain. Whether choice is framed in terms of gains or in terms of losses influences people’s decisions in ways that cannot be accounted for by conventional economic theory.Tversky and Kahneman (1981) illustrate the impact of presentation on framing by showing that simple wording changes (e.g., from describing outcomes in terms of lives saved to describing them in terms of lives lost) can lead to different preferences. In the experiment conducted by them, subjects are given the following choices: Imagine that the United States is preparing for the outbreak of an unusual Asian disease, which is expected to kill 600 people. Two alternative programs to combat the disease have been proposed. Assume that the exact scientific estimates of the consequences are as follows: Positive frame: • •

If Program A is adopted, 200 people will be saved. If Program B is adopted, there is a one-​third probability that 600 people will be saved and a two-​thirds probability that nobody will be saved.

Negative frame: • •

If Program C is adopted, 400 people will die. If Program D is adopted, there is a one-​third probability that nobody will die and a two-​thirds probability that 600 people will die.

The expected values of Programs A, B, C, and D are the same: 200 lives saved and 400 lives lost. Despite this, subjects receiving the positive frame solidly prefer

Behavioral economics and public policy  97

Program A (72%), and those receiving the negative frame strongly prefer Program D (78%). Thus, by restating the consequences of the alternative programs in terms of potential losses (‘will die’) rather than the potential gains (‘will be saved’), people’s preferences are reversed even though the choices are identical. Conventional economic theory assumes that consumer’s preferences are invariant with respect to his or her current endowment. However, there is diverse evidence of ‘reference-​dependence’: preferences depend sharply on an individual’s ‘reference point,’ which is usually equal to his or her current endowment. People evaluate identical facts differently, depending on whether they are presented in a positive or negative manner in terms of a reference point. A reference point is a base from which changes are assessed. It determines whether something is entered as gains or losses from an endowment or status quo point of view. This implies that values are attached to changes rather than final states. Prospect theory, developed by Kahneman and Tversky (1979), asserts that people evaluate risky choices not on the utility of the outcomes of these choices but rather on the basis of the gains and losses associated with the outcomes. Prospect theory accounts for the framing effect through a ‘value function,’ where values are calculated with regard to a reference point. The value function presents three main characteristics: (i) people implicitly evaluate outcomes in terms of gains and losses; (ii) people are more sensitive to variations between outcomes the closer they are to a reference point; and (iii) people experience gains and losses with different levels of intensity. Prospect theory suggests that people choose options as if they were evaluating outcomes with reference to a zero point, rather than in terms of total wealth. That is, people make decisions based on changes in wealth rather than total wealth. Prospect theory is based on experimental results that do not confirm the expected utility theory. Individuals do not evaluate choices on the utility of the outcomes of these choices. Rather, they frame and evaluate risky choices narrowly in terms of their gains and losses relative to a reference point. Prospect theory uses psychological information to attain a more accurate representation of preferences: losses tend to trigger negative emotional reactions. People have a heightened tendency to focus on avoiding losses, even if it means engaging in risky behavior to do so. For example, most people prefer the certainty of winning $500 over a 50-​50 chance of winning $1,000 or nothing, which demonstrates that they are risk averse. However, most people prefer a 50-​50 chance of losing $1,000 to the certainty of losing $500, which demonstrates that they are risk seeking. The implications of prospect theory are much different from standard consumer theory: people are risk averse with regard to the prospect of gains, but risk seeking in the face of large losses. People’s assessments of values are suggested to be based on an S-​shaped subjective value function of gains and losses. Its S-​shaped value function represents changes (i.e., gains and losses), in the objective value levels on the horizontal axis, and the subjective value resulting from these changes on the vertical axis. The same outcome may be evaluated differently by the same agent depending on whether it is perceived as a gain or a loss with respect to the reference point. The agent is risk averse in the domain of gains, and he or she is risk seeking in the domain of losses.

98  Behavioral economics and public policy

The function for gains is concave, and the function for losses is convex. There is an asymmetry in the value function under prospect theory. The value function is steeper for losses than it is for gains. The ‘kink’ in the shape of the function reflects the asymmetry in the evaluation of gains and losses. A loss would be felt more stronger than a gain of an equivalent amount.

4.5  Nudges Public economics typically lists four market failures that may justify government intervention in markets—​imperfect competition (or natural monopoly), externalities, public goods, and asymmetric information. Conventional economics and governmental economic policy have long been overwhelmingly based on a view of the human as a rational agent who engages in a search for information on the best available option, knows and considers all the costs and benefits, and follows one’s true preferences. However, people usually display bounded rationality, in the sense that they depart from the standard account in predictable and systematic ways. Recent research in behavioral economics highlights a fifth possible market failure—​individual may make mistakes in pursuing their own well-​being. Such a failure means that an individual fails to behave as predicted by rational choice theory. A ‘behavioral’ market failure is caused by an anomaly, bias, heuristic, misperception, fallacy, or illusion in economic decision-​making. As a result of behavioral biases, people fail to satisfy well-​defined preferences. Public policy is directed at better preference satisfaction. Behavioral economics is fundamentally concerned with the questions of how people actually behave in decision-​making situations and how their choices can be improved so that their welfare is enhanced. Governments are increasing making use of behavioral insights. One goal of behaviorally informed consumer policy is to design policy tools that ultimately guide consumers to make better decisions for themselves. People are prone to error in certain areas. They frequently fail to make welfare-​ enhancing decisions. A policy is called to be ‘paternalistic’ only if it makes people behave differently than they would otherwise have done. Successful paternalist policies help people to avoid the poor reasoning that leads them astray from the satisfaction of their own goals. Paternalism is generally construed as altering preferences. If a paternalistic policy increases the cost for participating in an activity one initially preferred, one may still want to engage in that activity, but one’s all-​ things-​considered preference changes and one will choose to do something else. Paternalistic interventions are designed with the primary aim to enhance the well-​ being of their target agents. However, such interference sometimes arouses strong opposition.The government forces individuals to do what it thinks is good for them, rather than what they themselves want. Paternalistic interventions are implemented without the explicit consent of these target agents. For an intervention to qualify as paternalistic, it is not required that its target agents actively oppose it. Many people feel that paternalism sometimes interferes with personal liberty. Liberty consists of having certain options, but coercive paternalism deprives people of options. The

Behavioral economics and public policy  99

objection is that coercive paternalist policies force people to do things that they do not want to do. People know best what they want. They lose the enjoyment of their preferred choice. The government fails to respect the individual’s values; it substitutes its own ends for those that the individual has chosen to pursue. John Stuart Mill ([1869] 1991, p. 91) argues against paternalism: “[a]‌ll errors which he is likely to commit against advice and warning, are far outweighed by the evil of allowing others to constrain him to what they deem his good.” Thus, according to Mill, the harm of interfering with an individual is far greater than having the individual commit errors. An individual knows best about one’s own interests, while no benevolent planner can know better. For Mill, the only freedom is that of pursuing our ‘own’ good in our ‘own’ way. That people own what they do is important. Self-​direction is what contributes to individual well-​being. Liberty makes available the practices of thinking and acting through which people develop the ability to discern what is best. For Mill, to own our preferences is to have an identity. This is how people come to have an individual identity. Our preferences are open to amendment in the light of experience. People have to feel responsible for their actions if those actions are to be a source of learning. The introduction of so-​called libertarian paternalism has become very popular for implementing behavioral public policies without compromising people’s freedom and autonomy. Libertarian paternalists help individuals to make welfare-​ enhancing choices without restricting the range of options available to them. People are free to continue with their existing behaviors if they wish (remaining freedom of choice is the best safeguard against any misguided policy interventions). At first glance, the aim to enhance individuals’ well-​being as judged by themselves, without restricting the range of options available to them, may seem an attractive policy ideal. Individuals should be allowed as much freedom as possible in their decision-​ making, but the government should establish conditions that lead to ex post good decisions. Libertarian paternalism is a program that employs ‘nudges’ as tools for influencing people without using economic incentives or coercion. A nudge is a specific type of behavioral policy intervention that alters people’s behavior in a predictable way but, in contrast to traditional behavioral intervention, it operates by modifying the ‘choice architecture’ of a decision situation, while preserving the same set of decision options and keeping unchanged the costs associated with a decision (Thaler and Sunstein, 2008). Choices themselves are not affected; people are free to make the same (inferior) choices they would have made without the nudge. Thus, “[i]‌f no coercion is involved, we think that some types of paternalism should be acceptable to even the most ardent libertarian” (Thaler and Sunstein, 2003, p. 175). Thaler and Sunstein think that governments can operate smarter by nudging people toward valued outcomes, rather than by forcing them to comply through the promulgation and enforcement of laws. As strategic use of default rules, nudges are increasingly being proposed and used as a policy tool by government agencies around the world. A nudge seeks to bring about beneficial ends by exploring or preventing systematic biases. Systematic biases lead to poor choices. Policymakers are asked to design

100  Behavioral economics and public policy

policies in light of these cognitive shortcomings. Choice architecture reflects the fact that what is chosen often depends on how the choice is presented. A nudge has an aspect of a policy intervention on the choice architecture; it is intended by the policymaker to steer the chooser’s behavior away from the behavior implied by the cognitive shortcoming and toward his or her ultimate goal or preferences. Nudges are adopted to help people make better choices in a large variety of areas of human life: retirement, investment, prescription drug coverage, social security, environment, organ donation, school choice, and even marriage. Different choice architectures affect how people choose. A prominent example used to describe nudging is the cafeteria redesign program: placing healthy food up front and at eye level in the cafeteria, while placing snacks and less healthy food alternatives behind the counter. People can choose between healthy food and unhealthy one.The ‘pro-​ self ’ (i.e., focusing on enhancing private welfare) nudge intervention in the cafeteria redesign example aims to make people eat healthy food. By setting a default rule, a choice architect determines the option that an individual will receive if he or she foregoes making a choice of his or her own. Collections of default settings determine the way individuals initially encounter products, services, or policies. Defaults are choices that apply to individuals who do not take active steps to change them. Default options can have a powerful impact on behavior without necessarily restricting choice. Default rules may well be the most effective nudges. Automatic pension enrollment is an example of the use of defaults in public policy, where an opt-​out default has been seen to significantly improve participation (Madrian and Shea, 2001).17 A substantial increase in money transfer to retirement savings is set as the default rule. Automatic enrollment in retirement plans can increase people’s savings significantly. People are more likely to make choices about retirement savings in line with their long-​term interests. Why do default rules work? According to Thaler and Sunstein (2008, p. 83): many people will take whatever option requires the least effort, or the path of least resistance … All these forces imply that if, for a given choice, there is a default option—​an option that will obtain if the chooser does nothing—​then we can expect a large number of people to end up with that option, whether or not it is good for them. The design of nudges largely relies on results from behavioral economics about the use of heuristics. The nudge approach assumes “somewhat mindless, passive decision makers” (Thaler and Sunstein, 2008, p. 37). Nudges are considered as having a social connection with the System 1 ‘intuitive thinking’ of dual-​process models. Consider the following illustration of framing as a policy intervention. A government aiming to encourage energy conservation informs the public in two different ways: (i) If you use energy conservation methods, you will save $350 per year. (ii) If you do not use energy conservation methods, you will lose $350 per year.

Behavioral economics and public policy  101

Thaler and Sunstein (2008) suggest that campaigns employing a ‘loss’ frame are much more effective than those using a ‘gain’ frame. Framing is a powerful nudge, because people tend to be somewhat mindless, passive decision-​makers. Nudges influence real-​world human beings. Every choice occurs in contexts; it would be impossible for a choice environment to be completely neutral about the way that alternatives are presented and arranged. Nudges tend to involve relatively small additions or changes to the decision environments that encourage, but do not force, changes in behavior. Usually, nudges are not financial mechanisms (they may involve changes in the way in which financial mechanisms are presented or constructed). Nudges are not disruptive to existing programs and preserve citizen choice. Nudges have been praised as soft regulatory tools that, while steering people in a particular direction, also preserve their freedom of choice. For some applications, the goal of nudging is uncontroversial. For example, suppose that an intervention is motivated by the fact that smoking causes negative externalities in third parties. Then, smokers may appreciate (or at least do not oppose) soft measures that help them smoke less, like a less prominent display of tobacco products in stores. The core idea in libertarian paternalism is that the context or environment people face can be redesigned such that their automatic decisions better align with their deliberative preferences. Nudges are supposed to work in a situation with inconsistent preferences subject to endogenous change. People’s preferences are taken as incomplete, inconsistent, and variable. Nudges may not only induce preference changes but may contribute to creating preferences where none existed before. The choice environment is rarely neutral, and choice architects will always be shaping decisions, whether people like it or not. Despite the attempt to preserve option-​freedom, one of the most prominent arguments against nudging is that it disrespects ‘autonomy’ (Grüne-​Yanoff, 2012; White, 2013). Autonomous choice is the cornerstone of democratic liberal societies. Autonomous agents are self-​ governing. They have the ability to discern and consider options and the capacities to act according to these preferences. The influence of nudges may not be easy to resist, even for such agents, especially if nudges are more effective. For example, default rules which preserve freedom to choose are often regarded as prototypical instruments of libertarian paternalism.When consumers are aware that defaults may be set as recommendations in some cases, or manipulation attempts in other cases, they successfully retain autonomy and freedom of choice. However, if default rules have an effect because consumers are not aware that they have choices, or because the transaction costs of changing from the default option are too high, defaults impinge on liberty. Making certain choice options more or less prominent, from this perspective, may prevent individuals from choosing freely among the available alternatives. However, nudges may incorporate values beyond individual autonomy (e.g., solidarity) when judging the suitability as a public policy instrument. Growing attention has been paid to the massive impact of the coronavirus disease 2019 (COVID-​19) pandemic on life safety and the global community. The COVID-​ 19 pandemic has largely restricted population mobility around the

102  Behavioral economics and public policy

world. Restaurants and many other establishments have closed their doors to the public, and demand for home deliveries has risen substantially. Social distancing has become a part of everyday life, and medical masks have become significant medical protective products. Handwashing with soap and water has received considerable attention as a simple, primary preventive measure that most people can do independently. Traditional handwashing promotion interventions have focused on messaging around the health-​related risks associated with germs. It is challenging for public health officials to create messages that are effective in motivating the behavior associated with preventing the spread of the virus. We need to find an efficient way to bring about a behavioral change toward better hygiene practices. However, behavioral change interventions which target everyday activities must overcome habits. Habits are hard to break through improved education and knowledge. Recent research has demonstrated success in handwashing promotion using a nudge-​based intervention rather than health-​based messages (Dreibelbis et al., 2016). In the Bangladesh example, Dreibelbis et al. (2016) painted footprints on footpaths guiding primary school students back from the latrine to the handwashing station, which created a disruptive cue to interrupt the returning to class and remind them to use the handwashing station. The primary goal for interventions is to attract attention, make the behavior convenient, and reinforce it as a social norm. Nudges will be effective if they can achieve a behavioral change toward increased handwashing at critical times. Nudges target automatic processes than may be more effective in promoting behavioral change than conscious reflection and knowledge-​based decision-​making.

4.6  Addiction and self-​control Some people drink excessively, others smoke, and still others use illicit drugs. Their decisions need to be improved. Common tendencies are failing to grapple with self-​control. Simply not assuming that people are making choices that are in their own best interests opens up a world of alternative policy interventions beyond just manipulating prices via taxes and subsidies. Excessive drinking remains a substantial public health problem and requires prevention approaches that are grounded on theoretical and empirical research. Behavioral economics may provide a novel analytical framework for generating prevention strategies with research on choice, decision-​making, and self-​ administration. It predicts that the choice dynamic is critically dependent on the temporal context of the decisions. As discussed later, substance abusers, including heavy drinkers, tend to devalue delayed outcomes; this myopia may be an important target for prevention programs.Though the value of rewards decreases as the receipt is delayed, there are substantial individual differences in the degree that delayed rewards are discounted. This discounting phenomenon may be a core feature of substance abuse. Heavy drinkers discount the value of delayed rewards more steeply than light drinkers.

Behavioral economics and public policy  103

Policymakers need to be aware that how options and information are presented matters to consumer comprehension. Consumers often lack or have difficulty in comprehending information in the face of a myriad of critical choices. If consumers are either poorly informed or misinformed about the effects of addictive goods, they may make bad decisions. If the government can provide relevant information effectively and efficiently, informational or educational policies are potentially beneficial.The use of graphic warning labels on cigarette packages can be seen as one educational intervention in an attempt to combat a smoking habit. Smokers have some knowledge of major health risks, but they have little knowledge about the likelihoods of smoking-​related disease. Most tobacco users overestimate the likelihood that lung cancer is curable. Graphic warning labels on cigarette packs provide affective meaning to smoking-​related diseases which are otherwise potentially abstract to smokers. According to Evans et al. (2015), placing graphic warnings on cigarette packages (i) increased negative affect to smoking, (ii) heightened scrutiny of smoking’s risks, (iii) improved the perceived credibility of the warnings, and (iv) increased label memory. Graphic warnings helped smokers to understand the risks of their habit better and encouraged them to quit. However, informational policy alone is limited because it cannot address the causes of temptation. The basic problem is self-​control, particularly when the present temptation is so salient, whereas the future health implication is remote and incremental. Decisions are made between the present and the future. It is natural to favor the present over the future. Consumers may want to adhere to a healthy lifestyle but frequently succumb to the temptation for immediate gratification. For example, many smokers say they want to quit, often try to quit, but continuously fail in their quit attempts. One element common to addictions is the addict’s inability to escape when he or she wants to. The immediacy of reward associated with smoking clearly has some importance in understanding why quitting can be difficult. The rewards from smoking are immediate, and the adverse consequences tend to be delayed. People who become addicted may get greater, or longer, pleasure from their substances, and they may also discount future rewards more sharply. Self-​ control can be defined as efforts made by the individual to avoid or resist behaving inconsistently, where one’s short-​run actions are perceived as sub-​optimal from one’s long-​run perspective. People regularly fail to meet their goals. One solution is to offer such people commitment devices. Broadly, a commitment device is an arrangement entered into by an individual who restricts his or her future choice set by making certain choices more expensive. People facing greater temptation impose higher costs on their own potential failure. A commitment contract refers to a commitment device that is an actual contract between two parties. As an example of a commitment contract, self-​funded deposit commitment contracts can be used as a low-​cost method of improving the overall performance of a health-​incentive program. Such commitment contracts allow users to put money at stake that is lost if the person does not meet a goal. In the program for smokers in the Philippines to deposit

104  Behavioral economics and public policy

funds into a savings account, Giné et al. (2010) show that such a contract can be an effective tool for smoking cessation. After opening the account, smokers were required to take a urine test for smoking cessation. In their experiment, if the smoker quit (which would be verified by the urine test), the money would be returned; if the smoker failed to quit, the money would be donated to charity. Those who were offered the commitment contract were roughly one third more likely to have quit smoking one year later than the control group. Exponential discounting yields time-​consistent preferences, which means that individuals act according to their long-​run interest when making decisions over time. Individuals are then assumed to be rational about their life-​cycle work and consumption patterns. They save and consume in a manner designed to maintain their preferred levels of consumption through their lives. However, empirical evidence has been collected to cast doubt on the hypothesis of constant discounting. Individuals regularly show time-​ inconsistent preferences in decision-​ making characterized by discount rates varying over time. This violates the principle of exponential discounting. Instead, hyperbolic discounting yields time-​inconsistent preferences, which means that people employ short-​ run discount rates which are higher than long-​run ones. Many people report that they would like to save more than they do. People want to save more, but not until next year, and when next year arrives they again want to save more, but not until next year, and so on. People make short-​sighted decisions when benefits are immediate. Such choices being dominated by immediate benefits are interpreted as a lack of self-​control or present-​biased preferences. Some of the manifestations of the time-​inconsistent preferences are inability to lose weight, stop smoking, and save enough for retirement. If individuals have time-​varying discount rates, their preferences change at some point in the future. An individual may prefer $110 in 31 days over $100 in 30 days, but he or she may prefer $100 now over $110 tomorrow. This is inconsistent with exponential discounting used by conventional economics: if individuals were to discount future utilities exponentially, time preferences would not reverse. Discounting is central in intertemporal choice, but an important consideration is what discounting model to employ. The conventional discounted utility model was introduced by Samuelson (1937). At any point in time, an individual has well-​ defined preferences representable by a utility function. Individuals are assumed to maximize the present value of a stream of current and future (intertemporally separable) utility. In Samuelson’s (1937) model, present value is calculated by discounting future payoffs by a constant amount in each time period. Thus, individuals are assumed to select consumption levels in each time period, c(t), to maximize: U ( c(0 ),…, c( I )) =


∑δ u(c(t )), t

t =0

subject to an income/​ wealth constraint. The intertemporal preferences from the perspective of period 0 is represented by U ( c(0 ),…, c( I )). Choices are time

Behavioral economics and public policy  105

consistent, that is, individuals will make the same utility trade-​off between two periods regardless of when they make the allocation. The discount function from the perspective of period 0 for utility experienced at period t is δt . The exponential discount function implies that an individual’s discount factor between two periods depends only on the length of the delay interval. The individuals in Samuelson’s (1937) model do exactly what they plan to do. Their state-​contingent plans are always perfectly aligned with their subsequent state-​contingent actions. An alternative framework based on Strotz (1955) suggests that individuals may exhibit systematic biases in their decision-​making. Labeling the phenomenon as ‘myopia,’ Strotz (1955) demonstrates that individuals exhibit a higher level of impatience in decisions involving short-​term trade-​offs than they do for decisions made over long-​term trade-​offs. The approach is hyperbolic discounting, i.e., the discount rate is not constant but declining over time. Hyperbolic discounting lacks the characteristics of stationarity. As a consequence, time-​inconsistent behavior can occur. Time inconsistency implies that an ex ante optimal decision is not carried out, because a later reevaluation suggests that it is not optimal anymore. Laibson (1997) assumes that individuals maximize a discounted utility stream that places disproportionately higher weight on the present payoffs relative to all future ones.The ‘quasi-​hyperbolic’ discounted utility function takes the following form: U ( c(0 ),…, c( I )) = u ( c(0 )) + β


∑δ u (c(t )) , t

t =1

where the parameter β ( < 1) corresponds to a time-​inconsistent preference for the current payoff, and δ ( < 1) is time-​consistent (long-​run) component of temporal preferences. If β = 1, these preferences are simply exponential discounting. But β < 1 implies that an individual has self-​control problems. With β < 1, the present is given special status relative to all other time periods. Thus, β models an individual’s preference for immediate gratification. A preference for immediate gratification implies time-​inconsistent preferences. To the extent that individuals are myopic, they will make risky choices inconsistent with long-​term utility maximization. The individual’s behavior can no longer be characterized by solving a single optimization problem. Policy interventions are potentially welfare improving. A possible solution to the problem of time inconsistency is a mandatory commitment. For example, if many individuals choose sub-​optimally low levels of saving, the introduction of mandatory savings may be welfare enhancing. Regarding obesity, an individual who would like to eat healthfully will attempt to build commitment mechanisms to overcome their anticipated short-​run desires to eat unhealthy foods. An individual’s self-​reputation may be used as a commitment device (Bénabou and Tirole, 2004). If an individual pledges to go on a diet, the decision to cheat is met with both benefits (the pleasure from consumption) and costs (the perceived loss of reputation). According to Bénabou and Tirole (2004), these reputation effects will be undermined if memories are opportunistically edited to alter the original

106  Behavioral economics and public policy

pledge or the portion of reputation devoted to fulfilling the pledge. Some decision-​ makers may lack completely or partially self-​understanding. To foresee the need for commitment, decision-​makers have to be aware of their self-​control problems with experience.

Notes 1 First, neoclassical economics presupposes that an agent has a ‘well-​defined’ utility (or profit) function. Second, it supposes that all alternative strategies are known to the decision-​maker. Third, all the consequences that follow on each of these strategies are assumed to be determined with certainty. Finally, the comparative evaluation of these sets of consequences is assumed to be driven by a universal desire to maximize expected utility (or profit). 2 In the preface (p. vii), Simon explains the choice of this title: his essays “are concerned with laying foundations for a science of man that will comfortably accommodate his dual nature as a social and as a rational animal.” 3 The cognitive revolution sought to undermine the dominance of behavioralism in psychology, focusing on internal psychological processes. 4 Simon (1969) explains scientific discovery by his information-​ processing approach. The scientists under investigation share one or more of the processes that are heuristic-​ driven search mechanisms. An important question is why one agent’s use of some shared processes would produce a scientific discovery, while another one’s use of it would not. Simon (1969) claims that the environment is the essential factor in producing different results with the same processes.That is, the environment is instrumental in proceeding the particular scientific discoveries. 5 In a search-​theoretic choice experiment, Caplin et al. (2011) find that many decisions can be understood using Simon’s satisficing model: most subjects search sequentially and stop the search when an environmentally determined level of reservation utility has been realized. 6 Along with Herbert Simon, Richard Cyert and James March are regarded as pioneers of behavioral economic associated with the Carnegie School. 7 Biological analogies for the evolution of organizational forms are conceptually attractive. In a biological context, the term ‘Lamarckism’ implies that genotypical inheritance may occur as a consequence of acquired characteristics. Nelson and Winter define their position as Lamarckian; “our theory is unabashedly Lamarckian; it contemplates both the ‘inheritance’ of acquired characters and the timely appearance of variation under the stimulus of adversity” (1982, p. 11, original emphasis). 8 For Nelson and Winter (1982), routines as well as being replicators are interactors, given that they act as both behavioral dispositions and actual behaviors. However, as Knudsen (2002) points out, this results in a definitional problem from an evolutionary perspective. Routines provide the necessary stability in behavior over time. If routines were the analogue to genes, we would expect them to be what codes for functional behavior and not to be defined in terms of the behavior itself. 9 Altman (2000) introduces effort discretion into path-​dependent modeling. Given effort discretion, incentives need not exist for agents to adopt superior economic regimes. The introduction of effort variability allows for the existence of sub-​optimal equilibrium in the long run. The existence of high and low productivity regimes might be a product of history.

Behavioral economics and public policy  107

10 According to Gigerenzer et al. (2008), research in the fast and frugal heuristics program focuses on three interrelated question: (i) What heuristics do people use to make decisions, and when do they rely on which heuristics from the toolbox? (ii) To what environmental structure is a given heuristic adapted, that is, in what situations does it perform well, say, by being able to yield accurate, fast, and effortless decisions? (iii) How can the study of people’s repertoires of heuristics and their fit to environmental structure aid decision-​making in the applied world? 11 For Smith (2008), constructivist rationality and ecological rationality are not opposed. Two concepts of rationality work together through an incessant back-​ and-​ forth movement. Constructivism projects real-​world rationality on a continuum of logical reasoning and statistical inference, and its results are used as a benchmark to judge actual behavior in the world or in the laboratory. One can use constructivist rationality to make sense of an observation. Once an observation is made, one can ask whether it corresponds to a constructivist model of rationality. 12 According to Baumol (1990), entrepreneurial talent is assumed to be reasonably and equally distributed across time and societies. Innovative entrepreneurship is incremental in nature. 13 According to de Jong and Marsili’s (2015) empirical study, Schumpeterian opportunities are correlated with an innovative behavior, a strategic emphasis on future needs. In contrast, Kirznerian opportunities are associated with a strategic focus on present needs. They are more frequently pursued by small ventures and by young entrepreneurs. 14 Kirzner defines the equilibrium state in neoclassical economics as “a state in which each decision correctly anticipates all other decisions” (1979, p. 110). 15 For Lachmann, Kirzner’s term ‘alertness’ is considered as a behavioral prerequisite for capital formation which is founded on continuous “alertness to change and a willingness to make frequent changes, the switching ‘on’ and ‘off ’ of various output streams as well as reshuffling capital combinations” (1976, p. 148, original emphasis). 16 Different responses to gains and losses also lead to the endowment effect—​an inclination to value more highly and pay more for an item which is already in one’s possession than items which one does not yet own. When a person comes into possession of something, he or she feels ownership toward it, thereby overvaluing it. That is, he or she expresses a preference for a particular reference point, the endowment. This contradicts standard economic theory arguing that preference is invariant with respect to the current endowment. Dividing students into three groups, Kahneman et al. (1990) present the following evidence: The first group was given a choice between a mug worth $4.95 and a chocolate bar worth $6.00, and 56% chose the mug over the chocolate bar.The second group first got the mug and was then given the opportunity to trade it for the chocolate bar; 89% chose to keep their mug.The third group was first given the chocolate bar and then the opportunity to trade it for a mug; 90% chose to keep their chocolate. In experiments, students have a strong tendency to overvalue whichever gift they were given initially, feeling ownership for it and, therefore, being reluctant to trade. 17 Furthermore, powerful effects of defaults on behavior have also been observed in organ donation decisions (Johnson and Goldstein, 2003).


5.1  Social incentives and social preferences Suppose that you are in charge of a local government and must confront a range of social, environmental, and health problems. Rates of recycling are waning; residents are evading local taxes; and obesity and smoking are up. The most obvious way to get people to do something is to pay them for doing so. You design and implement your incentive programs.You pay rebates for recycling, provide tax credits for prompt tax payment, and make small payments to families who lose weight or stop smoking. However, contrary to your expectations, your incentives actually increase tax noncompliance, weight gain, and smoking, while decreasing recycling. How do your incentives have counterintuitive and counterproductive effects on human behavior? Policymakers are interested in how to stimulate voluntary action through appropriate incentives. A factor that might increase the performance of prosocial activities (e.g., volunteering, civic duty, charitable donations, or other social contributions) concerns people’s incentives. People might simply not find it worthwhile to engage in prosocial activities if the benefits fall short of the opportunity costs. Policy failures are partly driven by neglecting behavioral aspects when applying policy interventions. Traditionally, policy interventions have tended to focus on providing ‘economic incentives’ that change the consequences of behavior, seeking to change the way people think about their behavior. These interventions draw on the assumption that people change behavior accordingly when extrinsic motivations are changed. Motivation denotes a reason for an individual choice. Here, motivations can be distinguished into two types: extrinsic and intrinsic. Extrinsic motivation comes from ‘outside’ the person in question, while intrinsic motivation comes from ‘within’ the person. Specifically, extrinsic motivation is the desire for a reward or DOI: 10.4324/​9780429344817-5

Social preferences and moral economy  109

avoidance of a penalty, while intrinsic motivation is the inherent satisfaction of a task. Economic approach to human behavior, due to the application of the relative price effect, is based on extrinsic motivation. Economic incentives are material rewards or penalties intended to shape individual behavior by supplying an extrinsic motivation. An extrinsically motivated reward can be monetary (cash payment) or nonmonetary (e.g., gift). It is designed to motivate a specific action and is offered before an action occurs. Economic incentives are framed in such a way as to induce self-​interest as a response. Once a monetary payment is introduced, the activity is no longer performed when payment is withdrawn.1 Economic incentives alter the costs and benefits of available options. In conventional economics, it has been widely accepted that desirable behavior can be promoted by making economic incentives contingent on performance. The promise of economic incentives is that individuals are rationally motivated and are more likely to pursue an activity if it is externally rewarded. Therefore, economic incentives, as external rewards, for donating blood should raise the perceived value of the exchange and increase the likelihood of a person donating. Recently, there is increasing recognition that individuals are not solely concerned with extrinsic rewards.2 Economic incentives are not always an ideal motivator. Behavioral economics provides a wide array of sources for motivation by taking into consideration intrinsic motivation based on the gratification derived from taking part in a project with no external rewards. People behave as if they were intrinsically motivated rather than stimulated by any economic incentives as suggested by conventional economic theory. Intrinsic motivation is an important source of employee and organizational performance. In fact, the ideal intrinsic incentive lies in the work itself. In this case, intrinsic motivation is based on personal gratification gained from carrying out a specific task. Many employees are affected not only by their pay but also by their perceptions of how the company treats them. Therefore, features like workplace flexibility, worker involvement in running the company, and procedural justice in promotions could be important for increasing workers’ nonpecuniary motivations. To understand what kind of incentives might encourage prosocial behavior, we must first have an understanding of the motivation behind such behavior. People may undertake certain prosocial actions guided by an intrinsic motivation. Under certain conditions, external rewards can inhibit intrinsic motivation. The expose to extrinsic monetary incentives crowds out intrinsic motivation and thus reduces subsequent interest in the task. Motivational crowding out occurs when introducing an economic incentive for a task provokes a loss of intrinsic motivation. That is, performance of intrinsically motivated tasks is harmed by pay for performance. In fact, offering extrinsic rewards to encourage prosocial behavior can sometimes backfire and decrease the desired behavior by crowding out intrinsic motivation to act altruistically. Blood donation has often been seen as an example of altruism with intrinsic motivations. Donating blood is a prosocial act in the sense that donors incur individual costs in exchange for a collective benefit and contribute by ensuring the blood supply system works well. People may respond

110  Social preferences and moral economy

negatively to extrinsic rewards, particularly if intrinsic motivation for the target behavior is already high. ‘Social incentives’ may be used to induce desirable actions. People occasionally undertake generous acts, such as those associated with charity and self-​sacrifice. Social incentives will give prominence to the role of nonpecuniary drivers of prosocial behavior. For example, social incentives in the form of praise, recognition, social approval, ostracism, or public embarrassment can serve as an internal reward or punishment to stimulate certain types of behavior.These social incentives operate through feelings of status, esteem, pride, or fear. People may be more motivated to act altruistically simply because it makes them feel good about themselves. The move toward considering social motives and incentives that take account of reasons why people behave the way they do is also reflected in the policy context. Economic incentives as external rewards may have counterintuitive and counterproductive effects on human behavior. Economic incentives may diminish intrinsic motivations such as altruism, professionalism, or civic duty. Providing financial rewards for the performance of activities that are originally motivated by intrinsic reasons leads to a reduction in the performance of those activities. The interaction between nonmonetary motives and external interventions can render certain policies less effective. Monetary payments contingent on performance would reduce the intrinsic motivation to carry out the task. With a payment, a person has no way to demonstrate one’s willingness to perform an activity for reasons other than monetary ones. Motivational crowding out not only reduces the effectiveness of policy but also may reduce activity below the pre-​policy level. The potential drawbacks of financial rewards have gained increasing attention, with concerns about motivational crowding out taking center stage. The concept of motivational crowding out derives from research conducted in the 1970s, beginning in payment for blood donation and expanding to a broad range of incentive programs. Most blood is donated to be used for other individuals. Blood donation is a prosocial act that imposes a cost to the individual but benefits others. Donation is intrinsically (as opposed to extrinsically) motivated. As Titmuss (1970) famously pointed out, providing economic incentives to blood donors may crowd out blood supply. A crowding out effect would mean that donation rates actually decrease after the introduction of payment. According to Titmuss’ work, a system of paid donation decreases opportunities to behave altruistically and thus harms altruism, while a system without payment fosters altruism in society. Pure altruism is the key driver for giving blood in his research. Purely altruistic donors may feel less inclined to donate if financial rewards are involved.3 They may feel their altruism is no longer needed, or they may perceive the use of payments as controlling. A financial reimbursement for blood donation would reduce the intrinsic motivation behind individuals’ donation behavior, producing a decline in supply from those individuals. The introduction of economic incentives thus causes crowding out. Contrary to a common assumption implicit in standard economic theory, the effects of extrinsic motivations such as economic incentives do not necessarily complement intrinsic motivations. Therefore, under relevant

Social preferences and moral economy  111

circumstances, it is not advisable to use the price mechanism to elicit a higher supply, and one should rely on intrinsic motivation. Moreover, due to its altruistic foundation of voluntary behavior, unpaid blood donation would be safer than paid donation. According to Titmuss’ analysis, hepatitis rates from blood transfusions significantly decreased when the blood was donated rather than purchased. That is, donors who are not paid for blood have no incentive to hide an illness, which provides higher guarantees for blood quality. The blood donated for transfusion is free from harmful substances or infectious agents. There are concerns that economic incentives may compromise the safety of the blood supply by attracting high-​r isk donors who conceal information to obtain the rewards and undermine donors’ intrinsic motivation to donate blood. Individuals may care about what others think of them. Especially, individuals may not want to be seen as driven by monetary concerns.Then, image concerns are important for giving blood. If blood donations are not rewarded, those individuals with image concerns would be willing to donate.The prospect that others will find out how much effort is put into charitable donations motivates individuals to generate more donations. Individuals would sometimes try to fulfill the expectations of others that he or she should join a group going to a blood donor event.The reputation mechanism, which is a direct result of the influence of others, is then a decisive factor in the donation behavior of a person.This implies that people tend to act less altruistically if no one observes their actions. However, we also seem to prefer to have a positive view of ourselves independent of the view of others (Akerlof and Kranton, 2000). People are motivated by their own view of themselves as well as by how others view them. Daycare centers face the problem that parents sometimes arrive late to pick up their children, which forces teachers to stay after the official closing time. A conventional economic approach would suggest introducing a fine for collecting children late. Such a punishment is expected to induce parents to reduce the occurrence of picking up their children late. In the Israeli daycare center study, however, Gneezy and Rustichini (2000) find that the introduction of a penalty system increases undesired behavior. In particular, small amounts of extrinsic incentives are expected to have large negative effects on observed prosocial behavior. In the daycare centers, a small fine was imposed for each instance where a parent was ten or more minutes late picking up a child. The fine was intended, by imposing an additional cost to dysfunctional behavior, to reduce the number of latecomer parents. Once a fine was imposed, being late was priced. The relatively small fine signaled to parents that being late was not important. Therefore, the lateness incidence increased. Furthermore, after the fine was removed, the increased frequency of lateness persisted. The message (that it was not so bad to be late) did not disappear once the fine was removed. It was difficult to revert back to the original level of arriving late. The formalization of the rules regarding collecting children late entailed a change in the nature of sanction, in the Israeli daycare center case, from social to material sanctions. According to Bowles (2016), this type of behavioral failure occurs partly because such rules are based on faulty assumptions that

112  Social preferences and moral economy

people are selfish and amoral. Economic incentives are appeals to self-​interest, and hence people might become more selfish, even in the absence of such incentives.

5.2  Social preferences and fairness People are said to have entirely self-​interested preferences if and only if they always choose in such a way as to maximize their own (expected) pecuniary payoffs. A purely self-​interested person refuses to contribute anything in the provision of a public good and free rides on the contribution of others.4 Conventional economic theory is thus built on the assumption that all people are entirely self-​interested and do not care about the well-​being of others. However, the self-​interested hypothesis has come into question. This hypothesis may be true for some, but it is certainly not true for all. Individuals exhibit prosocial behavior when they do not always make choices that maximize their own pecuniary payoffs. They often act pro-​socially, contribute to charities, and engage in pro-​environmental behavior, even if this imposes costs on them. According to conventional economics, people care only about their own consumption of public goods but do not directly benefit from their own contribution, nor are they directly affected by others’ consumption or contribution. Behavioral economics provides an alternative view to help explain why people make private provision of public goods. People making voluntary contributions to public goods, such as blood donations, and voluntary collection and recycling of waste, cannot be explained solely by pure self-​interest. They can care about others’ outcomes and place value on social goods. They value not only their own consumption but also the consumption of others: they are other-​regarding. The possibility that some individuals exhibit ‘social preferences’ (i.e., fairness concerns, reciprocity, and even pure altruism) has gained a more general acceptance among economists. Formal models of social preferences assume that some people may be self-​interested, but this need not be the case for all people. An individual exhibits social preferences if the individual cares not only about the resources allocated to him or her but also to others. People give to others who are worse off and pay to punish others who did them wrong. Different models have been proposed to explain this behavior. Some people do not like to think of themselves as motivated only by self-​interest. They may be intrinsically motivated by social preferences. The effect of social incentives depends on the structure of social preferences.To generate social incentives, we need to augment preferences to include the outcomes of others.That is, a player’s utility function not only depends on his or her material payoff but may also be a function of the allocation of resources within his or her reference group. More formally, given a group of N persons, let x = (x1, x2, …, xN) denote an allocation of physical resources out of some set X of feasible allocations. The utility of individual i may be any function of the total allocation. The self-​interested individual i’s utility only depends on xi, the monetary payoff of i. Individuals i has social preferences if, for any given xi, i’s utility is affected by variations of xj, j ≠ i. Especially, individual i is altruistic if the first partial derivatives

Social preferences and moral economy  113

of u(x1, x2, …, xN) with respect to x1, x2, …, xN are strictly positive (individual i’s utility increases with the well-​being of other people). People are averse to inequality between themselves and others, and they are thus motivated to reduce any inequality. Inequality aversion means that people resist outcomes that are perceived as inequitable (outcome-​based social preferences). People are then willing to give up some material payoffs to move in the direction of more equitable outcomes. People are uneasy, to a certain extent, about the presence of inequality in the model of Fehr and Schmidt (1999). Their model of inequality aversion assumes that one’s relative standing in the income distribution is important. Fehr and Schmidt (1999) set up a model of ‘self-​centered’ inequality aversion, in which inequality aversion means that people dislike inequitable outcomes and self-​ centered indicates that this aversion mainly stems from a comparison of their own and others’ payoffs. It is assumed that people derive utility from their own income situation but suffer from inequality in terms of being either better or worse off in material terms than others. Inequality is ‘advantageous’ if people are better off than others.This results in feelings of guilt.To reduce the inequality is one way to reduce the adverse feelings of guilt. The Fehr and Schmidt’s (1999) utility function, in addition to a standard neoclassical term, includes two terms: positive and negative deviations of one’s own payoff, each weighted with its parameter. People, in this setting, dislike being worse off than others more than being better off. Heterogeneity is explicitly accounted for by assuming a distribution of preferences in the population. Utility is assigned based on a player’s own payoff and an other-​regarding component that compares his or her payoff with the payoffs of others. The social comparison function is based on the difference between player i’s own payoff xi and the payoffs of all other players in the game. Given a group of N persons, the Fehr and Schmidt’s (1999) utility function of player i is given by: U i ( x ) = xi −

αi N −1

∑ max x j ≠i


− xi , 0 −

βi N −1

∑ max x


− x j ,0 ,

j ≠i

where αi is a parameter that measures how much player i dislikes disadvantageous inequality (an ‘envy’ weight), and βi measures how much i dislikes advantageous inequality (a ‘guilt’ weight). Player i’s utility positively depends on his or her own income xi and is negatively related to the difference between his or her income and that of other j. The second term expresses how player i dislikes being worse off than others, while the third term shows disutility from being better off than others. It is assumed that αi ≥ βi ≥ 0 and βi < 1. Here, αi ≥ βi ≥ 0 indicates that player i’s utility loss from disadvantageous inequality (xi < x j ) is larger than from advantageous inequality (xi > x j ). The disutility from being worse off than others is likely to be greater than the disutility from being better off. The player dislikes advantageous inequality less than disadvantageous inequality. βi < 1 indicates that player i does not suffer terrible guilt when he or she is in a relatively good position.

114  Social preferences and moral economy

By normalizing Terms 2 and 3 with N − 1, the impact of inequality aversion on i’s utility becomes independent from the number of players N. In the two-​player case, the utility function is simplified to:



U i xi , x j = xi − αi max x j − xi , 0 − βi max xi − x j , 0 ,

i≠ j

The second term in the utility function measures the utility loss from disadvantage inequality, while the third term measures the loss from advantageous inequality. That is, the second term reflects how much player i dislikes disadvantageous inequality, discounted by an individual sensitivity parameter αi, and the third term reflects how much player i dislikes advantageous inequality, again discounted by an individual sensitivity parameter not more than αi. The players do not like inequality for its own sake, but they dislike disadvantageous inequality more than they dislike advantageous inequality. Charness and Rabin (2002), the successor to Fehr and Schmidt (1999), introduce a model which allows for a wide variety of ‘distributional preferences.’ In Charness and Rabin’s (2002) two-​person model, a player’s propensity to sacrifice for another player is characterized by parameters: the weight on the other’s payoff when he or she is ahead and the weight on it when he or she is behind. Letting xi and xj be players i’s and j’s payoffs, they consider the simple formulation of i’s preferences as follows:



U i xi , x j = (ρr + σs ) x j + (1 − ρr − σs ) xi,

i ≠ j,

where r = 1 if xj > xi, and r = 0 otherwise; s = 1 if xj < xi, and s = 0 otherwise. The weight that i places on j’s payoff may depend on whether j is getting a higher or lower payoff than i. The parameters ρ and σ allow for a range of different distributional preferences. One form of distributional preferences is simply competitive preferences, which is presented to be σ < ρ ≤ 0. Then, players like their payoffs to be higher relative to others’ payoffs. Another form of distributional preferences is inequality aversion, which corresponds to σ < 0 < ρ < 1. Then, players prefer to minimize disparities between their own payoffs and those of others. Universal principles, such as fairness, individual rights, and justice, emphasize the welfare of all individuals equally. Fairness itself seems to matter for some individuals. On the other hand, the judgment whether others consider one-​self ’s behavior as fair is also taken into account in an individual’s decision-​making. The use of fairness comes from the classical solution of the problem of the fair division of a cake between two persons, where one person cuts and the other chooses. This is generalized so that a lack of envy means that i does not prefer j’s bundle of commodities (in this case, j’s slice of cake) to i’s own. No interpersonal

Social preferences and moral economy  115

utility comparisons are then made since each person compares the valuation of his or her bundle to the other’s bundle, not the other’s utility. Prosocial behavior is conditional on other people’s cooperative behavior. Reciprocity is related to the concept of conditional cooperation: people cooperate if others cooperate too. The existence of conditional cooperators renders the individuals’ beliefs about others’ behavior important. These beliefs can be based on past behaviors in a repeated interaction, but they can also be based on the knowledge that the interacting members are alike. A reciprocal altruist is a conditional cooperator in the positive sense. Reciprocity might therefore be used to inform the appropriate system-​wide incentive mechanism if cooperation is deemed desirable. Reciprocal acts are often attitudinal. An unkind response to an unkind action is an act of negative reciprocity. Consider Fehr et al.’s (1997) experiment in which respondents are asked to assume they are either employers or employees. Following the neoclassical economic assumption, those who place themselves in the position of employees would be entirely self-​regarding and, therefore, choose a zero-​cost effort level in a hypothetical contract irrespective of the wage offered to them. According to Akerlof (1982), however, a wage higher than the minimum one necessary would often be perceived by employees as a ‘gift,’ and they would consequently work harder than self-​interest dictates. In Fehr et al.’s (1997) experiment, the higher the wage offered by employers, the higher the effort level to which employees committed. Employers recognize that employees would be predisposed toward strong reciprocity.

5.3  Altruism and envy Individuals care about how well they perform in comparison with the relevant others.They are influenced by their reference group and who their reference group is. People may derive utility from being perceived as fair. Some individuals maintain their image of fairness by adjusting their actions to suit their moral judgments.They are encouraged to behave less selfishly because of reputational concerns. Altruism means that people help others while making sacrifices. It involves acting to increase the welfare of others, incurring personal costs but lacking an increase in personal welfare. People motivated to help by altruism are driven by an ultimate desire to help others, at a personal cost, without any personal benefit. Blood donation can be regarded as a prototype of the altruistic act: the unawareness of who donates and who receives. An unpaid blood donor receives considerable emotional gratification as a benefit of donation, which constitutes a powerful motivation. Relative concerns seem to be asymmetric. Individuals may compare themselves to others around them. An increase in the situation of others will depress their evaluation of their own life and undermine their dignity. They may become malevolent or envious in this context. To examine what happens to interpersonal comparisons, we extend the scope of relevant preferences to include preferences over one’s relative position.

116  Social preferences and moral economy

The model proposed in this section has the following features. There is a finite number of workers in an organization. The strategy of an individual is his or her effort level. An individual’s utility function captures both self-​regarding and other-​ regarding preferences. The individual’s underlying motivation is modeled explicitly. Two possible types of other-​regarding preferences, which exhibit ‘envy’ and ‘altruism,’ are considered. Each psychological factor, which is denoted by a parameter, enforces a certain ‘internal’ norm. Individuals are initially endowed with their norms, but they can acquire and internalize norms through ‘socialization.’ In order to trace the evolution of norms through socialization, this section formulates a random-​matching model in the organization. While individuals act to maximize their utility, it is the prevalence of norms that determines the social norm they obey in the long run. Individuals with different norms will typically take different actions. Chance events have large and persistent effects due to positive feedbacks. It is shown that different social norms generate different equilibrium effort levels within an organization. We have two steady states in the long run: an equilibrium with high morale and an equilibrium with low morale. This multiplicity can explain why economic performance varies across organizations. We consider a group in which there are N individuals.5 Each individual j chooses his or her effort level, ej (≥ 0), j ∈ {1, …, N}. Suppose that output is some function of each agent’s effort, given by f(e), where e is an N-​dimensional vector of agents’ effort levels. Consider a work situation in which each agent’s compensation is determined as f(e)/​N. It is painful to put forth effort, and the pain that an agent feels is given by c(ej) that is exogenous, where c′ > 0 and c″ > 0. Following Teraji (2007), agent i’s utility function is given by: U i (e; µi ) =

f (e ) f (e ) − c (ei ) + µi (N − 1) N N


The first two terms of the right-​hand side of (5.1) capture self-​interest, and the last term represents other-​regarding preferences. The parameter μi is the weight (or the ‘binding power’ of the internal norm) that agent i’s preferences put on all other agents’ well-​being. Also, (N − 1)f(e)/​N is the aggregate compensations of all agents other than i in the N-​member group under consideration.The last term is an attempt to formalize the discussion of internal norms, which are social and endogenous. Suppose there are two possible types of norms that can be internalized by individuals. We say that agent i’s internal norm is enforced by ‘envy’ if μi ≡ μ− < 0. With the binding power of this sort, i’s utility is a decreasing function of the compensations of other agents. On the other hand, we say that agent i’s internal norm is enforced by ‘altruism’ if μi ≡ μ+ > 0. Then, agent i’s utility is an increasing function of the compensations of others. Initially, each agent is endowed with a particular norm. Each agent chooses his or her effort level, given the parameter μi. As the utility function Ui is strictly concave in ei and the efforts are continuous, there would be a unique utility maximizer for Ui. High (or low) effort level corresponds to high

Social preferences and moral economy  117

(or low) morale in the group. We next present an equilibrium concept somewhat different from the standard Nash equilibrium of a game. Definition 5.1: Given the binding power of agent i’s internal norm μi, an effort profile e* is a morale equilibrium if:





U i e * ; µi ≥ U i ei , e*−i ; µi ,

∀i , ∀e i .

A morale equilibrium holds so that each agent takes others’ effort as given. It is necessary to compare morale equilibria, one with μ+ and one with μ−. We will show that equilibrium effort is higher with μ+ (> 0) than it would be with μ− (< 0). For analytical simplicity, we assume that f(e) = Σjej and c(ei) = ei2/​2. Thus, (5.1) can be rewritten as: U i (e; µi ) =


j j


e ei2 j j + µi (N − 1) 2 N

The model is constructed in discrete time, indexed by t = 0, 1, 2, …. We model the evolution of norms, namely, those that are acquired and internalized through socialization in the organization.The model represents each individual’s updating as a process of switching from one norm to another through learning from others.We introduce the socioeconomic interaction between individuals in the group. At the beginning of time t, the number of agents with trait μ+ (> 0), that is, the binding power of the internal norm enforced by altruism, is denoted by n(t). And let τ(t) be the fraction of individuals with trait μ+ at the beginning of time t, that is, τ(t) = n(t)/​ N. Thus, the fraction of individuals with trait μ− (< 0), that is, the binding power of the internal norm which is enforced by envy, is 1 − τ(t) = (N − n(t))/​N at the beginning of t. We formulate a random-​ matching model in the organization. The model introduces chance events that affect population dynamics through socialization. Individuals try to socialize themselves to particular norms at the beginning of each time. Socialization occurs as follows. With a certain probability of socialization, agent i is matched with an individual in the organization and then adopts the norm of that individual. The probability of socialization is allowed to depend on the fraction of individuals in the organization. In this model, socialization acts as a type of ‘complementarities.’ That is, an agent with trait μ− (or μ+) has more incentives to socialize with an individual with trait μ+ (or μ−) as μ+ (or μ−) is the more widely dominant one in the group. Specifically, socialization to agent i with trait μ− in t occurs with probability q(τ(t)), which is a continuous, strictly increasing function in τ(t). On the other hand, socialization to agent i with trait μ+ in t occurs with probability q(1 − τ(t)), which is a continuous, strictly increasing function in 1 − τ(t). With probability 1 − q(τ(t)) or 1 − q(1 − τ(t)), agent i is ‘naïve’ and does not get randomly matched with somebody else in the group. Moreover, it is assumed that q(0) = 0 and q(1) = 1.

118  Social preferences and moral economy

Let P+ −(t) denote the transmission probability that an agent with trait μ+ is socialized to trait μ− at the end of time t. The socialization mechanism is then characterized by the following transition probability:  N − n(t )  P + − (t ) = q 1 − τ (t )  .  N −1 




This represents the coefficient of norm transmission, where an agent with trait μ+ (at the beginning of t) is matched with an individual with trait μ− of the remaining individuals and adopts and internalizes that trait at the end of t.The transition probability, P+ +(t), that an agent with μ+ sticks to that trait in the socialization mechanism is given by:  n (t ) − 1  P ++ (t ) = 1 − q 1 − τ (t ) + q(1 − τ (t ))    N − 1 




In the term 1 − q(1 − τ(t)) of the right-​hand side of (5.3), an agent with μ+ does not get matched with somebody else in the organization and sticks to that trait in t. In the last term, an agent with μ+ at the beginning of t has a chance of socialization but is matched with an individual with μ+ of the remaining individuals in the organization (the agent does not change his or her trait). Similarly, for an agent with trait μ− at the beginning of t, we get:  n(t )  P −+ (t ) = q( τ(t ))    N − 1 


 N − n (t ) − 1  P −− (t ) = 1 − q τ (t ) + q( τ(t ))    N − 1 



( )

Given these transmission probabilities, the fraction τ(t + 1) of individuals with trait μ+ at the beginning of time t + 1 is calculated to be:

τ (t + 1) = τ (t ) + P −+ (t ) (1 − τ (t )) − P + − (t ) τ (t ) . Substituting (5.2) and (5.4), we obtain:  N − n (t )  τ (t + 1) − τ (t ) =   τ(t ){q(τ (t ) − q (1 − τ (t ))}.  N − 1 


This is the equation for the population dynamics through socialization. It is easy to derive conditions that guarantee that the population dynamics of (5.6) converges

Social preferences and moral economy  119


1 q(1−τ) q(τ)


O (1)

1 (O) τ

(1−τ FIGURE 5.1  Population



to a ‘homogeneous’ population in the limit. In Figure 5.1, the probabilities of socialization, q(τ) and q(1 − τ), are measured vertically: q(τ) slopes upward to τ and q(1 − τ) slopes downward to τ. And q(1 − τ) is a mirror image of q(τ). Assume that τ(0) > 1/​2 holds initially. Then, we have ∂{τ(t + 1) − τ(t)}/​∂τ(t) > 0 from (5.6). Thus, we have τ(t) → 1 for any τ(0) > 1/​2. Everyone adopts the same norm eventually. Similarly, the dynamics implies that τ(t) → 0 for any τ(0) < 1/​2. To summarize: Proposition 5.1: For the fraction of individuals having the internal norm which is enforced by altruism, τ(t), in the group at the beginning of time t, (1) τ(t) converges to 1 if τ(0) > 1/​2 holds initially, and (2) τ(t) converges to 0 if τ(0) < 1/​2 holds initially. Thus, the process by which norms evolve through socialization exhibits generalized increasing returns. Social norms take the form of conventions, that is, the modes of behavior to which a majority of agents subscribe. People follow

120  Social preferences and moral economy

a certain norm through socialization. In the presence of generalized increasing returns, chance events have large and persistent effects due to positive feedbacks. In the self-​enforcing process that depends on the initial prevalence of norms, only one social norm emerges. The social norm, which is enforced by altruism or envy, persists in the long run if a majority of agents adopt that norm initially. In the socialization mechanism, individuals behave in conformity with a particular social norm eventually. Agent i’s utility function, whose binding power of the internal norm is μ+ or − μ at the beginning of time t, is given by:



∑ e (t ) − e

2 i



∑ e (t ) − e

2 i

U i e (t ) ; µ + = U i e (t ) ; µ − =

j j


j j


∑ e (t )

(t ) + (P ++

(t ) µ + + P + − (t )µ − )(N − 1)

(t ) +

(t ) µ + + P −− (t ) µ − ) (N − 1)





j j



j j


(t )


at the end of t. Agent i chooses i’s effort level at the end of each time. Then, there is a unique effort level maximizing the utility function. The utility-​maximizing effort levels for agent i in t, ei∗(t; μ−) and ei∗(t; μ+), are given by:

( ) (t; µ ) = arg max

) ( U (e (t ) ; µ ) .

ei* t ; µ + = arg max ei ≥ 0U i e (t ) ; µ + , ei*

ei ≥ 0


The first-​order conditions are: ei* t ; µ + =

( )

1 N −1 , + P ++ (t ) µ + + P + − (t ) µ − N N

( )

1 N −1 . + P −+ (t ) µ + + P −− (t ) µ − N N

ei* t ; µ − =







Suppose that τ(0) > 1/​2 holds initially. From Proposition 5.1, τ(t) → 1 in the limit. Then, we have P+ +(t) → 1 and P+ −(t) → 0 in (5.2) and (5.3). Thus, it holds that ei∗(t; μ+) → {1 + μ+ (N − 1)}/​N in (5.7). Furthermore, in (5.4) and (5.5), n(t)/​ (N − 1) → 1 and (N − n(t) − 1)/​(N − 1) → 0 for the remaining individuals other than agent i. Then, we have P− +(t) → 1 and P− −(t) → 0. Thus, ei∗(t; μ−) → {1 + μ+ (N − 1)}/​N in (5.8). Similarly, if τ(0) < 1/​2 holds initially, we have ei∗(t; μ+) → {1 + μ− (N − 1)}/​N and ei∗(t; μ−) → {1 + μ− (N − 1)}/​N in the limit. Morale is norm-​based motivation. The social norm generates the equilibrium effort level in the long run. Everyone chooses the same effort level in equilibrium because of a particular social norm that prevails.When there are multiple equilibria, these equilibria can be ranked. Morale is higher in the high-​effort equilibrium than

Social preferences and moral economy  121

in the low-​effort equilibrium. Thus, we have two morale equilibria: in one equilibrium with high morale, every agent’s effort level is {1 + μ+ (N − 1)}/​N, and in the other equilibrium with low morale, every agent’s effort level is {1 + μ− (N − 1)}/​ N, where μ+ > 0 > μ−. Thus, we have the following: Proposition 5.2: (1) Suppose that τ(0) > 1/​2 holds initially. Then, we have a morale equilibrium e = (e, …, e), where e = {1 + μ+ (N − 1)}/​N. (2) Suppose that τ(0) < 1/​2 holds initially. Then, we have a morale equilibrium e = (e, …, e), where e = {1 + μ− (N − 1)}/​N. The evolution of norms allows us to study long-​run equilibria with different effort levels. Which norm is persistent determines the equilibrium effort level within an organization. If the norm is enforced by altruism, every agent chooses the high equilibrium effort level in the long run. In this morale equilibrium, every agent is able to maintain high morale. Thus, norms inculcated through socialization can lead members to create incentives within an organization. On the other hand, if the norm, which is enforced by envy, persists, the low equilibrium effort level can be sustained. In this morale equilibrium, everyone shirks. Then, we have the decay of economic performance in the long run. Thus, the economic performance can vary depending on the norm that prevails within an organization. A key finding of path dependence is a property of ‘lock-​in’ by historical events (Arthur, 1994). In a world of increasing returns to scale (positive feedbacks), initial and trivial circumstances can have important and irreversible influences on the ultimate market allocation of resources. Path-​dependent economics is a theory of equilibrium selection with positive feedbacks, and a path-​dependent outcome is associated with characteristics of persistence and uniformity. The economy has a multiplicity of possible equilibrium solutions, where the dominant solution can be the sub-​optimal one. In this section, effort depends on the prevailing norm within an organization. In a world of effort discretion, the economic performance can vary across organizations. Furthermore, the process by which norms evolve through socialization exhibits generalized increasing returns. One social norm emerges in the process that depends on the initial prevalence of norms. The social norm, enforced by altruism or envy, persists in the long run. Individuals try to socialize themselves to one particular norm. Which norm is persistent determines the equilibrium effort level. It is possible for low-​effort organizations to survive in the long run. Then, the norm, enforced by envy, can be a cause of economic inefficiency. If the norm is enforced by envy, there does not exist the private incentives for individuals to choose the high-​effort equilibrium. In a world where norms are determinants to effort levels, the path-​dependent low-​effort equilibrium can exist. These discussions connect with what Harvey Leibenstein referred to as X-​ inefficiency. A large number of X-​inefficiency studies exist, dating from Leibenstein

122  Social preferences and moral economy

(1966). He referred to the difference between maximal effectiveness of utilization and actual utilization as the degree of X-​inefficiency. According to him, individuals supply different amounts of effort, where effort is a multidimensional variable, under different organizational and environmental circumstances. Individual motivations and interactions between individuals are important. Leibenstein (1987) presents a reasonable vision of human behavior within the organizational context, where X-​inefficiency can persist stubbornly. Given effort discretion, the quantity and quality of effort can vary. Labor productivity is affected by the quantity and quality of effort inputted into the process of production. The introduction of effort variability allows for the existence of path-​dependent high-​and low-​productivity firms in the long run. Individuals may make different decisions, depending on social contexts. Differences in context-​dependent behavior can be interpreted as being generated psychologically, ethically, and sociologically. This section has provided theoretical insights into the diversity of economic performances within organizations. The approach taken here can shed light on issues on endogenous norms in society. This section has argued that diverse morale and the evolution of norms interact in nontrivial ways. Agents can acquire and internalize norms through socialization. The process by which norms evolve through socialization is self-​enforcing and depends on the initial distribution of individuals with certain norms in the group. In this formulation, different norms typically exhibit different morale equilibria in the long run. The resulting dominance of a certain norm implies that a certain social structure appears more often than otherwise. The model has proposed two morale equilibria with different effort levels. If the norm, enforced by altruism, persists in the limit, a ‘good’ morale equilibrium emerges, in which everyone chooses a high effort level and high morale is sustainable within an organization. Norms inculcated through socialization can lead individuals to create incentives. On the other hand, if the norm, enforced by envy, persists in the limit, a ‘bad’ morale equilibrium emerges, in which everyone shirks. Morale can be thus norm-​based motivation. Given effort discretion, it is possible to expect a multiplicity of possible equilibrium solutions, where the dominant solution is sub-​optimal. The economic performance can vary because of the norm that prevails in organizations.

5.4  Public goods and charitable giving Prosocial behavior includes voluntary contributions to public goods. Individuals share a common interest in the provision of public goods. Typically, there is a choice of how much to contribute to a common pool, which then benefits all participants equally. Public goods provision is costly, which implies a tension between the individual and collective interest. The maximum payoff for the group is achieved if everyone contributes the full amount, but free riders increase their own payoff by withholding their contribution and still benefitting from the public goods. Prosocial behavior is usually at odds with standard economic analysis, which predicts that, in a noncooperative setting, individuals only make

Social preferences and moral economy  123

negligible contributions to public goods. Although mutual cooperation leads to the best possible collective outcome, individuals have an incentive to free ride on the contributions of others. That is, collective welfare is maximized if every agent makes maximal contributions, while selfish rationality implies zero contributions. According to the voluntary contribution model of public goods provision, as a general presumption, self-​ interested individuals who independently choose their contribution levels will generate a sub-​optimal level of the public goods (Bergstrom et al., 1986). People reduce their cooperation if others contribute less than themselves to public goods provision, which causes selection for noncooperation that eventually undermines collective action. As the free riding effect dominates in a population, the share of individuals making contributions to a public good tends to zero. In principle, governments can enforce the efficient provision of public goods. However, they generally lack the necessary information to do so, and the attainment of the social optimum is hampered by the selfish interest of each individual to give false signals. Laboratory experiments have been intensively used to investigate whether individual and collective decision-​making and behavior correspond to theoretical predictions.6 The public goods game is a suitable research tool for studying cooperative problems. Each of group members, in this game, receives an endowment of tokens. Players have to decide how many tokens to keep for themselves and how many to contribute to a group project. According to experimental findings concerning the voluntary provision of public goods, there is much less free riding and much higher voluntary contribution in one-​shot versions of the standard public goods games than theory suggests, but the public goods provision is still below the efficient level. Conditional cooperators contribute more if they expect others to contribute more.7 Individuals would be willing to contribute to public goods if they know that others do not free ride and also contribute. They adapt their expectation of others’ contribution on the basis of their experience of the average collective contribution in the previous rounds. In repeated public goods games, free riding increases and contributions decline over time. Conditional cooperators decrease their contribution accordingly, which causes the average contribution to further decline. Individuals are often tempted to cheat in social interactions, thereby gaining a benefit at the expense of cooperative partners.8 To encourage partners to behave cooperatively, individuals might therefore use control mechanisms that render cooperative behavior a more profitable option than cheating. One such mechanism is punishment. The opportunity to impose sanctions on free riders can potentially solve the collective action problem because being punished induces free riders to cooperate more. Rational and selfish free riders never sanction when this is costly. However, anticipation on being sanctioned will induce them to contribute, provided that the loss due to received punishment offsets the payoff advantage of free riding. Punishing a free rider, in cooperative group situations, can be considered a cooperative act because all group members benefit from the resulting increase in the free rider’s level of cooperation.

124  Social preferences and moral economy

Altruism has traditionally been studied in the context of helping, but punishment can also qualify as altruistic act. An altruistic punisher is willing to punish unfair others at a personal cost. Altruistic punishment describes behavior for which individuals are willing to incur personal costs to punish wrongdoing, even though they do not personally suffer from the misconduct and are an unrelated stranger. That is, people punish norm violators not for what they did to the punisher but for what they did to others (third-​party punishment). A strong reciprocator is defined as an individual who is both a conditional cooperator (reciprocal altruist) and an altruistic punisher (Gintis et al., 2005). If the number of altruistic punishers in a population is too small, selfish actions would become more profitable and drive out reciprocal altruism. Since it is costly to punish, no selfish third party will ever punish. When there is a sufficient proportion of strong reciprocators, a high-​level cooperation in a group could be attained. Fehr and Gächter (2000) introduce a two-​stage punishment mechanism in a public goods experiment. The first stage corresponds to a single period of the voluntary contribution mechanism. In such a mechanism, each subject is asked to allocate his or her endowment between a private and a group account. In the second stage, individual decisions are anonymously revealed to the group, and each subject has an opportunity to punish each other. Punishment is costly, both to the person imposing the punishment and the person being punished. Since punishment is costly, a self-​interested person will never punish. Nevertheless, Fehr and Gächter (2000) report frequent punishments even in the last period of the experiment. The fear of punishment has a strong positive effect on cooperation. Thus, public goods provision is significantly higher in the experiment with opportunities to sanction than in the experiment without opportunities for sanctions. People trust punishers more than non-​punishers. When punishing a free rider is good for a group, it could signal the punisher’s trustworthiness, concern with fairness, or commitment to that group. Others might be more willing to enter into relationships with people who have demonstrated that they would not tolerate unfairness. If punishment is a signal of trustworthiness or fairness, punishers may receive more benefits from cooperative partnerships than non-​punishers. The economic value of the resources devoted to charitable giving or volunteering is considerable. In economics, charitable act is a voluntary transfer that is not motivated by market exchange. It is a form of economic sacrifice by the giver in exchange for the receivers’ benefit for which the giver expects no return. Charitable donations are a form of indirect helping, often without any direct exposure to the beneficiary or direct knowledge of how the money will be used. There are two sources of utility that may encourage an individual to behave pro-​socially and contribute to the public good. First, individuals care about the provision of the public good (‘pure altruism’). Second, individuals may also gain utility from contributing, meaning that they value their own contribution more highly than that of someone else’s (‘impure altruism’). That is, individuals may experience a feeling of moral satisfaction derived from their contribution to the common good environment. When contributing to the public good yields some utility benefit to

Social preferences and moral economy  125

the individual, voluntary contributions can be consistent with standard economic models. Andreoni (1990) develops a model of donation with ‘warm glow’: the individual’s utility is not just as a function of the consumption of the private and public goods but also of the individual’s contribution to the public good itself.That is, the utility of individual i is typically specified as: Ui = Ui(xi, Y, gi), where xi is private consumption by i, Y = ∑ i gi is the total supply of charitable giving, and gi represents i’s private gift, the form of a warm glow associated with charitable giving. In particular, when Ui = Ui(xi, Y), the individual i cares nothing for the private gift. Then, the individual is thought to be purely altruistic. Likewise, when Ui = Ui(xi, gi), the individual is motivated to give only by warm glow, hence is purely egoistic. When both Y and gi are arguments, the individual is impurely altruistic. Donors report feeling proud or experiencing a warm glow after or during the blood donation. Warm glow may increase an individual’s likelihood of engaging in charitable behavior and can induce helping behavior. A prominent proximal explanation for charitable giving is the warm-​glow utility flow (i.e., feeling good from doing good) from the act of giving. In the warm-​glow model, individuals do not donate because of pure altruism but partly because of a utility gain from donating. The individual driven by pure altruistic motives is indifferent about the source of the increased welfare of others. Individual contribution to the cause can be substituted by the contribution of another one. Warm glow is the personal satisfaction an individual enjoys from being actively involved in an activity independent of any consideration of outcome.Warm glow is not to imply selfishness. It motivates people to help all others. Donors never know who receives their donation. As warm glow is unrelated to the characteristics of the recipient, it can be characterized as an unconditional motivation for helping. Combining warm glow and pure altruism results in impure altruism. The individual then donates to attain warm glow and benefit others simultaneously. When people make donations to public goods, instead of being motivated by the increase in value of the common good, they rather experience a direct, personal benefit arising from the contribution itself.That is, the individual driven by impure altruism receives a private benefit through the act of giving that cannot be substituted by the contribution of another individual.

5.5  Philanthropy and corporate social responsibility ‘Corporate social responsibility’ (CSR) has become a high-​profile public issue nowadays. It denotes a situation in which firms aim to combine their economic goals with the taking of social responsibility for their impact on eco-​systems and human beings. Many companies contribute to social goals, investing more resources in public goods provision. A company’s social reputation influences purchase decisions

126  Social preferences and moral economy

for consumers. Nowadays, society at large increasingly expects companies to behave as good corporate citizens.9 Good citizenship metaphor emphasizes voluntary self-​ restraint and altruism concerning the realization of broad duties. Social responsibility derives from the moral legitimacy the corporation achieves in society. An exclusive focus on profit maximization is too limited as it neglects the company’s responsibilities toward the socially conscious consumers. Recent corporate scandals involving large companies have highlighted CSR as one of the principal issues confronting the free-​market system. The term ‘CSR’ has been defined in various ways.10 According to one of the most frequently cited definitions, CSR is “a concept whereby companies integrate social and environmental concerns in their business operations and in their interaction with their stakeholders on a voluntary basis” (Commission of the European Communities, 2001, p. 6).11 Companies have responsibilities to take ethical and moral issues into account while they make a profit. There is a growing interest in the relationship between corporate governance and the social performance of a company. Contemporary economic theory, however, has argued that so long as economic agents are rational and self-​interested, they will be able to affect a better set of consequences than if their intentions are benevolent. In fact, according to Friedman (1970), the social responsibility of a business is to increase its profits. In Friedman’s framework, firms are owned by shareholders who are principals, and business managers are agents with the duty to serve the interests of their principals. Managers are obliged by contract to shareholder value; it is their primary task to maximize profits. Following this view, CSR is a misallocation of corporate resources that would be better spent on value-​added internal projects or returned to shareholders. Extra costs of CSR activities do not necessarily translate into profits. In addition, CSR strategies may restrict a firm’s strategic decisions since certain strategic alternatives cannot be chosen if they are not aligned with the firm’s CSR activities. The assumptions of neoclassical economic theory are subject to market forces. In the neoclassical analytical framework, the market system creates a decision-​ making situation that favors practices to ensure profit maximization at the expense of more morally preferable alternatives. However, the lens of neoclassical economic theory is distorting practical realities. A positive view of managers’ support of CSR is articulated by Freeman (1984) who encourages consideration of external stakeholders, beyond direct profit maximization. Freeman’s broad definition of stakeholders includes any group or individual who might affect the business objective or might be affected by its realization. Stakeholder engagement can be defined as the practices which an organization undertakes to involve stakeholders in a positive manner in organizational activities. The company’s CSR response is viewed and evaluated by stakeholders. Stakeholder engagement is, accordingly, deemed to play a vital part in the development of CSR strategies. According to the stakeholder theory approach, it can be beneficial for the company to engage in certain CSR activities that stakeholders perceive to be important; unless stakeholders and their interests are dealt with, they might withdraw their support for the company. Freeman’s (1984) work helps to reconceptualize the nature of the company to

Social preferences and moral economy  127

encourage consideration of new external stakeholders, legitimizing new forms of managerial understanding and action. CSR should be understood as a broad concept since it takes in the whole set philosophical and normative issues relating to the role of business in society. Stakeholder theory suggests a moral relationship between companies and stakeholders. ‘Corporate social performance’ (CSP) is voluntary activities. Moral management yields CSP. Supplying CSP is not just like managing traditional market-​based activities. In fact, the relationship between CSP and profits has not been adequately determined from empirical studies. CSP entails a company’s recognition of broad responsibilities, and it should be concerned with more than just profit. Neoclassical economic theory, however, treats the firm as a black box, completely self-​interested entrepreneur that arranges inputs so as to maximize the material well-​ being. Friedman (1962) questioned the ability of business managers to pursue the social interest as follows: If businessmen do have a social responsibility other than making maximum profits for stockholders, how are they to know what it is? Can self-​selected private individuals decide what the social interest is? Can they decide how great a burden they are justified in placing on themselves or their stockholders to serve that social interest? Friedman, 1962, pp. 133–​134 Instead, Friedman (1962) argues that a socially responsible firm is one that conducts the business in accord with shareholders’ desires, which generally makes as much money as possible while conforming to the basic rules of the society. His position, based on free-​market ideology, has come under increasing attack since the time of writing. According to the shareholder perspective, business is about economic and not social goals. In real life, however, economic decisions also have social consequence, and the boundaries between the social and economic world become blurred. Firms have responsibilities toward their environment that go beyond their legal and economic obligations. The stakeholder model reflects the modern understanding of companies as integrated in, rather than separated from, the rest of society. Within our economic system, there exists a growing cadre of companies where social, environmental, and ethical goals are on equal footing with the profit motive. According to stakeholder theory, CSR includes various kinds of responsibilities or dimensions: each dimension of CSR can be examined in relation to the various stakeholders, including consumers, employees, investors, suppliers, and the communities, or the organization. CSR is broadly understood from distinct stakeholders’ perspectives suggesting that firms must create value not only for their shareholders but also for other stakeholders.To consider what it means for a company to be socially responsible, this section focuses on the ‘socially conscious’ consumer. Managers do business in a way that maintains or improves the consumer’s and society’s well-​ being. Consumers may expect companies to protect the environment and conduct

128  Social preferences and moral economy

business ethically. When consumers are given information about a company’s level of social responsibility, they evaluate the company. Lack of awareness of the level of social responsibility is likely to be a major inhibitor of consumer responsiveness to CSR. The socially conscious consumer is a person who takes into account the public consequences of his or her private consumption or who attempts to use his or her purchasing power to bring about social change (Webster, 1975).This concept is based on the psychological construct of social involvement. The socially conscious consumer must be aware of social problems, be active in the community, and believe that he or she has the power to make a difference. A higher perceived CSR could also increase workers’ motivation. Koppel and Regner (2014) propose a gift-​exchange game variant, in which CSR is captured by donating a certain share of a firm’s profit to charity. In their experiment, firms choose the share of profits they want to give to a charity, while workers have to decide, for each possible share, the effort level they want to provide. The level of share of CSR and the effort level are then positively correlated. Consumers are a major stakeholder group particularly sensitive to CSP. Consumers’ perceptions of a company as socially oriented are associated with a higher level of trust in that company and its products.12 It is important for socially responsible companies to develop consumer trust and engage in CSR programs that are meaningful to their consumers. Trust is a fundamental asset in the company–​ consumer relationship. Commitment to a brand, defined as trust, is the consumer’s desire to maintain the relationship. Socially oriented companies can use trust to improve their competitive performance (brand loyalty). Brand loyalty includes tolerance to pay a higher price for the product.13 Ethical firms producing a particular product are not producing the same one as unethical ones. Even if the product is identical in all other characteristics, this same product can be differentiated by the extent to which it is produced in accordance with a specified set of ethical standards (Altman, 2005). Bhattacharya and Sen (2003) argue how non-​product aspects of a company, such as CSR, can lead to consumer loyalty and other positive post-​purchase outcomes. There are a number of studies in which consumers claim to be ready to pay higher prices for products from socially responsible companies or to take the social responsibility profile of the producer into consideration when comparing different brands (e.g., Mohr et al., 2001). From a demand-​side perspective, a good reputation increases the value of the brand, which, in turn, increases the company’s goodwill. Companies are faced with the challenge of becoming more attractive to socially conscious consumers. It is important to investigate how corporate decisions are received by consumers. Consumers are engaged in constructing the ethical identity of companies. Managers are encouraged to behave in an ethical manner because information about a company’s ethical behavior is thought to influence consumers’ image of the company. Corporate identity emerges from the interaction between the company and its consumers. Identity theory at the individual level has been used to explain role-​ related behavior. Social structures affect the extent to which an individual commits to a particular role. Corporate identity denotes the characteristic way in which

Social preferences and moral economy  129

an organization goes about its business or how it behaves and interfaces with the social environment. Corporate behaviors regarding ethical conformity, in light of normative expectations, represent a response mechanism to the social environment. Socialization, through which individuals learn the perspective necessary to perform roles, is expected to prime ethical identification. Corporate identity interplays with social concerns to influence the response to values, norms, and beliefs. CSP can be then regarded as voluntary corporate commitment to exceed the explicit and implicit obligations imposed on a company by society’s prevailing ideas and opinions. On the one hand, we have the interests of the economy arising from a largely unregulated market of companies pursuing their self-​interest and, on the other hand, we have the interests of society arising from ethical or moral conformity that may suppress the effects of self-​interest. Managers have to legitimate their activities within the field of tension between ethics and economic success. These aspects are strongly reflected in the dimensions of ethical or moral perceptions. In this section, we identify CSP with the private provision of public goods (socially responsible or environment-​friendly activities) or the private redistribution of profits to social causes (Teraji, 2009). Social expenditures represent redistribution through community projects and support, philanthropy, training and educational programs, workers’ rights initiatives, environmental abatement and protection, and alternatives to animal testing (Baron, 2008). What motives a business to make a contribution to these public goods? We provide a theoretical framework for analyzing corporate incentives to engage voluntarily in socially responsible activities. According to social exchange theory (Blau, 1964), individuals are motivated to take voluntary actions when they expect they could get something in exchange. That is based on the assumption that people base all their decisions on the calculation of costs and benefits in the pursuit of their own material well-​beings.14 Some studies assume that private provision of public goods is motivated in part by warm glow (Andreoni, 1989; 1990). If the warm-​glow motive is strong enough, individuals may continue to make direct donations. CSP is modeled as a response to consumers’ preferences over public goods (Besley and Ghatak, 2007). Bagnoli and Watts (2003) study the strategic use of CSP to appeal to ‘green’ consumers with warm-​glow preferences for public goods.15 In this section, the manager does not engage in strategic CSP in an attempt to maximize only ‘pecuniary’ profits.The manager may act to meet the expectations of socially conscious consumers and thereby to receive the resulting ‘social satisfaction’ from CSP. The nonpecuniary motivation of agents for being involved in activities that benefit others is an equilibrium outcome. The present model is related to Baron (2008) in the organizational forms for private provision of public goods. However, the model developed here provides an alternative explanation for understanding how different social preferences can be created in manager–​consumer interactions. The model investigates how the consumer’s taste for CSP is endogenously determined in the social environment. In response, the manager’s taste for CSP is also endogenously determined in the environment. As consumer awareness of the need for CSP increases, managers increasingly recognize the responsibilities for implementing ethical programs to enhance social welfare.

130  Social preferences and moral economy

Through evaluation of the manager–​consumer relationship, the manager chooses whether to integrate CSP within corporate strategy. CSP commitment increases the company’s trust, making its operations more legitimate to consumers. Whether a company is being ethical or not is dependent on the manager awareness of creating trust in the manager–​consumer relationship. Justifying the role of CSP as an extended scheme of corporate governance is an essential part of the requirements of a just society understood as a joint venture for mutual benefit. In this section, following Teraji (2009), CSP is discussed as an outgrowth of ‘collaboration’ between managers and consumers in the social environment. The term ‘collaboration,’ or ‘dialogue,’ describes the involvement of consumers in the managerial decision-​making processes that concern social and environmental issues.The realization of the company’s social responsibilities is shaped by the way the manager interacts with consumers. The model focuses on the views of both managers and consumers about CSP. The manager has a key role in the supply of CSP, using the company’s resources to achieve its social goals. Active CSP is conceptualized as follows. The manager can anticipate social preferences of consumers and transform them into social benefits. Social benefits then reflect the ‘price’ a socially conscious consumer puts on the supply of CSP. If social benefits can be created in the manager–​consumer collaboration, the consumers’ ethical attitudes facilitate the corporate social responsiveness. These collaborative relations can reinforce positive responses with each other.The socially conscious consumers’ attitudes to encourage CSP are influenced by the manager’s view to practice CSP and vice versa. Thus, both mangers and consumers are influenced by comprehensive perception toward CSP.This establishes the basis for understanding the central role of both ethical self-​ regulation and consumers’ activism. Ethical self-​regulation establishes a new context for manager–​consumer interactions, and consumers’ activism becomes effective in this context.This framework thus focuses on how CSP is understood and perceived by both managers and consumers. The model considers two key aspects of social preferences: the consumer’s taste for CSP and the manager’s taste for CSP. The firm in the model economy normally produces an identical (numeraire) private good which can be consumed. There are potential shareholders in the firm. The shareholders of the firm are assumed to contact with a manager. The manager operates the firm after accepting the contract. Let π be the (pecuniary) operating profit when the project is implemented. After the operating profit is specified, compensation is paid to the manager and the remaining amount of the profit is distributed to shareholders. Let pπ, where 0 < p < 1, denote the manager’s compensation when the profit is specified. Similarly, (1 − p)π is distributed to shareholders when the profit is specified. Moral conduct refers to a pattern of behavior that goes beyond normal business management. The manager chooses units of social expenditures, g, which affect social welfare. By choosing g, the manager contributes to the private provision of public goods as CSP. The provision is then bounded. It is assumed that the manager picks up his or her own contribution g in some interval [0, G], where G > 1 is given.

Social preferences and moral economy  131

In this model, CSP is discussed as an outgrowth of collaboration between the manager of the firm and its consumers in the social environment. The model concentrates on two different dimensions of social preferences: the consumer’s taste for CSP and the manager’s taste for CSP. As a consequence, social preference formation interacts with economic decisions taken by others. The model analyzes the decision problem of the manager who receives the compensation alone and the manager who relies on social values. Consumers choose whether they consume the private good or become concerned about the CSP of the firm. Consumers have an initial endowment they can allocate between purchasing the private good and personal giving to social causes. They may care about moral management or CSP. The model considers the relative weight between the consumer’s concern for the private good and that for the amount of CSP supplied. Let the consumer’s taste for CSP be described by a fraction α (≥ 0). The fraction α represents the value that a consumer receives from the amount of CSP supplied by the firm, g. It can be also considered as the ‘price’ the consumer puts on the perceived benefits from the CSP of the firm. On the other hand, the consumption for the private good becomes 1 − α. Then the consumer benefits from both the private good and the amount of CSP supplied, g. More specifically, the consumer’s utility function is given by: U = ln(1−α) + αg. (5.9) The greater α is, the more utility a consumer gets from the CSP of the firm. The manager allocates his or her time between producing the private good and providing CSP (the private provision of public goods). The manager provides CSP by allocating a portion of β (≥ 0) of his or her time to social causes. The fraction β is considered as the manager’s taste for CSP. If the fraction β is positive, the manager chooses the amount of CSP supplied, g. On the other hand, the manager earns the payoff ln(1 − β) by allocating 1 − β of his or her time in producing the private good. Thus, the operating profit is specified as: π = ln(1−β) − βg. (5.10) In (5.10), g reflects the firm’s social expenditures, the cost of CSP. It may be also considered as the private redistribution of profits to social causes.The manager may view CSP as an expense rather than an investment. Managerial perceptions of CSP are important. The manager may receive a noneconomic or psychological benefit reflecting ‘social satisfaction’ associated with the CSP of the firm. In the model, the manager perceives social satisfaction, h, from one unit of CSP provided by the firm. Thus, if the manager’s taste for CSP is β, the manager gets βhg from providing the amount of CSP, g. Social satisfaction h reflects the manager’s social and environmental awareness. From moral management or CSP, the firm has an ethical responsibility to society. The socially responsible firm seeks to meet the expectations of consumers and to get a positive image from

132  Social preferences and moral economy

consumers in the economy. Thus, social satisfaction h depends on the consumer’s taste for CSP. More specifically, let h = h(α) be a strictly increasing, strictly concave function of α. This implies that social satisfaction h increases when the value the consumer puts on the CSP of the firm increases. It is assumed that h(0) = 0. That is, if consumers are excluded from the collaboration, the value h is zero. When the profit is specified, the manager receives his or her compensation pπ. The manager’s utility function is then given by: V = pπ + βh(α)g. (5.11) The manager struggles in deciding how to reconcile the benefit from CSP with the cost of CSP. The benefit from socially responsible programs is not guaranteed. To the extent that increased CSP results in improved benefit, the manger will be encouraged to become more socially responsible. The consumer chooses α to maximize (5.9) subject to α ≥ 0. The first-​order condition for the consumer’s choice of α is: −

1 + g ≤ 0. 1− α


For α > 0, (5.12) holds with strict equality. Then: 0  α =  g −1  g 

if g ≤ 1 if g > 1


The optimal α depends on the amount of CSP supplied by the firm, g. From (5.13), we can say that the consumer is less concerned about the CSP of the firm if the social expenditures are not sufficient (g ≤ 1). The consumer’s benefits from CSP can increase with the amount of the public goods provided. Thus, the fraction α measures consumer’s awareness for CSP. Lack of awareness is likely to be a major inhibitor of consumer responsiveness to CSP. For a positive α, we have: 1 d( g − 1) / g = 2 > 0. dg g Thus, α is a strictly increasing function of g if g > 1.The amount of CSP supplied by the firm has a significant impact on consumer responses. The manager’s choice problem is to choose g and β to maximize his or her utility in (5.11). Using (5.10), (5.11) can be rewritten as: V = p{ln(1−β)−βg} + βh(α)g. (5.11)′

Social preferences and moral economy  133

In (5.11)′, for a positive β, the first-​order condition for the manager’s choice of g over the interval [0, G] is: if h( α ) ≤ p if h( α ) > p

 0 g= G

Thus, we can say that the manager is encouraged to provide CSP if his or her social satisfaction is larger than the ‘marginal’ cost of CSP, or h(α) > p. In the model, h increases in α. Then, the manager’s choice of g is associated with the consumer’s taste for CSP.The manager assesses the social demand for CSP and then determines the optimal level of CSP to provide. The manager chooses β to maximize (5.11)′ subject to β ≥ 0. The first-​order condition for the manager’s choice of β is: −

p + h ( α ) − p g ≤ 0. 1−β



For β > 0, the above inequality holds with strict equality. Then, we have: 0  β =  h (α ) − p g − p  h (α ) − p g 









if h ( α ) − p g ≤ p if h ( α ) − p g > p.

The optimal β depends on the manager’s social satisfaction, h(α), the marginal cost of CSP, p, and the amount of CSP supplied, g. Vogel (2005), examining the links between ethics and profits, argues that: The emergence of ‘companies with a conscience’ represents a particular vivid expression of the contemporary reconciliation of social values and the business system. These are companies whose vision of social responsibility was integral to their business strategies from the outset. They were formed by individuals with strong personal social commitments who regarded their businesses both as vehicles to make money and as a means to improve society. Vogel, 2005, p. 28, original emphasis Vogel (2005, p. 2) defines CSR, or business virtue, as “practices that improve the workplace and benefit society in ways that go above and beyond what companies are legally required to do.” CSR belongs to strategy, and, in this respect, it is a matter of corporate policy. Companies choose to behave in a socially responsible way, in the same way they choose to spend more on marketing or production. Standard economic analysis, built on the assumptions of neoclassical theory, pays little attention to ethics and morality beyond direct profit maximization.

134  Social preferences and moral economy

The self-​interested economic agent has no social preferences. For instance, in the principal–​agent literature, the principal is the person whose material welfare should be maximized, and the agent should execute the orders of the principal by receiving an adequate compensation. Efficiency, in the Pareto sense, is the main goal of the theory.The agency conception is based on the pecuniary conflict of interests of the principal (shareholder) and the agent (manager). A model of business as an independent system has been represented by neoclassical economic theory. In a world of independent systems, the firm exists to maximize its profits. In the behavioral theory of the firm, on the other hand, organizational slack is a prerequisite for being able to afford corporate social strategy. Such strategies rely on individual managers making decisions based on their own values. Nowadays, society at large increasingly expects companies to behave as corporate citizens. Profits are not an end per se; they must be compatible with other social needs. To respond positively to moral responsibility, managers must find an alternative way of doing business. In reality, we recognize that the interactions among various factors of society are complex. From a behavioral theory perspective, this type of complexity is understandable as multiple and conflicting managerial goals. The individual is not separate from society and thus not separate from the corporation. Therefore, business is represented to be part of society, affecting it and being affected by other aspects of it. A relational view thus emerges with corporations embedded in society. CSR is about the basic idea that businesses have to meet society’s expectations in their practices. CSR involvement is fueled by various social demands and enhances the company’s access to various resources. Consumer attention focuses on a company’s decisions and actions. In particular, consumers may expect companies to behave ethically and to be actively involved in helping society. Consumers then need to be aware of a company’s level of social responsibility. Thus, companies are under pressure to behave in socially responsible ways. Consumer expectations regarding the CSP of the firm are a motivating target and must be considered carefully. Managers must grasp the nature of the values, attitudes, and behaviors of their consumers and accordingly, respond. The attitude would serve as a reference point in consumers’ evaluations of a firm’s involvement in CSP. CSP communicates that the company can be trusted to act as a partner that will respond to social needs. CSP is thus internalized in the consumer dialogue process. Managers have to develop communications that provide information about their social performance. Social interactions are driven toward dialogically motivating and sustaining mutual understanding. People construct the content in a subjective, meaning-​creating process. Sense is a continuous process oriented toward placing current experiences in a frame of reference. Social environments are characterized by a shared understanding among individuals about the nature of a particular issue and the available behavioral alternatives. The mental process in the case of CSP is directed at the creation of a common, context-​bound view. Corporations are increasingly engaging as a form of social involvement in response to increased expectations for companies to contribute to CSP efforts. Companies bring their resources to rebuild trust in the social

Social preferences and moral economy  135

environment. By pursuing social and environmental objectives, socially responsible companies increase their trustworthiness, which in turn supports the process of value creation. The framework presented in this section identifies CSP with the private provision of public goods or the private redistribution of profits to social causes. In order to consider what motivates a business to make a contribution to public goods, the model focuses on manager–​consumer interactions. The model analyzes two dimensions of social preferences: the consumer’s taste for CSP and the manager’s taste for CSP. As a consequence, social preference formation interacts with economic decisions taken by others. The motivation underlying CSP is modeled in terms of social satisfaction that the manager perceives and receives in the interactions. Consumers experience the results of corporate social behavior, so they influence the ethical standards by which corporate behavior is judged and evaluate how well the manager performs according to those standards. CSP communicates that the company can be trusted to act as a partner that will respond to social needs. The collaborative relations can reinforce positive responses with each other. Consumers’ attitudes to encourage CSP are influenced by the manager’s view to practice CSP, and vice versa. CSP is considered as levels of economic, ethical, and discretionary activities of a business entity as adapted to the values, norms, and beliefs of society. Managers operating at the low level of moral conduct do not take socially conscious consumers’ interests into account, and their decisions are perceived to be immoral by consumers. No public good is then provided by firms. A behavioral view identifies an alternative source of inertia as the low level of moral conduct insensitive to the values, norms, and beliefs of society. On the contrary, managers operating at the high level of moral conduct focus on meeting the expectations of socially conscious consumers according to ethical requirements. Then the provision of public goods can be improved substantially. The corporate perspective on CSP and the resulting behavioral change are therefore products of social interactions. Managers who rely on social values facilitate the ethical development to be motivated beyond their self-​ interest.The primary force in determining the level of corporate moral conduct is the resulting social satisfaction in manager–​consumer interactions.The social satisfaction that the manager receives is influenced by the consumers’ expectations regarding CSP. More of public goods can be provided when consumers are more likely to value CSP, and managers are attuned to the consumers’ expectations about CSP.

Notes 1 In an example of a boy cutting the lawn for his father, once the father agrees to pay for the lawn to be cut, the boy is no longer willing to do it without payment. 2 A reward is offered after an action is completed, while an incentive is more appropriately described as a strategy employed to motivate action. 3 Payments may crowd out the prosocial attitude in certain situations where people act to support a common good without being paid to do so. In the environmental arena, for

136  Social preferences and moral economy

example, Frey and Oberholzer-​Gee (1997) found that Swiss residents were willing to accept nuclear waste disposal in their community purely out of a sense of public spiritedness about twice as frequently as when they were offered compensation for accepting the negative externality. 4 In a public good game, players are endowed with a certain number of tokens that they can either contribute to a project that is beneficial for all or keep for themselves. All players profit equally from the public good, no matter whether they contributed or not; each player receives a lower individual profit from the tokens contributed to the public good than from the tokens kept. 5 N is an even number in this analysis. 6 Zelmer (2003) and Chaudhuri (2011) provide surveys on experimental findings concerning the voluntary provision of public goods. 7 Economists have assumed that each player chooses a public good contribution that maximizes his or her utility, given his or her beliefs about levels of contributions by others. Sugden (1985) studies what the levels of contributions imply about individual beliefs about others’ behavior. In order to rationalize observed high levels of voluntary contributions, the beliefs that individuals must hold in order are mutually inconsistent. As Sugden (1985, p. 123) argues, “[i]‌t would hardly be satisfactory to claim to explain voluntary contributions to public goods by assuming individuals hold systematically false beliefs about one another’s behavior.” 8 In the prisoners’ dilemma game, two players simultaneously choose between cooperation and defection. If both decide to cooperate, they both obtain a high outcome; if both defect, they both receive a low outcome; and if one player cooperates and the other defects, the cooperator receives a very low outcome and the defector obtains a very high outcome. Thus, it is better for a player to defect for any given strategy of the opponent. 9 The term ‘corporate citizenship’ is used to connect business activity to broader social accountability and service for mutual benefit, reinforcing the view that a corporation is an entity with status equivalent to a person. 10 Bowen (1953) first pointed out that corporate decision-​making processes have to consider not only the economic dimension, but the social consequences deriving from their business behavior as well. 11 The Green Paper identifies four factors, which lie behind the growing success of CSR concept (p. 4): (1) the new concerns and expectations of citizens, consumers, public authorities, and investors in the context of globalization and large-​ scale industrial change; (2) social criteria, which are increasingly influencing the investment decisions of individuals and institutions both as consumers and as investors; (3) increased concern about the damage caused by economic activity to the environment; (4) transparency of business activities brought about by media and modern information and communication technologies. 12 Brown and Dacin (1997) support the idea that what consumers think about a company does influence their beliefs and attitudes toward the products of that company. They show that a high CSR grade leads to a higher evaluation of the company and corporate evaluation is positively related to product evaluation. In their experiments, subjects were given a description of a fictitious firm, along with a CSR report card with various grades indicating above and below average community involvement. Subjects were asked to rate products made by the firm as well as the firm itself. 13 In the ‘gift exchange game,’ Fehr et al. (1993) design competitive goods market experiments that allow for the emergence of reciprocal interactions. Buyers make a ‘gift’ to the sellers by paying prices above the competitive level. Sellers in turn respond reciprocally by choosing quality levels above what is dictated by their pecuniary interests.

Social preferences and moral economy  137

14 Many societies face the problem of how to provide public goods. Free riders are those selfish individuals who take advantage of the benefits provided by cooperators without contributing to those benefits themselves.When people face strong material incentives to free ride, the self-​interest model predicts that no one will contribute to the public good. However, if there are individual opportunities to punish others, those who cooperate may be willing to punish free riding, even though this is costly for them and even though they cannot expect future benefits from their punishment activities (Fehr and Gächter, 2000). 15 Eco-​labeling is an example of what Baron (2001) calls strategic CSR: attempts to increase profits by attracting ‘green’ consumers.


6.1  Society and coordination problems Interactions of people in society make adaptation possible. At an individual level, actions are based on subjective perception of what exists. However, a correspondence between individual actions and a social order is inherently problematic. People live in a world of expectation about interactions with others’ actions. For a society to exist, rules must be habitually obeyed. A social system can be viewed as a massively complex structure of rules that have evolved over a long period of time. According to Hayek (1973), one of the main characteristics of human behavior consists of following rules of conduct. Humans have a tendency to fit themselves into prescribed behavior in specific circumstances: Homo sapiens are “a rule-​ following animal” (Hayek, 1973, p. 11). The word ‘rules’ is sometimes used broadly to refer to any kind of directive for decision-​making or any behavioral regularity. Rules are behavioral patterns that individuals expect each other to follow. Like prices, rules coordinate and motivate interdependent behavior. The rules of conduct governing our actions are adaptations to our ignorance of the external environment in which we have to act. Because of our ignorance, we ought necessarily to rely largely on rules, instead of attempting to design the system of rules. Relying on rules is a device we have learned to use, because our reason is insufficient to master the details of complex reality. Thus, following rules helps people make decisions with some degree of certainty about which behavior is acceptable and which is not. The knowledge need not be articulable. We can know more than we can tell. Tacit knowledge refers to a nondeductive and nonscientific kind of knowledge. Hayek (1973, p. 99) relates tacit knowledge and practical learning of rules as follows: So long as individuals act in accordance with rules it is not necessary that they be consciously aware of the rules. It is enough they know how to act in DOI: 10.4324/​9780429344817-6

Social incentives and interaction  139

accordance with the rules without knowing that the rules are such and such in articulated terms. Rules of conduct are largely tacit.1 People are not capable of explaining these rules. However, respecting these rules, to a great extent, conditions the coordination of individual actions. Human knowledge is at once practical and abstract. It is embodied in the abstract schemata that compose the mind and manifests itself through the rules that guide individual actions. ‘Knowing how’ corresponds to tacit knowledge, while ‘knowing that’ relates to conscious knowledge. Knowing how consists in using habits and following rules whose nature and definitions do not need to be explained in the individual’s mind. Habit can be understood as a form of social embeddedness of individual agents. By acting on the basis of habits, individuals make their behavior predictable for others. In the face of uncertainty, each agent has to make decisions. Uncertainty brings the rules of social life that individuals rely on to make decisions.The problem of uncertainty provides a starting point from which it becomes possible to understand why human behavior is rule driven. Hayek’s thought largely rests on the concept of spontaneous order. Spontaneous orders in human affairs are patterns that arise as the unintended consequences of individual actions. Individuals must coordinate their actions with those of others. A mutual adjustment process of individual actions makes it possible to realize such an order. According to Hayek (1973, p. 36), the term ‘order’ refers to a state of affairs in which a multiplicity of elements of various kinds are so related to each other that we may learn from our acquaintance with some spatial or temporal part of the whole to form correct expectations concerning the rest, or at least expectations which have a good chance of proving correct. Hayek’s discussions of spontaneous order in social life can be summarized as the following themes (Sugden, 1989): (i) Unintended consequences: Spontaneous orders are the unintended social consequences of individual actions but not the execution of any human design. (ii) Rule following: Spontaneous orders are formed when individuals follow abstract rules of conduct. However, the individuals themselves may not be able to articulate the rules they follow. (iii) General predictability: The ‘general’ properties of a spontaneous order are predictable. The formed patterns have general characteristics that are highly predictable. (iv) Specific unpredictability: The detailed features of a spontaneous order are unpredictable. Not knowing specific circumstances, individuals cannot predict exactly how others will act. The problem is not merely the ‘quantity’ of knowledge that would be required in order to make accurate predictions but also the ‘kind’ of knowledge.

140  Social incentives and interaction

(v) Division of knowledge: Each individual makes use of his or her specific knowledge in deciding how to act. Therefore, spontaneous orders embody a totality of knowledge that is not known to any single planner. A system is explained by both its constituent elements and the connections by which they are related. Emergence implies that the system as a whole is able to perform functions or solve problems that elements cannot achieve separately. Any whole may be said to be a higher level entity. The higher level entities possess emergent properties that arise as a result of interactions between their lower level parts. When certain elements stand in particular relation to one another, the whole has properties that are not possessed by its elements taken in isolation. Emergent properties are irreducible to the properties of the lower level elements of which the higher level whole is formed. Emergent order comes from the processes that require little or no foresight but considerable experimentation and trial and error. Agents organize their rule-​ based structure through the interaction with their external environment. Even though the rules are quite simple, they can produce global patterns that may not be at all obvious.2 A social system can be viewed as a complex structure of rules that have evolved over a long period of time. An equilibrium situation is considered as one in which individual actions are fully coordinated. Individual knowledge, expectations, and the actions based on these, in the equilibrium situation, are mutually consistent. Any change in the relevant knowledge leads individuals to alter their actions, and it disrupts the equilibrium relations between their actions taken before and those taken after the change. The analysis of coordination of individual actions is the main problem for economists supporting the tradition of Austrian economics.3 For Mises ([1949] 1966), “[t]‌he market phenomena are social phenomena. They are the resultant of each individual’s active contribution.” Exchange is undertaken in the expectation of having something in return which renders two individuals tied to each other in a social relation. The study of purposeful human action is, for Mises, the key to understanding the market process. As Mises ([1949] 1966, p. 245) argues, [t]‌he final state of rest is an imaginary construction, not a description of reality. … What makes it necessary to take resource to this imaginary construction is the fact that the market at every instant is moving toward a final state of rest. According to Lachmann (1976), the market process is characterized by inconsistency of individual plans. Inconsistency of individual plans is the direct consequence of the introduction of subjectivism to expectations. Lachmann (1970; 1976) particularly emphasizes the problem of divergent expectations. Lachmann (1970) is interested in the influence of institutions on the formation and revision of individual plans. Institutions are explained as follows: “[t]‌hey [institutions] are … co-​ordinating the actions of millions whom they relieve of the need to acquire and

Social incentives and interaction  141

digest detailed knowledge about others and form detailed expectations about their future action” (Lachmann, 1970, p. 50). There always exists a large variety of beliefs in a social structure. Lachmann (1970) regards people as social beings whose beliefs, expectations, and plans are profoundly shaped by the institutional environment in which they live. Individual actions are guided by intersubjectively shared social structures. This implies that people are able to understand and anticipate how others will act in particular situations, thereby facilitating the formation of reliable expectations and mutually compatible plans. If people have no grounds for believing that a particular set of institutions will endure into the future, they will have no reason to orient their plans toward those institutions. In order to serve their ‘rule-​influenced’ purpose, institutions must display stability. Institutions play a crucial role in enabling people to formulate expectations that are sufficiently accurate for the successful coordination of plans.

6.2  Social norms People frequently face conflicts between individual self-​ interest and collective rationality in the course of social life. An opposition exists between what is best for oneself versus what is best for others. Sociologists have traditionally emphasized informal, interpersonal mechanisms such as relationships, shared values, and social norms. People are strongly influenced by what others do. Social norms play a significant role in behavioral prediction and explanation. The behavior of individuals is often determined by social norms reflecting what is approved or common in society. Social norms are rules of behavior shared by the members of a social group that “specify what actions are regarded by a set of persons as proper and correct, or improper and incorrect” (Coleman, 1990, p. 243). Social norms specify a limited range of behavior that is acceptable in a situation, facilitating confidence in the course of action. Norms can impact our judgments. For example, according to Cialdini et al. (1990), people are more likely not to litter when the floor is nearly spotless in public places. When the apparent social norm is not to litter, people are more likely to conform to that norm by refraining from littering. Norms describe the uniformity of behavior that characterizes a group. The more people adhere to a norm in a group, the more likely its adoption for an individual is. Using norms as cues for behavioral change is usually based on telling people what others are doing in a similar situation. The concept of social norms is important to the explanation of cooperation and prosocial behavior. Reciprocity can be seen as an example of a social norm. In a situation where people are motivated by social norms, their willingness to contribute to good social causes increases with their perception of the contribution of others. People become more cooperative in a group decision situation, compared to when acting as individual decision-​makers. A norm is constructed as the tendency of the distribution of a certain attribute in a group. Norms thus come into existence as a product of group interactions. How these interactions take shape depends on the specific institutions. A subsystem

142  Social incentives and interaction

of cultural institutions and social processes is responsible for system-​wide norm promulgation. Norm conflict is then considered as a disequilibrium phenomenon rather than a normal state of affairs. A process of socialization leads to the internalization of norms. Habit and tradition become ‘institutionalized’ behavior more or less in accordance with the norms of prior generations, which is again passed over to the next generations. Norms provide the content of social roles, create common expectations, and facilitate social cooperation. Different norms provide the contours of different groups. Norms serve to guide an individual’s behavior, and they are thus an attractive method of social control. Examples range from table manners to business practices. A rule against poor table manners is not suitable for embodiment in law. A norm specifies a behavior that is seen as desirable or undesirable in the shared view of group members. For example, according to Ostrom (2000), social norms are defined as “shared understandings about actions that are obligatory, permitted or forbidden” (pp. 143–​144). It is necessary that people believe a social norm exists and know the class of situations to which the norm pertains. This allows that conflicting norms may exist simultaneously. People must be aware that they are in a situation where a particular norm applies. In social psychology, norms consist of two major categories: they are both descriptive (‘is’ statements) and injunctive (‘should’ statements) (Cialdini and Trost, 1998). Descriptive norms describe what people generally do in a situation, while injunctive norms describe what people ought to do there. Descriptive norms are simply regularities of behavior. On the other hand, injunctive norms are considered to be behavioral expectations that are backed by social or material sanctions. It is considered a belief system about what constitutes morally approved and disapproved conduct.Work on how norms impact cooperation has focused on injunctive norms. Norms broadly encompass conventions. Behavioral regularities that lack normative attitudes are referred to as conventions (McAdams and Rasmusen, 2007). Conventions are thought of as descriptive norms that are simply what people do. A specific convention is an actual behavioral regularity in a given group. It implies the behavioral pattern, actually followed by the group members in a recurrent situation of social interaction. In game theory, a social norm is seen essentially as a convention. Without the normative obligation, the behavioral regularity would be simply an equilibrium that results when each player takes his or her best step considering the actions of others. Norms emerge not from a collective need but from the decentralized interaction of players according to their own interests.A convention is a stable solution to a recurrent coordination problem; it is arbitrary in the sense of being the realization of only one of multiple equilibria that a coordination game displays.There must be at least one alternative norm which could have served more or less equally well as a solution to the coordination problem. There are multiple equilibria—​two or more outcomes that satisfy the Nash criterion that no player would benefit by unilaterally switching strategies. The notion of convention

Social incentives and interaction  143

is thus characterized in terms of coordination equilibrium, that is, a strategy combination in which no one would have been better off had any one player alone acted otherwise. Every player strictly prefers that all conform to the equilibrium. Pure coordination games are games with a perfect coincidence of interests, having at least two Nash equilibria with equal payoffs. A pure coordination game is commonly illustrated by the choice between driving on the left and right sides of the road.There are two pure strategy Nash equilibria with identical payoffs—​where everyone drives on the left and where everyone drives on the right. All players share some interests in avoiding the noncoordinated outcome—​where some drive on the left and others on the right. It is costly to try to change the convention once it is established. The convention is stable because no one has an interest in deviating from it. Nash equilibria then correspond to population equilibria of players who interact in accordance with the rules of the game. In an evolutionary explanation of a trait, its frequency in a population is used in terms of selection pressures that favor (or disfavor) its replication relative to its alternatives in a particular environment.The primary determinant of the payoff of any particular trait is the frequency distribution of itself and its alternatives in the population. In successive rounds of the game, players are assumed to adjust their social traits in response to the payoffs from expressing those traits in the previous round. When an evolutionary model endogenizes strategies, it explains the frequency distribution of these strategies. Boyer and Orléan (1992) consider how conformity effects can lead to the emergence of a group consensus around one convention. Let us consider the symmetrical game defined by the payoff matrix as given in Figure 6.1. It is assumed that there is a large population from which pairs of players are repeatedly drawn at random to play this symmetric game. Players do not understand that their own behavior potentially affects the future play of their opponents. Let p(t) be the frequency of players who play strategy A at time t in the population. Then, a player who chooses strategy A will obtain the utility U(A, p(t)), given by the following formula:



U A, p (t ) = p (t )UA. Likewise, for the frequency of players having chosen strategy B at time t in the population, 1 − p(t ), a player who chooses strategy B will obtain the following utility U(B, p(t)): A

FIGURE 6.1  Payoff












144  Social incentives and interaction Utility



O FIGURE 6.2 Two






) (


U B, p (t ) = 1 − p (t ) UB. Two utilities, U(A, p(t)) and U(B, p(t)), are described in Figure 6.2. There exists a value p* such that U(A, p*) equals to U(B, p*): p* =


For p > p*, U(A, p) is greater than U(B, p), even if UA is smaller than UB. Successful behavior becomes more prevalent just because players imitate successful behavior. The frequency p(t) increases if U(A, p(t)) is greater than U(B, p(t)), which can be formulated in the following way: dp (t ) dt





= Z U A, p (t ) − U B, p (t )  ,

where Z is a nondecreasing, sign-​preserving function. This is a differential equation that, together with an initial condition, determines a path for the population that describes the state of the population for any time t. Thus, for the sufficient number of players having chosen strategy A, its representation in the population grows (dp(t)/​dt > 0), and the system eventually converges

Social incentives and interaction  145

to the convention A which is Pareto inefficient.That is, for p(t) > p*, the frequency p(t) converges to 1. If no fraction of the population plays strategy B at any point of time, then it is never played. An equilibrium can be viewed as the steady state of a community whose members are myopically groping toward maximizing behavior. In game theory, an equilibrium is a profile of strategies, one for each player participating in a strategic interaction. What distinguishes an equilibrium from other profiles is that no player can do better by changing his or her strategy unilaterally. If others do their part in the equilibrium, no player has an incentive to deviate. When a game has only one equilibrium, Nash equilibrium must be the only rational prediction of players’ behavior. However, a game can have multiple Nash equilibria. What more than one equilibrium are possible, a player’s choice of strategy is not fully determined by the payoffs. In real life, individuals are skilled at coordinating their actions. Schelling (1960) develops the study of mechanisms allowing people to coordinate. What is necessary is to coordinate predictions: People can often concert their intentions and expectations with others if each knows that the other is trying to do the same. Most situations—​perhaps every situation for people who are practiced at this kind of game—​provide some clue for coordinating behavior, some focal point for each person’s expectation of what the other expects him to expect to be expected to do. Schelling, 1960, p. 57, original emphasis According to Schelling (1960), when the problem is selecting one means of coordinating among many, certain solutions stand out from the others as the sort that will attract the attention. “Some kind of prominence or conspicuousness” (Schelling, 1960, p. 57) or “uniqueness” (p. 58) of a coordination equilibrium is ‘focal,’ in the sense that it creates convergent expectations.4 This is a property that captures the attention of players as a result of common perceptions, impressions, and associations. Players tend to select a unique equilibrium in some non-​payoff dimension, merely because that uniqueness causes each player to expect every player to focus on it. In a coordination game, all players are better off as common effort increases. Schelling’s (1960) focal-​point idea is important as an explanation of how players coordinate. Each member of the population expects every other member to behave in accordance with the relevant regularity. Individuals often coordinate at a point that, in some sense,‘stand out’ from the others. In explaining the concept of salience, Schelling (1960) describes players as searching for a ‘key,’ ‘signal,’ or ‘suggestion’ that is hidden in their decision problems, Players tend to select an equilibrium that is ‘unique,’ merely because that uniqueness causes each player to expect each player to focus on it. A shared vision of what should be obvious to each player leads to the emergence of a focal point. Schelling (1960) contends that coordination is inherently dependent on empirical evidence.There is a logic to decide what solution has the unique property. But the uniqueness often depends on the contingent associations of the individuals involved.

146  Social incentives and interaction

Lewis (1969) gives a more formal structure to characterize coordination problems. Individuals’ interests roughly coincide in pure coordination games. In order to have a sufficient reason for choosing a particular action, an agent needs to have a sufficient degree of beliefs that the other agent will choose a certain action. When individuals interact with each other, their actions and the deliberate reasoning by which they choose their actions are interdependent. According to Lewis’ (1969) argument, coordination may be achieved with the aid of mutual expectations about action. Past experience of a convention, or ‘precedent,’ is the source of such expectations. A specific convention is assumed to be arbitrary, in the sense of being the realization of only one of the multiple potential regularities that could emerge in a recurrent situation of social interaction.5 In the driving game, the drivers condition their choices on the history of play. That is, if everyone has been driving on the right side of the road up until now, everyone will continue to drive on the right; if everyone has been driving on the left, everyone will keep doing the same. A certain action stands out as the one that most likely everyone will pick. Lewis (1969) justifies the stability of conventions through the reconstruction of the actual processes of reasoning of those involved in the interaction where a particular convention emerges. A specific convention is an actual regularity in the behavior of a given population. The stability of conventions is justified through the reconstruction of the actual processes of reasoning of those involved in the interaction situation where a particular convention emerges. When people interact with each other, their actions and the deliberate reasoning by which they choose their actions are interdependent. In Lewis (1969), the definition of convention requires not only that agents engaged in a game play a coordination equilibrium but also that they have common knowledge that they all prefer to conform with the equilibrium, given that every agent conforms with this equilibrium. Following Lewis (1969), a proposition P is common knowledge for a set of agents if and only if: 1. each agent k knows P. 2. each agent i knows that each agent k knows that P, each agent j knows that each agent i knows that each agent k knows that P, and so on. For Lewis (1969), a coordination equilibrium is a convention only if the agents have common knowledge of a ‘mutual expectations criterion’ (MEC): each agent has a decisive reason to conform to his or her part of the convention, given that he or she expects the other agents conform to their parts. In general, in order to have a sufficient reason for choosing a particular action, an agent needs to have a sufficient degree of belief that the others will choose certain actions.6 Such a sufficient degree of belief is justified by a system of mutual expectations. Lewis (1969) adds this common knowledge requirement to the definition of convention to rule out cases in which agents coordinate as the result of false beliefs regarding their opponents. If the members of a group coordinate successfully, they will start to notice a regularity. The members develop a preference for conforming to this regularity if

Social incentives and interaction  147

others do so as well. A general expectation that the members will conform to the regularity is formed. Consequently, a stable pattern of behavior, or a convention, emerges based on the general expectations of such members about the way that others will behave. Suppose that the members of a population face some coordination problem in a recurrent situation where there is a history of conformity with a behavioral regularity R. Everyone prefers to conform with R if almost everybody else does the same. Then, R is a population-​wide regularity in behavior by which the members reliably pick a coordination equilibrium. Such a regularity R is a convention if it in fact occurs. Conformity to the prevailing convention is not explained solely from the premises of classical game theory, due to the Nash equilibrium selection inherent in any coordination problem. Lewis’ (1969) theory of conventions provides a way to allow external information to play a role in reasoning about one’s choice of strategy. To understand conventions, we must consider how they are mean to figure in the practical reasoning of people. Their appreciation of the practical force of conventions depends on their expectation, or implicit understanding, that the conventions are practiced in their community. Lewis (1969) adds an account of the mechanism by which conformity is maintained. Salience plays a crucial role in the emergence and reproduction of conventions. Individuals try for a coordination equilibrium that is salient: one that stands out from the rest by its uniqueness in some conspicuous respect. Conventions are understood as salient regularities of behavior. Salience constitutes a focal point that facilitates coordination when purely rational considerations are insufficient to identify the best move. Focal points may be determined by cultural, cognitive, or even biological factors. Then, Lewis’ account exposes the limitations of purely rational consideration for the analysis of collective behavior. Salience ultimately comes down to a ‘non-​rational’ (but not ‘irrational’) propensity to choose in certain ways when reason gives out (Sugden, 1995). In order to understand how conventions emerge from individual interactions, we have to consider how a focal point facilitates coordination. In case of conventions, salience is determined by past conformity (salience by precedent). Lewis (1969) argues that the precedent would make one of the coordination equilibria salient. Each individual has some tendency to repeat actions that have proved to be successful in previous interactions. The precedent will make one of the coordination equilibria salient. It is because a regularity of behavior already exists that there is a precedent to follow. Each individual’s expectation that others will do their parts, strengthened perhaps by replication using his or her higher-​order expectations, gives him or her some reason to do his or her own part. When players have no reason to choose one strategy rather than another, their default choices tend to favor strategies with certain properties of salience. Knowing this, rational players of coordination games choose salient strategies deliberately. Salience can result from players correlating their strategies and their expectations with certain events external to the game. This particular convention is conceived as a ‘correlated equilibrium’ of coordination games with multiple equilibria. In a correlated equilibrium, players tie their strategies to various ‘state of the world’ or

148  Social incentives and interaction

other pieces of information which are not formally part of the game.7 Correlated equilibrium allows players’ actions to be statistically dependent on some random signals external to the model (Aumann, 1974; 1987). In a correlated equilibrium, individual actions are statistically dependent on an event or signal sent by a correlation device. One interpretation of correlated equilibrium is that a correlation device (here, a convention) instructs individuals to take actions, sending signals to them regarding what they have to do in a given situation. A convention is a signaling mechanism that the individuals can use to coordinate their actions in social interaction.This suggests that correlation of actions in the interaction is not unrealistic. While the correlation required by predictable behavior may be asking too much, it is possible to obtain correlated equilibrium in the interaction. Correlated equilibrium only requires rationality and common priors, while Nash equilibrium requires stronger premises. Different players can take different actions potentially, but their actions are conditional on some external signals. ‘Nature’ first gives a publicly observable signal. Players’ strategies assign an action to every possible observation. If no player has an incentive to deviate from the recommended strategy, the strategy distribution is a correlated equilibrium. In general, social norms regulate the strategic interaction of social actors. If individuals play a game with several Nash equilibria, social norms serve to choose the most socially desirable among these equilibria. Social norms are considered to play the role of a ‘choreographer’ who leads people to take actions according to some commonly known probability distribution (Gintis, 2009). A social norm prescribes which strategy every player should choose depending on the observed signal, in such a way that it is rational for him or her to follow the suggestion, knowing that others are following it too. A game is a description of a strategic interaction involving two or more players. Formally, a game G is defined as a triple (N, S, u) where: N = (1, 2, …, n) is the set of the players of the game; S = S1 × …×Sn is the Cartesian product of the set Si of available pure strategies for each player i; and u : S →  is the payoff function mapping S into a n-tuple of real numbers for each strategy combination in S. The triple (N, S, u) allows us to specify any strategic interaction. It is assumed that each player knows everything about the structure of the game. Each player chooses a strategy to maximize his or her payoff, given his or her beliefs about the strategies used by the other players. The idea of Nash equilibrium is that if the players choose strategies that are best responses to each other, then no player has an incentive to deviate to an alternative strategy. The system is thus in a kind of equilibrium, with no force pushing it toward a different outcome. A correlated equilibrium can be considered as a Nash equilibrium in an ‘extended game’ where a choreographer is a player. The choices can be connected in some way: if players relate their choice of strategies with some jointly observable feature

Social incentives and interaction  149

of their interaction, then they can achieve a correlated equilibrium. A choreographer instructs individuals to take actions according to some commonly known probability distribution.The choreographer is then considered as a social norm that assigns roles to individuals in society. For any game (N, S, u), define Ω as the set of all possible worlds, and ω ∈ Ω as the state of the world that actually obtains. Each player i possess an information partition Hi on the states of the world. Then, Hi can be thought as player i’s private information. Define Ψ = ∩i H i as the ‘common information partition,’ that is, the intersection of the information partitions of all players. Ψω means that, at ω , each player knows which cell in Ψ obtains. An event P is ‘public’ when Ψω ⊆ P since, for any ω ∈ Ψω , everyone knows P, everyone knows that, and so on. Define a function x: Ω → S as a system of strategies correlated with the state of the world, with x( ω ) = ( x1 ( ω ),…, xn ( ω )) ∈ S . Finally, define (Ω,q) as the probability space over all possible states of the world, in which q is a probability distribution over Ω. A strategy profile x * ( ω ) = ( s1* ,…, sn* ) forms a correlated equilibrium if, for each player i, for each state of the world ω , and given the probability space (Ω,q):





Ei ui x * ( ω ) |H i  ( ω ) > Ei ui si ; x * ( ω ) |H i  ( ω ) .     Given the conjectures conditional on his or her information partition and the state of the world ω , each player should have no interest to deviate from the strategy prescribed by x * ( ω ). Any norms can be interpreted as a correlated equilibrium by linking an institutional element with the occurrence of a public event P. As indicated above, a public event is an event such that it refers to a world ω that belongs to the common information partition of the players. Therefore, P is common knowledge. A public event P is governed by an institutional element if, whenever P occurs, it is common knowledge that everyone follows the correlated equilibrium. Correlated equilibria define parts for group members to play in schemes that focus the expectations and correlate the choices of players who find themselves in norm-​ governed situations of social interaction. There is a scheme of actions such that it is common knowledge in the community that most members expect most other ones to do their parts in the scheme. A correlated equilibrium can capture the idea that players use the same strategy, while potentially performing difference actions. An obvious solution in such circumstances is to adopt a correlation device. Whatever the nature of the correlated device, the effect which is captured by the correlated equilibrium operates through the formation of the players’ beliefs. A correlated equilibrium can be interpreted as consistent with a social norm. On the assumption that other players follow the signal, no one is served better by acting differently.This implies the set of actions is a correlated equilibrium. Social norms affect individual actions through providing guidelines as to what is normal. Some behavioral interventions aim to influence actions through increasing awareness of social norms. In order to correct for free riding, policies should take into account social norms which may help to overcome the under-​provision of

150  Social incentives and interaction

public goods. Individuals tend to be motivated partly by others’ perceptions. They try to signal traits defined as good based on the community’s norms. Being altruistic is often seen as good, but being greedy or selfish is not. Image concerns are an important motivation for contributing to public goods. The image motivation depends on how much people care for their reputation. People may be more likely to provide public goods if their contributions are publicly acknowledged. Information about reputation would be a key factor in initiating and sustaining cooperation in society. People share reputational information about those with whom they have interacted. Positive reputations, built on a history of cooperative interactions, provide useful information about partners’ trustworthiness. People will act more pro-​socially in the public sphere than in the private one. A community makes reputational judgments of a person’s cooperativeness or generosity. People derive intrinsic value from a self-​image desire. For example, in a model by Bénabou and Tirole (2006), a ‘reputational payoff ’ from contribution to a public good is a function of the belief that others have regarding the consumer type. The reputational rewards that go to those who behave in cooperative or generous ways can be viewed as a powerful force shaping prosocial action. Consistent with the benefits that accrue to those with a prosocial reputation, prosocial behavior will increase in a situation where reputational rewards are possible. The individual’s intrinsic incentive to be prosocial increases as the share of the population acting that way increases. The prospect of reputational gain serves to promote prosocial actions. Individuals thus have an incentive to acquire a good reputation for keeping promises and performing acts that appear trustworthy. The tendency for individuals to comply with social norms strengthens a good reputation when reciprocity is indirect. A relatively small group is better able to sustain cooperation. If a person tries to free ride in a small group, it is easier to identify and punish the culprit. Reciprocity could be sustained if a group is not too large and the membership is not fluid. On the other hand, Olson (1965) argues that large groups could not organize themselves voluntarily for coordinated and cooperative action, even though they have good reason to do so. The more people who have to share a collective benefit, the less that each individual can singularly gain, making the return on their cooperation meaningless to them. Thus, any rational, self-​interested agent would choose to free ride on large group endeavors, and effort within the group would be insufficient to produce the collective good. People judge the appropriateness of their behavior based, in part, on the behavior of others. Providing households with the consumption information of their peers, as an indicator of social norms, may be successful in eliciting behavioral change. According to Allcott (2011), consumption of electricity can be reduced by informing people about their power use compared to their neighbors’ and to that of ‘efficient users.’ Increasing energy efficiency can play a significant role in reducing overall energy consumption and associated emissions. Advising consumers that people similar to them are using less energy will likely motivate them to conform to these energy-​saving norms and reduce their consumption accordingly. Social

Social incentives and interaction  151

norms could be used to present information to people about decisions made by others like them in a similar situation. People tend to follow social norms if they are not very familiar with the circumstances of decisions. They perceive decisions presented as norms as likely to be effective and adaptive responses. For example, Goldstein et al. (2008) show that a social comparison treatment that presents people staying in hotels with a norm such as ‘the majority of people reused their towels’ has a larger effect on their behavior than standard messages about the benefits of reusing towels.

6.3  Norm compliance Social norms often direct individuals to undertake actions that are inconsistent with selfish actions. For example, in the dictator game, 50-​50 division is generally viewed as norm compliant (Andreoni and Bernheim, 2009).8 People may deviate from such norms. In the case of legal compliance, individual incentives most often refer to deterrence (Becker, 1968). That is, individuals are deterred from criminal activities by a higher fine and by a higher probability of conviction. Unlike legal rules, social norms are not supported by formal sanctions.Why do social norms not simply collapse from the violation? This section studies two distinct mechanisms on norm compliance. The incentive to comply with norms derives not only from the enforcement of costly punishment by others but also from reputation building for oneself. The importance of decentralized punishments (i.e., punishments carried out by individuals without the intervention of a central authority) is documented in experimental studies. Since subjects knew that they would interact with their group members again, they could use punishment to discourage free riding, anticipating greater cooperation and greater payoffs in subsequent rounds. The fear of punishment has a positive effect on cooperation. In public goods experiments, subjects begin by contributing on average about half of their endowments to the public account. However, the level of contribution decays over the course of multiple rounds (Andreoni, 1995). When costly punishment is permitted, cooperation does not deteriorate. Fehr and Gächter (2000) indicate that many individuals are willing to punish unfair behavior at a personal cost in public goods games. Costly punishment is administered by ‘third parties’ (Fehr and Fischbacher, 2004). Potential punishers are not themselves the victims but have merely witnessed unfair behavior. This is called ‘altruistic punishment’ as individuals sacrifice for no direct benefits (Gintis et al., 2003). It suggests that cooperation has evolved through the sacrifice of altruistic punishers who are ready to incur some costs to prevent unfair behavior. People are fair because they have a psychological motivation to restore fairness. Then, punishment can be seen as a consequence of a sense of fairness. Certain groups of individuals can maintain a strong reputation over time. Akerlof (1980) develops an economic model to show that social norms that involve pecuniary disadvantage to individuals may persist without erosion.9 Disobedience to the norm may involve a loss of reputation. People want to achieve the reputation of

152  Social incentives and interaction

being fair. People are fair because they care about their reputation. They may not be genuinely fair. They have to be rewarded for good reputation, and they have to be willing to comply with the norm. Individuals are influenced in their convictions by what they think others will do.10 Conformity to the norm is conditional on expectations about others’ behavior. Norms are constituted by expectations shared by group members in a population and are jointly recognized among them.11 Thus, we have a set of solutions to the problem of norm compliance. The first solution is the ‘punishment-​based’ account. Following this account, people comply with norms because the threat of punishment makes it in their interest to do so. Altruistic punishment seems to have a solid foundation in human interaction. However, such a costly punishment leads to a large increase in losses for altruistic punishers.The second solution is, on the other hand, the ‘reputation-​based’ account. Social norms can be sustained if the pecuniary advantage from breaking norms is not sufficient to offset the forgone reputation effect. This is related to indirect reciprocity. According to Alexander (1987), indirect reciprocity is arranged in the form of chain: a person is eventually helped by someone else who may not have been directly helped by him or her.12 Altruistic actions can be sustained if people who support others receive support in return. To achieve such indirect reciprocity, building up a positive reputation is needed.13 This section analyzes the interaction between the potential for costly punishment and building personal reputation. The model considers two groups of agents in a society with one norm. Agents in one group (say, group i) choose whether to comply with the norm by incurring some cost. They acquire utility from the reputation derived from complying with the norm. This utility depends positively on the proportion of motivated compliers. Individuals may differ in their motivation to comply with the norm. Punishment will be imposed on individuals who deviate from the norm. In the other group (say, group j), there are agents who value compliance and potentially punish noncompliance (i.e., the sanctioning individuals). The model investigates individual punishment decisions. Agents choose to punish violators at some cost (decentralized punishment). This section asks how individual values evolve endogenously over time and analyzes the long-​run dynamics of norm formation. The present framework systematically investigates the different forces to account for the long-​run stability of the norm. There are two scenarios as follows: In one scenario, there is some possibility that the erosion of the reputation effect induces individuals to break the norm. Then, the norm is enforced due to a higher level of punishment of noncompliance. Punishment would be used to enforce the norm if a substantial fraction of people has little reputation-​derived utility by obeying the norm. This section, however, suggests that altruistic punishment may play a limited role in sustaining the norm. In another scenario, everyone is motivated due to reputation formation, despite a lower level of punishment by others. For a lower level of punishment, effective reputation building provides a way to sustain the norm. Following Teraji (2013), we consider a society populated by a continuum of agents at each period of time t. The population size is constant over time and

Social incentives and interaction  153

normalized to 1. We assume that the population is composed of two groups, i and j, which differ in characteristics. The first group i amounts to a share 1/​2 of the whole population. Accordingly, the half part of the population belongs to group j. Matching between two groups takes place randomly. A member of group i randomly matches with the opponent of group j at each period. This may suggest a large society in which one-​shot encounters with unrelated strangers are common and information is rarely transparent. Suppose that there is only one norm in the society for simplicity. We assume that it is only possible to either follow the norm or not. An agent of group i (‘he’) chooses an action x ∈ {0, 1} at each period. His action is represented by a discrete variable, one or zero. That is, x = 1 if the agent of group i complies with the norm; x = 0 if he violates it. Inertia is introduced with the assumption that every agent of group i cannot switch actions at each point in time. He must make a commitment to a particular action in the short run. Opportunities to switch actions arrive randomly. Some fraction α, 0 < α < 1, of individuals is drawn randomly from group i and makes a new choice of either x = 1 or x = 0. Thus, we may interpret a norm as a prescription indicating how a person ought to behave at any situation at which he may be called to move. An agent in group i may deviate from the norm, but this deviation is costly. Punishment (such as ostracism) will be imposed on group i members who deviate from the norm. If deviation from the norm is observed (x = 0), the opponent of group j (‘she’) decides whether to punish the deviator, that is, chooses p on the closed interval [0, 1]. Here, with probability p, the agent of group j punishes noncompliance, and with probability 1 -​ p, she does not. The agent of group j is only an outside party who happens to know that norm violation has occurred. Altruistic punishment is motivated to restore norm compliance even though they are not expected to interact again in the future. It is costly for agents of group j to punish norm violators. Thus, in the model, agents in group i choose whether to comply with the norm, while agents in group j value compliance and potentially punish violators. Punishment is then confined to interactions with others who share the same norm in the population. Individuals in group i are assumed to be heterogeneous with respect to their social concerns. We denote an agent’s type of group i as g. An agent of type g = 0 is not concerned with the social meaning of a certain action. A higher g implies higher social concerns. The distribution of g is assumed to be exogenous and uniform in the model. Let the uniform distribution of g be F(g), with g ∈ [0, G]. Furthermore, let the density of the uniform distribution be f = 1/​G. In group i, the short-​run utility function of an agent of type g is given by: U = x{R(g, µ) -​D} -​(1 -​x) p C. (6.1) Each individual in group i is assumed to maximize the utility function (6.1), which constitutes the short-​run equilibrium. The first term of the utility function (6.1) reflects the reputation value of a norm if the agent of group i chooses to comply

154  Social incentives and interaction

with the norm (x = 1). It depends positively on the agent’s type g and the fraction µ of individuals who comply with the norm in group i. The agent internalizes the norm with the reputation value R. The strength of the norm is determined by the proportion of agents who follow it. Group i consists of µ compliers and (1/​2 -​ µ) violators, where 0 < µ < 1/​2. The fraction µ is given in the short run and is known by all members of group i. A person who deviates from the norm forgoes reputation-​derived utility. The second term reflects a fixed cost, D > 0, of compliance if the agent of group i chooses x = 1.The final term reflects a fixed cost, C > 0, as the penalty for violation if the agent of group i chooses not to comply with the norm (x = 0) and the opponent of group j chooses to punish him with probability p. The punisher may disapprove of the violator, which reduces his level of satisfaction and his well-​being.14 To sum up, the agent of group i has to pay a compliance cost, D, if x = 1 and the penalty for violation, C, if (1 -​ x) p > 0. It is assumed that D > C. The reputation value R is assumed to take the following form: R(g, µ) = g µ. (6.2) Absent social concerns (g = 0), the agent of group i would not follow the norm. An agent in group i will consider following the norm when the value of his action is greater than the cost of violation. The higher g and the higher µ (the proportion of norm followers) are, the greater the utility from following the norm. The value of following it may take the form of abstract reward, such as social approval from others. The reputation value can also be interpreted as the value of social image. People care about how others perceive them, and these concerns influence a wide range of decisions (Andreoni and Bernheim, 2009). The model also encompasses individual punishers who are not direct victims for norm violation. They punish violators to restore norm compliance. For an agent of group j, whether her opponent in group i follows the norm or not is a random variable. Let the probability that her opponent in group i chooses to comply with the norm be Pr(x = 1). The short-​run utility function of an individual in group j is given by: V = Pr(x = 1) B -​{1 -​Pr(x = 1)} p T. (6.3) Each individual in group j is assumed to maximize the utility function (6.3), which constitutes the short-​run equilibrium. The first term of the utility function (6.3) reflects a psychological benefit, B > 0, received by the agent of group j when her opponent from group i chooses to follow the norm with probability Pr(x = 1). The second term is a cost T (> B) associated with punishment (p > 0) when her opponent from group i violates the norm with probability 1 -​ Pr(x = 1). Punishment can be psychologically costly. Punishers suffer a psychological cost if their opponents violate the norm. This psychological cost includes the punisher’s anger (negative emotion) for violation of the norm.

Social incentives and interaction  155

The benefit–​cost ratio b for the agent of group j is given by: b=

B (6.4) T

As the punisher’s anger (T) increases, the benefit–​cost ratio (b) decreases. The reputation value of the norm affects individual’s short-​r un utility as defined in (6.1). Individuals of group i choose to comply with the norm if the utility from so doing is as great as the utility derived from not complying. The reputation value of the norm depends on the proportion of compliers in group i. At each period, the fraction (1/​2)α of individuals in group i has a chance of making a new choice. In the aggregate, the proportion of compliers is (1/​2)αPr(x = 1), and the proportion of violators is (1/​2)α{1 -​ Pr(x = 1)}. Here, α parameterizes the inertia of the adjustment process. In the adjustment process, it takes some time until individuals find out how many others follow the norm. The proportion µ will increase (or decrease) if (1/​2)αPr(x = 1) is larger (or smaller) than (1/​2)α {1 -​Pr(x = 1)}in group i. Therefore, the long-​run adjustment process is described by the following form: dµ 1 1  = α  Pr ( x = 1) − 1 − Pr ( x = 1)  dt 2 2  



From the above equation, we have: dµ 1  = α Pr ( x = 1) −  (6.5) dt 2  Thus, the speed of norm evolution is influenced by the rate α at which the group members acquire the knowledge necessary to appreciate the norm. If Pr(x = 1) differs from 1/​2, the proportion µ will increase or decrease at rate α until µ is equal to 0 or 1/​2. The relationship between short run and long run is summarized as follows: At each period t, the proportion of norm followers, µ, is given, and agents of group i decide whether to follow the norm or not, given this proportion (short run).The proportion µ is formed over a period of time, and it converges to a steady state (long run). An agent, who is drawn randomly from group i, chooses x ∈ {0, 1} to maximize his utility function: U = x (g µ -​D) -​(1 -​x) p C, from (6.1) and (6.2). Maximization of this utility function constitutes the short-​run equilibrium. If g µ -​D > -​p C, he chooses to comply with the norm (x = 1); otherwise, he does not (x = 0). Thus, his best-​response function is

156  Social incentives and interaction

D − pC µ D − pC if gµ − D + pC ≤ 0, or g ≤ µ

 1  x= 0 

if gµ − D + pC > 0, or g >

It is useful to consider the probability of heterogeneous individuals choosing to comply with the norm in group i, Pr(x = 1). Then:  D − pC  Pr ( x = 1) = Pr  < g  µ   D − pC  = 1− F  .  µ  Using the density of the uniform distribution, f = 1/​G, we have: Pr ( x = 1) = 1 −

D − pC (6.6) f. µ

Thus, in the short run, the probability to comply with the norm, Pr(x = 1), is a continuous increasing function in the probability of punishment, p. Only in the long run, the incentive to comply with the norm adapts to the fraction of norm followers, µ. The strength of the norm is endogenously determined by the collective behavior of the group members. An agent in group j optimally chooses to punish the norm deviator in group i with the probability of punishment, p. That is, at each period, the agent of group j chooses p ∈ [0, 1] to maximize her utility function:  D − pC V = 1 − µ 

 D − pC  fB−  µ 

 f  pT 

from (6.3) and (6.6). This maximization also constitutes the short-​run equilibrium. The first-​order condition is: ∂V C D − pC pC = fB − fT + fT = 0, ∂p µ µ µ which, for an interior solution, equates the marginal cost of punishment to the marginal benefit of punishment. Using the benefit–​cost ratio b in (6.4), we have:  D − bC  p = min  , 1 (6.7)  2C 

Social incentives and interaction  157

For (D -​bC)/​2C < 1 in condition (6.7), it follows that: ∂p 1 = − < 0. ∂b 2 A lower benefit–​cost ratio (b) implies a higher probability of punishment (p). In Figure 6.3, if b1 > b2, then p1 < p2. As the punisher’s angry for the norm deviation increases (or T increases), the ratio b decreases (see (6.4)) and the probability p increases. Once a norm deviation is observed, an agent of group j feels anger toward the norm deviator and attempts to reduce his payoff.Thus, anger plays an important role in punishment decisions. The proportion of individuals who comply with the norm changes over time. From (6.5) and (6.6), the long-​run adjustment process is described by:  1 D − pC  (6.8) dµ f . = α − dt µ 2  Using (6.8), a critical value of µ, or µ*, is defined as follows: 1  µ * = min 2 ( D − pC ) f ,  2 


Condition (6.9) describes the critical value µ* as a decreasing function of the probability of punishment, p, for 2(D -​pC)f < 1/​2.Then, the value µ* in condition (6.9) does not include the benefit–​cost ratio b. p




b b1

FIGURE 6.3 The

relationship between b and p


158  Social incentives and interaction p


µ* 1/2

FIGURE 6.4  Increase

in p

The adjustment process is not necessarily continuous. The proportion µ (> 0) is decreasing for µ < µ*. Then, the adjustment process converges to µ = 0 (everyone in group i deviates from the norm). On the other hand, if µ* is less than 1/​2, µ is increasing for µ > µ*.Then, the adjustment process converges to µ = 1/​2 (everyone in group i follows the norm). Thus, there are two stable equilibrium points. In an (µ*, p) diagram shown in Figure 6.4, there is one intersection between conditions (6.7) and (6.9). Condition (6.7) describes the probability of punishment (p) for a given benefit–​cost ratio (b), which is a straight line; while condition (6.9) describes the critical value µ* as a function of p, which is a declining line. Figure 6.4 also describes the effect of a higher probability of punishment (p) due to a lower benefit–​cost ratio (b). A lower b (or a higher T) implies a higher p in Figure 6.3. In Figure 6.4, a higher p implies that the intersection point between the two lines shifts to the left, which implies a smaller critical level of µ* (< 1/​2). When µ is above the critical value µ*, the adjustment process converges to µ = 1/​2. Anticipating a higher probability of punishment, more agents of group i consider following the norm. On the other hand, Figure 6.5 describes the effect of a lower probability of punishment (p) due to a higher benefit–​cost ratio (b). A higher b (or a lower T) implies a lower p in Figure 6.3. In Figure 6.5, a lower p implies that the intersection point between the two lines shifts to the right, which implies a larger critical level of µ*. If the fraction of norm followers, which is above a larger µ*, is larger, the adjustment process converges to µ = 1/​2. For a higher reputation value, more agents of group i consider following the norm. People conform to maintain their reputation. The strength of the norm is determined by the proportion of the members who

Social incentives and interaction  159 p


µ* 1/2

FIGURE 6.5  Decrease

in p

follow it. Even though p is small, individuals choose to comply with the norm. If the norm is initially widespread, the equilibrium converges to one steady state in which everyone complies with the norm. If instead the initial fraction of norm followers is small, the society ends up in another steady state with opposite features. Depending on the initial conditions, the society might converge to one or the other steady state. We conclude the following: Proposition 6.1: (a) A higher p induces individuals to sustain the norm in the society with a smaller µ. (b) A larger µ induces individuals to sustain the norm in the society with a lower p. An agent does not automatically comply with a norm but deviates whenever the gains from deviation are sufficiently large. Proposition 6.1(a) shows that the norm is enforced due to a higher level of punishment of noncompliance. The norm is enforced due to the expectation that the violation will be punished. There is some possibility that the erosion of the reputation effect induces individuals to break the norm. Punishment would be used to enforce the norm if a substantial fraction of people has little reputation-​derived utility by obeying it. Punitive behavior is a way to restore norm compliance by penalizing violators. Proposition 6.1(b), on the other hand, shows that everyone is motivated due to reputation formation, despite a lower level of punishment by others. For a lower level of punishment, effective reputation building provides a way to sustain the norm.Then, an agent must believe

160  Social incentives and interaction

that sufficiently many other agents will comply with the norm as well. Punishment and reputation are interacting mechanisms. Reputation mechanisms generate an environment where the execution of costly punishment is less frequent without taking away its deterring force. The interaction of two mechanisms—​punishment and reputation—​not only comes closer to real life but also provides a convenient way to norm compliance. Even in societies that have strong governments, norms are both a source of law and often a cheap and effective substitute for law (Posner, 1997). The incentives for obeying laws are clear enough, but why do people obey norms? An individual suffers disutility from deviating from one’s norms, which causes behavior to conform to those norms even when it encompasses sacrifices to one’s own well-​being. We develop a theoretical framework to consider the problem of norm compliance. Some norms often require people to sacrifice for the group. The existence of such norms presents an evolutionary puzzle. This section analyzes the interaction between the potential for costly punishment and building personal reputation. The model considers two groups, group i and group j, in a society with one norm. Agents in group i choose whether to comply with the norm by incurring some cost. They acquire utility from the reputation derived from complying with the norm. This utility depends positively on the proportion of motivated compliers in group i. Behaving in a manner that violates the norm is costly. On the other hand, in group j, there are agents who value compliance and potentially punish noncompliance. They are altruistic punishers who are not direct victims for norm violation. We have a set of solutions to the problem of norm compliance.The first solution is the punishment-​based account. Following this account, people comply with the norm because the threat of punishment makes it in their interest to do so. Altruistic punishers are ready to incur some cost to prevent unfair behavior. They have a psychological motivation to restore norm compliance. Although altruistic punishment seems to have a solid foundation in human interaction, such a punishment leads to a large increase in losses for punishers. The second solution is, on the other hand, the reputation-​based account. People want to achieve the reputation of being fair. People comply with the norm because they care about their reputation. The present framework systematically investigates two scenarios to account for the long-​r un stability of the norm. In one scenario, there is some possibility that the erosion of the reputation effect induces individuals to break the norm. Individuals may have a selfish interest in violating it. Then, the norm is enforced due to a higher level of punishment of noncompliance. Punishment would be used to enforce the norm if a substantial fraction of people has little reputation-​ derived utility by obeying it. In another scenario, everyone is motivated due to reputation formation, despite a lower level of punishment by others. For a lower level of punishment, effective reputation building provides a way to sustain the norm. The two mechanisms, punishment and reputation, are interacting. The interaction not only comes closer to real life but also provides a convenient way to norm compliance.

Social incentives and interaction  161

6.4  Identity and culture Social identity is often understood as referring to group membership based on ascriptive characteristics. A person comes to view oneself as a member of a particular social entity. Through cognitive processes of categorization, one forms self-​ categories of organizational membership and one’s similarities with others in the organization, as well as dissimilarities with others in different organizations. People are defined and understood not only as individuals but also as belonging to certain social categories such as age categories, economic categories, or cultural categories. In economics, an individual’s identity is typically equated with one’s utility function which is exogenously given and stable. This is the notion of a unified identity. A person’s sense of self is ‘identity’ in Akerlof and Kranton’s (2000) terminology. Identity is envisaged as a sense of self, or self-​image, which is nested in social categories. At a very basic level, most people are aware of how they are differentiated from their surroundings. People often perceive themselves as members of social groups. Such group memberships may affect behavior in ways that cannot easily be reduced to material self-​interest. According to Akerlof and Kranton (2000, p. 716), [b]‌ecause of its explanatory power, numerous scholars in psychology, sociology, political science, anthropology, and history have adopted identity as a central concept. … We incorporate identity into a general model of behavior and then demonstrate how identity influences economic outcome. Akerlof and Kranton (2000) propose a utility function that incorporates identity as a motivation for behavior. Identity is based on social difference. In a seminal work, Akerlof and Kranton (2000) expand the utility function by modeling individual decisions as driven by internalized social norms. Identity is bound to social categories, and individuals identify with people in some categories and differentiate themselves from those in other categories. Identity gives rise to utility gain from conforming individual’s actions to the prescriptions of each social category. In their framework, individual utility is gained when actions conform to social norms and lost when they do not. The self-​image argument is a vector reflecting the different social categories. In Akerlof and Kranton’s (2000) prototype model, a person j’s utility depends on actions and identity as follows: Uj = Uj (aj, a-​j, Ij), where aj denotes the actions of person j, a-​j are the actions of others, and Ij is the identity of j. In their model, identity is based on actions, categories, and prescriptions. Each person belongs to one or more social categories, and the prescriptions indicate social norms associated with each social category. Thus, identity can be modeled as follows:

162  Social incentives and interaction

Ij = Ij (aj, a-​j; cj, P), where cj refers to social categories, and P refers to prescriptions. In Akerlof and Kranton’s (2000) model, identity depends on two social categories, Green and Red, and the correspondence of one’s own and others’ actions to behavioral prescriptions for their category. There are two possible activities: Activity One and Activity Two. Each person has a taste for either Activity. They consider an interaction between an individual with a taste for Activity One (Person One) and an individual with a taste for Activity Two (Person Two). If Person One (or Two) undertakes Activity One (or Two), he or she earns some positive utility. A person who chooses the activity that does not match his or her taste earns zero utility. As behavioral prescriptions, Greens should engage in Activity One, and Reds should engage in Activity Two. For example, by choosing Activity One, a person could affirm his or her identity as Green. First, identity changes the payoffs from one’s own actions. Person One who chooses Activity Two would lose his or her Green identity. Second, identity changes the payoffs of others’ actions (externality). If Person One and Person Two are paired, Activity Two on the part of Part Two diminishes Person One’s Green identity. Third, the choice of different identities affects an individual’s economic behavior. While Person Two could choose between Green and Red, he or she could never be a ‘true’ Green. Finally, the social categories and behavioral prescriptions can be changed, affecting identity-​based preferences. Furthermore, Akerlof and Kranton (2000) show that identity can explain many economic phenomena (e.g., gender discrimination in the labor market, the household division of labor, and the economics of social exclusion and poverty). There is an increasing awareness of cultural variations across societies and countries in the way that individuals behave under the same sets of circumstances. Groups develop common shared meanings or culture. The group shapes individuals’ preferences by defining a sense of belonging. Individuals internalize their social environment by adopting the values of a social group, which is reflected in differences in social identity. Culture can be understood as heterogeneous. There is cultural variation in the way people think about themselves and about others. These differences are responsible for differences in the way people behave. Cultural differences are, to a large extent, due to environmental differences. Therefore, patterns of social interactions affect the structure of cultural systems. Henrich et al. (2004) propose as the mechanism reducing ‘intra-​g roup’ differences and maintaining ‘inter-​g roup’ differences, by biasing individuals in favor of copying the common ideas, beliefs, and values. Culturally transmitted ideas, beliefs, and values are important for understanding human behavior. As Henrich et al. (2004) suggest, cultural evolution is likely to proceed much more rapidly than genetic evolution because cultural transmission can spread novel behavior, ideas, and practices among populations within a single generation. Cultural transmission involves learning from others. It allows individuals to

Social incentives and interaction  163

adapt more quickly to changing environments than is possible under either a strictly genetic mode of transmission or a system that includes only individual trial-​and-​ error learning. Using socially transmitted information, people can make predictions about the intent of others. Once a cultural trait is acquired and internalized, it tends to direct actions without significant cognitive effort or reflection. The ultimatum bargaining game involves two parties with payoffs that are monetary and transferrable. This sequential-​move bargaining game begins when the first player, the ‘proposer,’ proposes a division of a fixed monetary pie (typically provided by the experimenter). This proposed split represents a take-​it-​or-​ leave-​it offer. The second player, the ‘responder,’ may accept or reject the offer. The interaction between the players is anonymous. If the responder accepts the proposed division, then the division is, as proposed by the proposer, implemented. If the responder rejects the proposed division, each player earns nothing. Game theory predicts that a self-​interested responder will accept any proposed positive payoff, no matter how small. Anticipating this choice, a self-​interested proposer will offer nothing more than the smallest amount possible. The equilibrium offer (i.e., the division for which no player has anything to gain by doing something differently) allocates the smallest positive payoff to the responder. The proposer offers the responder the smallest possible amount of money, and the responder accepts: this is the subgame perfect equilibrium. The subgame perfect equilibrium gives all the bargaining power to the proposer. That is, any demand that leaves the second player with anything should be accepted, and consequently, the proposer should get almost all of the pie. Experimental subjects repeatedly violate the theoretical predictions.15 Results from ultimatum game experiments in developed countries show that proposers do not demand nearly this amount, generally demanding between 50% and 60% of the total. Research on the ultimate game experiments has consistently found that responders tend to reject low offers and thus behave at odds with the assumption that they simply maximize their self-​ interest (Camerer, 2003). In experimental research, games have been used recently in anthropology to investigate the basis of prosocial human behavior such as fairness, altruism, and cooperation (Henrich et al., 2004). Using games, researchers standardize protocols across field sites and compare behavior and groups to explore the range of cross-​ cultural variation. The experiment in Roth et al. (1991) is the first cross-​country comparison study on ultimatum bargaining in the United States, Israel, Japan, and Slovenia (Yugoslavia). In this experiment, subjects were first assigned a role: proposers and responders. They played the game in their respective roles for ten rounds, each time with a different, anonymous, and randomly selected opponent. The pie was worth the equivalent of $10 in terms of purchasing power in all four countries. Subjects were paid according to their monetary payoffs in one randomly selected round. The responder can choose to accept or reject the offer. If the responder accepts the offer, the proposer gets the demanded amount and the responder gets the remainder. If the responder rejects, neither player receives anything.

164  Social incentives and interaction

The experimental results in Roth et al. (1991) can be summarized as follows: • • •

The frequencies of the offers implied by the subgame perfect equilibrium of the game ($0 or $1) ranged from 1% of the time in Slovenia to 9.3% in Israel. Offers were highest in the United Sates and Slovenia, then in Japan, and lowest in Israel. The conditional frequency of rejected offers was inversely related to the fraction of the pie offered. That is, low offers were rejected more frequently than high offers. The conditional frequencies of acceptance of offers were lowest in Slovenia, then in the United States, and highest in Israel and Japan. Furthermore, the probability that a given offer is rejected was lower in countries where lower offers were observed.

Thus, Roth et al. (1991) find ‘small,’ significant differences in the ultimatum bargaining game. They conclude that these differences can be explained as ‘cultural differences.’ Camerer (2003) argues that cross-​cultural comparison raises at least four difficult methodological problems: stakes, languages, experimenter interactions, and confounds. First, controlling for stakes requires the experimenter to match the purchasing power of the stake in different cultures. Second, keeping the meaning of instructions as constant as possible is important. Third, the biggest mistake in controlling for identity is to use a different experimenter in each culture. The ideal experimenters speak both languages and are perceived similarly in both cultures. Fourth, it is extremely difficult to avoid the effects of potentially important variables that are confounded with culture (causing ‘identification problems’ in econometrics terms). In the ultimatum bargaining game, existing experimental data and analyses have shown that subjects from industrial societies behave almost similarly; the mean offer from proposers averages between 40% and 50% of the total, and responders often reject offers lower than 20% of the total. However, Henrich (2000) shows that the Machiguenga of the Peruvian Amazon behave very differently from subjects drawn from industrial societies in the ultimatum bargaining game. The Machiguenga are described as ‘socially disconnected,’ with economic life centering on the individual family and little opportunity for anonymous transactions. For the Machiguenga, the economic unit is the family; families fully produce for their own needs (food, clothing, etc.) and do not rely on institutions or other families for their social or economics welfare. Cooperation above the family level is almost unknown, and the Machiguenga are quite socially disconnected. In Henrich’s (2000) test, Machiguenga proposers offered only 26% of the total sum, and responders almost always accept offers less than 20% of the total. Machiguenga proposers seem to possess little or no sense of obligation to provide an equal share to responders, and responders seem to have little or no expectation of receiving an equal share. Cultural differences greatly influence economic behavior. Henrich et al. (2001) recruited subjects from 15 small-​scale societies to expand the diversity of economic and cultural conditions in various games including the

Social incentives and interaction  165

ultimatum bargaining game. Their sample consisted of three foraging groups, six slash-​and-​burn horticulturists, four nomadic herding groups, and two sedentary, small-​scale agricultural societies. The experimental results in Henrich et al. (2001) can be summarized as follows: • •

• •

The canonical model of self-​interested behavior was not supported in any society studied. There was considerably behavioral variability across groups. The mean offers from proposers ranged from 26% to 58%. The most common behavior for the Machiguenga was to offer zero. In some groups, rejections were extremely rare, even in the presence of very low offers, while in others, rejection rates were substantial, including frequent rejections of offers above 50%. Group-​level differences in economic organization and the degree of market integration explain a substantial portion of the behavioral variation across societies; the higher the degree of market integration and the higher the payoffs to cooperation, the greater the level of cooperation and sharing in experimental games. Individual-​level economic and demographic variables do not explain behavior either within or across groups. Behavior in the experiments is generally consistent with economic patterns of everyday life in these societies.

Why would people reject a positive amount of money? The responder would reject the offer because he or she is angry at being treated unfairly, and he or she is willing to pay to hurt the person who perpetrated this unfairness. This is not just a matter of developing a capability for anger, but rather a matter for developing a cognitive understanding of whether an angry response is appropriate. What matters is that the ‘right’ response is elicited in the ‘right’ settings. Prosocial emotions (the empathy people feel toward those who have been kind to them and the hostility toward whose who have not) thus serve to enhance the individual and group fitness in the social conditions. This reflects strong reciprocity, namely a propensity to reward those who have behaved cooperatively and correspondingly to punish those who have violated norms of acceptance behavior, even when reward and punishment cannot be justified in terms of outcome-​oriented preferences.

6.5  Social preferences and internal moral constraints Moral problems are problems of justice and fairness. Income differentials are usually reduced by means of progressive income taxation and monetary transfers to the needy.The government may promote equality also by publicly providing some private goods. A crucial issue is then to identify the factors which derive the support for income redistribution in society. The support for redistributive institutions largely depends on individuals’ willingness to redistribute. There is plenty experimental evidence that people have an innate desire for fairness. For example, in

166  Social incentives and interaction

the ultimatum game, people are ready to suffer a monetary loss themselves just to punish behavior that is considered unfair. A prosocial individual (the responder) concerned about fairness will reject unequal offers in the ultimatum game. David Hume’s ethical theory is rooted in his account of human sentiments and of how some sentiments become specifically moral sentiments. In A Treatise of Human Nature, Hume ([1739–​1740] 1983) considered the imagination a prerequisite to feelings of sympathy. Individuals morally evaluate the circumstances of others by imaging themselves in the observed situation. Hume emphasized the primary role of human sentiments in moral judgment. For Hume, reason is morally inert and appropriate moral action is necessarily based on human sentiments. Human beings are made moral because they are motivated by the approbation that they receive from others with whom they sympathize. The term ‘sympathy’ originates in the works of 18th century moral philosophers such as David Hume and Adam Smith. Sympathy is characterized as an outcome of imagination simply because people do not have access to other minds. People only have access to others’ observable situations. The imagination allows them to place themselves in the witnessed circumstance. We can postulate a new taste to explain moral behavior. Utility functions can take a variety of forms.They can combine self-​directed and social or other-​oriented preferences. This approach extends the assumption of the self-​interested individual to include social preferences. For a two-​person interaction, i and j, the specific form of altruism can be described as follows: ui = ωi + αω j , where ui describes the individual i’s utility, ωi i’s welfare, and ω j the welfare for the social partner j. Here, α (0 ≤ α ≤ 1) describes how much individual i cares for the welfare of one’s partner. When α = 0, i cares only for one’s own welfare. This formula is a purely self-​oriented utility function and induces i to ignore j’s welfare. When α = 1, i values the partner j’s welfare as much as one’s own. This may be referred to as egalitarian altruism. However, moral judgments might not depend on the individuals’ identities and other particular circumstances. Impartiality is then a distinctive feature of moral judgments on collective life. Moral judgments are not simply expressions of an individual’s interests or preferences. Moral judgments involve expressions of attitudes. They are claims to universality in their context and transcend the interests of those concerned. Therefore, moral dispositions cannot be adequately summarized by any preference function. Morality and preferences may be different things. For Immanuel Kant, the 18th-​century German philosopher, good will is the only thing good in itself: the highest good is the good will.16 It is an intrinsic moral value that should influence human actions. The Groundworks of the Metaphysics of Morals is Kant’s central contribution to moral philosophy and has inspired controversy ever since it was first published in 1785.17 According to Kant, an action has moral worth only if it is performed from a sense of duty. Duty rather than purpose

Social incentives and interaction  167

is the fundamental concept of Kantian ethics. Kant assumed that the ordinary person already knew that one had a duty to be honest, industrious, charitable, truthful, and so forth. An individual’s decision is made due to his or her moral sense of duty. To act from a good will is to act from duty. If a person strictly acts for pure pleasure and reward, or when he or she is afraid of social sanctions, he or she is not guided by good will. Kant’s ethics is thus an ethics of duty rather than an ethics of consequences. The ethical person is one who acts from the right intentions. Morality for Kant is reason in action. Following Kant, the emotional feelings that derive moral behavior are to be distrusted. Kant looked to the application of reason to establish universal moral laws since reason is shared by all humans. For Kant, the primary problem is whether there are any genuine moral principles underlying our moral judgments in the first place. If the moral problem is only the psychological problem of motivation, it appears impossible that the moral judgments hold universally. Reason must discern universal principles of right conduct without regard for any particular end or desire in a given situation. Human beings have free will; they are able to act from laws required by reason. Moral action consists of dutiful adherence to moral rules. Thus, we have a duty to follow the precepts that our reason reveals. Duty is the practical unconditional necessity of action, and it can be a law for all human wills. Kant distinguished between the conditional universality of ‘hypothetical imperative’ and the unconditional universality of ‘categorical imperative.’ Only the latter has the formal character that Kant claimed to be the distinctive feature of reason. The hypothetical imperative recommends actions as means to some goal: “if you want to achieve X, you should do Y.” Since the hypothetical imperative would make moral behavior conditional on inclinations or personal goals, it cannot be moral from Kant’s standpoint. The categorical imperative, on the other hand, has the form: “act only according to that maxim whereby you can, at the same time, will that it should become a universal law.” This formula represents the familiar ethical standard of universalization, which holds that any rule of action one adopts must be consistent with everyone else’s adopting it as well. A moral rule must take the form of universalizable categorical imperative. Kant’s categorical imperative, derived from a priori reasoning, is independent of personal desires or subjective ends. It merely reflects common reasoning concerning moral principles. To apply the categorical imperative to decision-​making reveals duties that must be followed in order to act morally. The action can only be undertaken if the principle on which the action is based passes the test of the categorical imperative. The categorical imperative is the principle for judging one’s proposed plans of action, or maxims, in order to assess their moral status. It holds universally for everyone regardless of the circumstances or situations in which one finds oneself. In Kant’s view, a morality depending on contingent circumstances such as bad luck would be unacceptable. Kant’s ethical system is based on the moral status of actions themselves, rather than their consequences. Moral rules are seen as rules of duty in Kantian ethics. The moral philosophy of Kant is thus considered as ethical impersonalism—​the rigid separation of morality from the pursuit of personal goals.

168  Social incentives and interaction

Orthodox neoclassical economists are reluctant to appeal to internal moral constraints. They usually attribute the cooperation they observe to external constraints such as laws and regulations. People only cooperate for fear of external constraints. All behavioral prescriptions for individuals require external rules. People respond to changes in legal rules just as consumers respond to changes in prices. However, government enforcement is far from perfect. Government does not have perfect information about all potential problems. People appear to cooperate out of a sense of justice. According to Kant, it is possible to derive morally binding rules of conduct. Internal moral constraints are the rules that people choose to follow independent of external rules. People appear to cooperate out of a sense of justice. They may act cooperatively because they adopt internal moral constraints as certain moral rules. Laffont (1975) asks why (at least in some countries) people do not leave their beer cans on beaches even though the impact on their own welfare from doing so is negligible. According to Laffont (1975, p. 431), “[e]‌very economic action takes place in the framework of a moral or ethics.” Kant’s categorical imperative can explain the prosocial behavior observed in the beach example. An individual then assumes that, according to Kantian principle, the others will act as he or she does. Individuals behave as if they expect that choosing an action increases the likelihood that the others also select the action. Kant’s categorical imperative is considered as a cooperative norm. Recall that a Nash equilibrium is a situation in which if a player deviates from one’s intended action, one will get a worse payoff, assuming that others do not deviate. Roemer (2010; 2015) presents a ‘Kantian equilibrium’ of a game, which is defined with respect to a class of counterfactual deviations that each player envisions.That is, a player is at a Kantian equilibrium if and only if one would receive a lower payoff upon a deviation (i.e., increasing or decreasing one’s equilibrium activity level by some factor), assuming that all others would deviate likewise. In Kantian equilibrium, each player chooses the common strategy to be adopted by everyone to maximize one’s own utility. A Kantian equilibrium is characterized as a configuration of activity levels in a community where everyone follows Kant’s principle of behavior. Rabin ([1995] 2019) considers an agent who chooses whether or not to engage in an activity which is personally pleasurable but possibly causes harm to others. In Rabin’s ([1995] 2019) model, an agent chooses x = 1 or 0, where x = 1 means he or she engages in the activity and x = 0 means he or she does not. The activity brings enjoyment V(x) to the agent personally. It is assumed that V(0) = 0 and V(1) = v ∈(0,1), where v is a parameter which measures how pleasurable he or she finds this activity. However, this agent may not want to engage in the activity which possibly causes harm to others. Let H(x) denote a measure of the social harm caused by the activity.The agent believes that the activity causes harm with probability q: H(1) = 1 with probability q and H(1) = 0 with probability 1 -​ q. On the other hand, not engaging in the activity does not cause harm: H(0) = 0. Rabin ([1995] 2019) defines an agent with moral preferences as follows:

Social incentives and interaction  169

Definition 6.1 (Rabin [1995] 2019): An agent has moral preferences if he or she maximizes the von Neumann-​Morgenstern utility function: v −q U P (v , q ) =  0

if x = 1 if x = 0.

This specification implies that the agent will engage in the activity if and only if the probability of social harm is less than or equal to the personal benefit from engaging in the activity, q < v. Instead of maximizing social surplus, Rabin ([1995] 2019) supposes an agent with the internal moral constraint not engaging in an activity that is too likely to harm others. In Rabin ([1995] 2019), an agent guided by a moral rule is defined as follows: Definition 6.2 (Rabin [1995] 2019): An agent follows a moral rule if there exists y > 0 such that he or she maximizes the utility function: v − g (q ) U R (v , q ) =  0

if x = 1 if x = 0,

where g(q) = 1 if q > y and g(q) = 0 if q < y. That is, if the probability that his or her activity causes social harm is too high (q > y), the agent does not engage in the activity. On the other hand, if the probability of social harm is not deemed so high (q < y), he or she engages in the activity. To recognize that people have internal moral constraints provides us with more options. Thus, we have two explanatory components at the individual level: moral preferences and internal moral constraints. Both having moral preferences and following moral rules are important for the choice whether or not to engage in the activity that may cause social harm.

Notes 1 Polanyi (1969) uses the concept of ‘tacit knowledge’ to refer to all those kinds of scientific knowledge that cannot be expressed in explicit form (spoken words, formulae, maps, graphs, mathematical theory, and so on).The concept of tacit knowledge has been widely used in social sciences in order to incorporate it within a more comprehensive theory of practices and their role in social reality. 2 According to Markose (2005), the von Neumann models based on cellular automata have laid the ground rules of modern complex system theory regarding (1) the use of large ensembles of microlevel computational entities or automata following simple rules of local interaction and connectivity, (2) the capacity of these computational entities to self-​reproduce and also to produce automata of greater complexity than themselves, and (3) the use of the principles of computing machines to explain diverse system-​wide or global dynamics. 3 Rizzo (2013, pp. 50–​51) lists eight highly interrelated themes about Austrian economics: (1) the subjective quality of human decision-​making, (2) the individual’s perception of the

170  Social incentives and interaction

passage of time, (3) the radical uncertainty of expectations, (4) the decentralization of explicit and tacit knowledge in society, (5) the dynamic market processes generated by individual actions, (6) the function of the price system in transmitting knowledge, (7) the supplementary role of cultural norms and other cultural products in conveying knowledge, and (8) the spontaneous evolution of social institutions. 4 See Sugden (1995) for a formal theory of focal points. In his model, there is a one-​to-​one relationship between labels and strategies, so that players are always able to distinguish between their strategies. 5 As an example, in Lewis’ home town of Oberlin, Ohio, there was a time when local calls were cut off without warning after three minutes. Then, in the population of phone users, an interaction occurs between two phone users when a call between them is cut off. One convention is that the original caller calls back. Another one is that the person called calls back. In fact, the first of these was the convention. 6 There are several arguments as to whether the formalization of a wider notion of rationality is necessary to endow the players in Lewis’ definition of conventions with a sufficient reason to coordinate their behavior. For example, according to Bicchieri (2006), players only need empirical or plain expectations of the conformity of others to a particular convention to actually conform to that convention. 7 In Vanderschraaf (1995), Lewis’ (1969) conventions are shown to be correlated equilibria. 8 The dictator game in theory gives rise to very inequitable distributions of resources. However, when the game is played for real, fair allocations figure prominently. Many game experiments offer abundant evidence that contradicts the hypothesis that all players are motivated only by their own material interest (see Camerer, 2003). 9 Building on Akerlof ’s (1980) model, Naylor (1989) explains the logic of collective strike action. 10 Sliwka (2007) considers the notion of trust as a credible signal of a social norm. 11 Tirole (1996) considers the joint dynamics of individual and collective reputations. 12 See Nowak and Sigmund (1998) for a mathematical model of indirect reciprocity. Their model is based on image scoring; agents develop a positive reputation for cooperating and only cooperate with others whose score is above a threshold (image score). 13 Engelman and Fischbacher (2009) assess the interplay of indirect reciprocity and strategic reputation building in an experimental helping game. When indirect reciprocity is not contaminated by incentives for strategic reputation building, they call this pure indirect reciprocity. 14 The cost C can be endogenized by assuming that individuals have beliefs about how they are judged for the norm deviation. This allows individuals to hold strong or weak beliefs in the sense that the cost C can be high or low. 15 See Güth et al. (1982) for early experiments. 16 To understand Kant, it would be necessary to realize that he stands in the Aristotelian and scholastic tradition still influential in his time. 17 Kant published his Groundworks 2 years before he finished his Critique of Practical Reason (1797), 5 years before he completed the Critique of Judgment (1790), and more than 12 years before he came out with his conclusive statements on ethics in the Metaphysics of Morals (1797/​1798).


7.1  The tragedy of the commons The term ‘commons’ is widely used in political discourses on how to tackle challenges such as climate change, food security, and transmission of knowledge. Climate change can be viewed as an illustration of collective action problems detrimental to commons. The greenhouse gases that any individual emits have an imperceptible impact on climate change, while activities producing greenhouse gases (e.g., driving, heating, and air conditioning) have noticeable benefits to the individual’s personal well-​being. In academic writings, the commons have come to represent an alternative model of social organization, going beyond the market–​state dichotomy. Governance is about forming institutional structures: the concerns are making social priorities, resolving conflicts, and facilitating coordination. Polycentricity is a concept that describes a complex form of governance with multiple centers of decision-​making, each of which operates with some degree of autonomy. Governance arrangements exhibiting polycentric characteristics may be capable of striking a balance between centralized and decentralized (or community-​level) governance. In a polycentric governance system, many autonomous units take others into account through processes of cooperation, competition, conflict, and conflict resolution. The tragedy of the commons is a thought experiment used by Hardin (1968). It is a metaphor that refers to the ultimate destruction of a common. According to Hardin (1968), users of a common are selfishly rational, and they will consequently overuse the common, resulting in the tragedy of its destruction. Imagine some herdsmen who have access to a common pasture. Each of them can increase one’s herd by at least one sheep. If the herdsmen behave fully rationally and self-​ interestedly, they bring one more sheep to the common. This process is repeated, and the herds increase progressively, which causes exploitation from which the DOI: 10.4324/​9780429344817-7

172  Governing the commons

pasture cannot recover. Thus, the excessive search for the individual good results in the destruction of the common. Environmental degradation and resource depletion have been globally pervasive concerns over the last few decades. The problem of the management of common-​pool resources has received relatively increased attention in economics. The properties of common-​pool resources lead to at least two types of collective action problems in the management: those of resource appropriation and provision. An appropriation problem can result in overconsumption of a subtractable resource, in which an individual benefits from personal consumption at the expense of the community and the conditions of the resource. A provision problem can result in under-​provision of the infrastructure needed to appropriate a resource. Conserving common-​ pool resources is problematic because many individuals rely on the benefits of extraction for their livelihoods. A common-​pool resource exists when multiple parties appropriate from a resource, and appropriation by any one party reduces the resources available to others. It differs from a public good in that resource units harvested by one appropriator are not available to any other appropriator. A public good has two essential attributes: non-​excludability and non-​r ivalry in consumption. A common-​pool resource is, on the other hand, non-​excludable but rival.1 Non-​excludability means the difficulty to exclude non-​ paying consumers from consumption, which is a feature that both types of goods share. The possibility of non-​r ival consumption by multiple consumers is the major feature that distinguishes public goods from common-​pool resources. One person’s enjoyment of a sunset does not subtract from others’ enjoyment of a sunset. On the other hand, if one fisherman lands a ton of fish, that ton is not available for others. Common-​pool resources are expected to suffer from collective action problems when individuals act in a self-​interested manner. Individuals make independent and anonymous decisions and primarily focus on their own immediate payoffs in a common-​pool resource setting. Hardin’s (1968) metaphor is based on the assumption that human behavior is driven by selfish identity. Individual selfish users are unwilling to pay the costs of conservation because the benefits of doing so are shared collectively. Resources involving open access are much more vulnerable to overharvesting than those with restricted access. Joint users may harvest resources without gaining prior permission. Unrestricted individualistic decision-​making in relation to common-​pool resources will result in tragedy: the greater the individual effort, the worse off people become. The action of one agent affects the resource stock, which in turn affects the well-​being of another agent through decreased resource availability. Given their rival and non-​excludable properties, common-​ pool resources are expected to suffer from overexploitation. The fact that consumption of the resource by one person reduces the amount available to others implies that these resource users face a typical social dilemma. Social dilemmas are group interactions in which an individual maximizes his or her own payoff when he or she does not cooperate, but attaining the social optimum requires cooperation. In order for a governance arrangement to be successful, it should stop the overuse of the resources it governs. Incentives need to be provided to users to limit

Governing the commons  173

their consumption levels and boost their provision levels. Several approaches have been suggested to mitigate the social dilemmas. The structure of decision-​making arrangements has to be modified to enable individuals to act jointly in relation to those resources as a common property.2 Appropriators may attempt to implement governance by investing in monitoring and enforcement. The assumption of self-​enforcement is consistent with the case-​study analysis by Ostrom (1990). If communities are able to design their own usage schemes, organize themselves, and enforce the rules they design, then collective action and self-​governance can be successful in reducing the impact of the social dilemmas. Ostrom’s (1990) work finds that, in common-​pool resource situations, collectively desirable outcomes arise when resource users are left to develop the rules and enforcement mechanisms themselves. Ostrom (1990) suggests that users are not necessarily always short-​term maximizers, and that they can develop self-​governing institutional arrangements to regulate the commons. For Hardin (1968), there are only two solutions to the tragedy of the commons: the coercive state and privatization. That is, one solution is control of natural resources by a central government agency, and the second solution is the imposition of private property. Ostrom (1990) suggests that user self-​governance can be a third option to prevent tragic outcomes. Community management is a system which is based on cooperation, in which individual decision units formulate both individual and common goals. In such a community management, users themselves have the decision-​making responsibility. Ostrom, based on her case-​study analysis, identifies various design principles needed to achieve long-​enduring institutions: (1) clearly defined boundaries that allow the exclusion of external resource users; (2) rules for the resource use adapted to local conditions; (3) collective arrangements that let most resource users take part in the decision-​making process; (4) effective monitoring under responsibility of the local users; (5) a scale of graduated sanctions for the users who violate the community rules; (6) cheap and easy mechanisms of conflict resolution; (7) community self-​determination that is recognized by higher level authorities; and (8) a larger common-​pool resource is organized in the form of multiple layers of nested enterprises, with small local common-​pool resources at the base level. Local communities often develop their own cooperative rules without enforcement from the top-​down and without imposing private property rights.3 Ostrom’s primary concern is on community-​based natural resource management within a polycentric governance system. A polycentric governance system can be combined and recombined in a variety of ways to provide solutions to certain local problems that result from complex economic interactions (Ostrom, 2005).

7.2  Social preferences and environmental sustainability This section offers a novel explanation for the conservation of a common-​pool resource.4 Namely, economic agents who consider the resource conservation to be their major concern are driven by prosocial or pro-​environmental identity with moral preferences. In the opening of Adam Smith’s The Theory of Moral Sentiments,

174  Governing the commons

[h]‌ow selfish soever man may be supposed, there are evidently some principles in his nature, which interest him in the fortune of others, and render their happiness necessary to him, though he derives nothing from it except the pleasure of seeing it. Smith, [1759] 1981, p. 9 Rizzolatti and Craighero (2005) describe that this famous sentence by Adam Smith contains the two distinct ideas. First, we are endowed with a mechanism that makes us share the fortunes of others. By observing others, we enter in an empathic relation with them. Second, because of our empathy with others, we are compelled to desire their happiness. Others’ unhappiness somehow intrudes into us; if others are unhappy, we are also unhappy. Humans have an innate interest in the fortunes of other people and desire for empathy with others. The cognitive dimension of empathy refers to the understanding of another person’s feelings or characteristics. We consider how changes in identity affect changes in collective action. Akerlof and Kranton’s (2000) attempt is the first one to emphasize the importance of identity for economic decision-​making. In their framework, identity is defined as a sense of belonging to a social category which is exogenous. An individual is understood to have a sense of self when one associates oneself with others in one’s social categories and differentiates oneself from non-​members. Insofar as individuals internalize the code of conduct linked to a prescribed behavior, cognitive dissonance may be evoked when violating the prescription. Thus, deviating from prescriptions associated with the social categories is inherently costly and generates identity losses. An individual takes one’s identity as given and only reacts to belonging to a specific social category. Differently from their framework, we emphasize the ‘interplay’ of different identities within an individual. In order to achieve collective action, the members of a resource user group need to develop an identity which is founded on a shared understanding. Identity plays an important role in differences in behavior. Such an identity can be considered as shared meanings which facilitate or impede collective action in the use of common-​pool resources. We argue that understanding the nature of the relationship between identity and collective action is fundamental in the management of common-​pool resources. This section studies the identity of individuals as a source of selfish or pro-​ environmental behavior. Consider a community which is made up of a fixed number n of individuals. Individuals have complete rights of access to a common-​ pool resource. They simultaneously decide the amount of resources to exploit from a common pool of renewable resources.The total stock of the resource in existence is denoted by F. Each individual is endowed with a fixed activity level, one, which he or she can allocate to harvesting the common-​pool resource or to an alternative activity (resource conservation). The individual selects activity allocation between the resource extraction and the resource conservation. The individual chooses a degree of effort expended on the resource extraction, e, where 0 < e < 1. That is, if the individual chooses to allocate effort e in the resource extraction, then he or she puts in the remaining 1 -​e in the resource conservation.

Governing the commons  175

The resource extraction entails a (direct) cost c(e). The individual puts in effort e at a cost c(e) in the resource extraction. It is assumed that c(.) is strictly increasing and strictly convex in e, that is, c′(.) > 0 and c′′(.) > 0. In addition, to ensure an interior solution, it is assumed that c′(0) = 0 and lim e →1 c'(e ) = ∞. Then, the marginal cost of harvesting, c′(e), becomes sufficiently large as e approaches a common maximum effort level that can be exerted by each individual. There are two types of identities that an individual can take: selfish (S) and pro-​ environmental (PE). Selfish and pro-​environmental identities receive the material payoff related to resource extraction e. Each identity receives a share of the total, equal to his or her share of total effort in the resource extraction. The material payoff b(e) is then given by: e b (e ) =   (1 / n ) F

if ne ≤ F (7.1) otherwise

If the sum of all harvesting efforts, ne, does not exceed the available amount of the resource F, then each individual receives a return equal to his or her own effort e. However, if the sum of harvesting efforts exceeds the resource stock, the resource might be exhausted. Then, it is supposed that each individual receives (e/​ne)F = (1/​n)F if ne > F. Let US denote the selfish identity’s payoff. The selfish identity only cares for the material payoff. Definition 7.1: An individual holds the selfish identity if the utility function is given by: US = b(e) -​c(e). (7.2) Besides the material payoff, the pro-​ environmental identity also obtains psychological payoff per unit of activity in the resource conservation defined by k, where 0 < k < 1. It is interpreted as the degree of moral satisfaction that the pro-​ environmental identity derives from the resource conservation. Let UPE denote the payoff for the pro-​environmental identity with moral preferences. The pro-​environmental identity with moral preferences is defined as follows: Definition 7.2: An individual holds the identity with moral preferences if the utility function is given by: UPE = b(e) + k(1 -​e) -​c(e), (7.3) where k reflects the degree of moral satisfaction from the resource conservation. The model focuses on ‘intra-​personal’ dimensions within an individual, not on ‘inter-​personal’ ones among individuals. The timing of the situation is the following:

176  Governing the commons

[Stage 1] The individual adopts either the selfish identity (Definition 7.1) or the pro-​environmental identity (Definition 7.2). [Stage 2] The individual chooses an effort e in the resource extraction. At Stage 1, each individual gets an opportunity to adopt either the selfish identity or the pro-​environmental identity. Let x be the individual’s subjective belief about holding the pro-​environmental identity, where 0 < x < 1. With probability x, the individual holds the pro-​environmental identity; with probability 1 -​x, he or she holds the selfish identity. At Stage 2, each individual chooses the effort level at Stage 2, taking into account his or her sense of identity. Using (7.2) and (7.3), the individual’s expected utility at Stage 2 is written as: (1 -​x)US + xUPE = (1 -​x){b(e) -​c(e)} + x{b(e) + k(1 -​e) -​c(e)} = b(e) -​c(e) + xk(1 -​e). (7.4) Individual identity is composed as a weighted sum of two different types: selfish and pro-​environmental. The individual chooses the optimal level of effort in the resource extraction to maximize his or her utility (7.4). The problem that he or she faces at Stage 2 is maxeb(e) -​c(e) + xk(1 -​e). The optimal level of effort equates the marginal benefit of providing effort to its marginal cost. The first proposition states the existence of an equilibrium: Proposition 7.1: If 1 < (1/​n)F, there exists a unique equilibrium where an agent chooses a positive effort in the resource extraction. Otherwise, there exists no equilibrium in the interval ((1/​n)F, 1] and there may not always exist an equilibrium in the interval [0, (1/​n)F]. Proof. Suppose that 1 < (1/​n)F. Since e < 1, e < (1/​n)F holds. Then, b(e) = e from (7.1). Thus, from (7.4), the agent’s payoff is given by (1 − x)US + xUPE = e + xk (1 − e) − c(e). By differentiating this payoff with respect to e, the first-​order condition for an interior solution is obtained as follows: 1 − xk = c′(e). (7.5) Since 0 < x < 1 and 0 < k < 1, the left-​hand side of (7.5) is a positive constant. Since c′(e) > 0 and c′′(e) > 0 for e > 0, the right-​hand side of (7.5) is strictly increasing; by the assumption (lime→1c′(e) = ∞), c′(e) goes through the positive real line 1 − xk as e goes from 0 to 1. Thus, there exists a unique value of e that satisfies (7.5) in the interval [0, 1]. This unique equilibrium satisfies the second-​order condition for a maximum. Now suppose that (1/​n)F < 1. Consider e in the interval ((1/​n)F, 1]. Then, b(e) = (1/​n)F from (7.1). Thus, from (7.4), the individual’s payoff is given by (1 − x) US + xUPE = (1/​n)F + xk(1 − e) − c(e). Differentiating this payoff with respect to e yields −xk = c′(e). Since the right-​hand side is strictly increasing, there exists no equilibrium in the interval ((1/​n)F, 1].

Governing the commons  177 c′ (e)

c′ (e)

1 – xk

O FIGURE 7.1  An




equilibrium: 1 –​xk and 1 < (1/​n)F

Consider e in the interval [0, (1/​n)F], where (1/​n)F < 1. Then, b(e) = e from (7.1). The individual’s payoff is given by (1 − x)US + xUPE = e + xk(1 − e) − c(e) from (7.4). Differentiating this payoff with respect to e yields (7.5). As shown above, there is a unique equilibrium in the interval [0, 1]. However, an equilibrium may not always exist in the interval [0, (1/​n)F]. ■ For 1 < (1/​n)F in Figure 7.1, the curve c′(e) intersects the real line, 1 − xk. Thus, if the resource stock is sufficiently large compared with the number of individuals, there is a unique equilibrium in the interval [0, 1]. On the other hand, for (1/​n)F < 1 (i.e., the resource stock is sufficiently small compared with the number of individuals), Figure 7.2 depicts the case in which the curve c′(e), does not intersect the real line, 1 − xk, in the interval [0, (1/​n)F]. For a steeper c′(e), there may exist an equilibrium in the interval [0, (1/​n)F] (Figure 7.3). However, as shown in Figure 7.2, it is not always true that an equilibrium exists in the interval [0, (1/​n)F]. Second, it is shown that choosing the pro-​ environmental identity reduces the resource extraction. A significant departure from conventional models of the commons is that an amount of effort expended on the resource extraction depends on the agent’s subjective belief that he or she holds the pro-​environmental identity. The subjective belief about the identity adoption affects the resource extraction and conservation.

178  Governing the commons c′ (e) c′ (e)

1 – xk

O FIGURE 7.2  No




equilibrium: 1 –​xk and (1/​n)F < 1

c′ (e) c′ (e)

1 – xk

O FIGURE 7.3  An


equilibrium: 1 –​xk and (1/​n)F < 1



Governing the commons  179

Proposition 7.2: If 1 < (1/​n)F, an equilibrium effort in the resource extraction is strictly decreasing in the probability that adopts the pro-​environmental identity, x. Proof. As shown in Proposition 7.1, if 1 < (1/​n)F, there exists a unique positive value of e that satisfies (7.5). Implicit differentiation of (7.5) with respect to x yields: de/​dx = −k(1/​c′′(e)). (7.6) Since 0 < k 0 for e > 0, the sign of de/​dx in (7.6) is negative. ■ Proposition 7.2 shows how the equilibrium value of e varies with the parameter x. The equilibrium resource extraction depends on the probability that the individual holds the pro-​environmental identity. The value e decreases as x increases: de(x)/​ dx < 0. The greater an individual’s exposure to x, the greater his or her effort in the resource conservation. In Proposition 7.2, a different probability that the individual holds the pro-​environmental identity, x, prescribes a different behavior in which a different amount of effort is selected in the resource extraction. As the probability that the individual holds the pro-​environmental identity increases, he or she is expected to provide a lower (higher) activity level to the resource extraction (conservation). The model in this section is set in discrete time, and time is denoted by t = 0, 1, 2, … Resource extraction and conservation occur at discrete points in time. This section considers a renewable resource with common-​pool characteristics, whose stock at any point in time t is given by F(t). If F(t + 1) -​F(t) > 0 for all t > 0, resource sustainability can be achieved. In the common-​ pool resource extraction problem, each individual would be better off if all would restrain their use, but it is not in the interest of the selfish identity to do so.5 The pro-​environmental identity can be deterred from overexploitation by moral satisfaction from the resource conservation. The conservation activity ‘regrows’ the renewable resource, while effort in the extraction activity depletes the resource. For a fixed number n of individuals, the aggregate resource conservation (or renewal) is given by n(1 -​ e(t)) at t, while the aggregate resource extraction is given by ne(t) at t.The interaction between these components governs the dynamics of the model. After the individual decisions, the resource stock is depleted by an amount equal to the sum of extraction efforts, and it regrows by an amount equal to the sum of conservation activities. The next period’s resource stock is given by F(t + 1), and it is available for the individuals at t + 1. Thus, the stock dynamics is the difference between stock accumulation due to regeneration and stock depletion due to harvesting, namely: F(t + 1) -​F(t) = n(1 -​e(t)) -​ne(t).(7.7) Common-​pool resources are sustainable over time if F(t + 1) -​ F(t) > 0 for any t. We consider whether a threshold value, e*, exists such that F(t + 1) -​ F(t) > 0 holds for any e < e*. As shown in Proposition 7.2, if 1 < (1/​n)F, an equilibrium effort e is strictly decreasing in the probability that the individual holds the

180  Governing the commons

pro-​environmental identity, x. Thus, we can consider e* = e(x*) such that e(x) < e(x*) holds for x > x*. Proposition 7.3 below shows that resource sustainability can be attained by increasing the probability that he or she holds the pro-​environmental identity. Proposition 7.3: Suppose that 1 < (1/​n)F(t) for t > 0. F(t + 1) -​F(t) > 0 if and only if x > x*, where e(x*) = 1/​2. Proof. If 1 < (1/​n)F(t), there is a positive solution e(t) > 0 to (7.5) for t > 0. From (7.7), F(t + 1) -​F(t) = n(1 -​e(t)) -​ne(t) = n(1 − 2e(t)). Consider x* such that e(x*) = 1/​2. Since sign{de/​dx} in (7.6) is negative, 1/​2 > e(x) for x > x*. Therefore, for any t, F(t + 1) -​F(t) > 0 if and only if x > x*. ■ Open access common-​pool resources are prone to overextraction. The required institutions are absent in many settings because of a lack of governance. Rather, individuals can devise their own enforceable institutional arrangements to escape the overexploitation of the commons. Bottom-​up processes are more appealing than authoritarian enforcement. Ostrom (1990) shows how local communities may succeed in overcoming the tragedy of the commons. Local problem-​solving does a better job of producing cooperation than the external imposition of rules. Local spheres must have some autonomy for systems to function properly; local actors resolve problems within their own sphere. For example, fishers using the same ocean grounds might come up with better and more durable solutions to the depletion of stocks than the set of rules imposed by state bureaucrats. The community-​ based governance is self-​regulating because individual interests are balanced against the collective interest for environmental sustainability. Some agents have been able to avoid destroying their own resource base, while others have not, which raises the question of what differences might exist between these communities.6 Resource conservation may be seen as a burden entailing costs due to foregone resource-​use opportunities. Then, financial incentives may be seen as critical to offsetting these costs. However, some recent studies demonstrate that common-​ pool resources are often successfully and sustainably governed without financial incentives. Especially, field experiments in many communities show that resource users balance self-​interest with conformity (e.g., Kerr et al., 2012). Individuals who have social preferences tend to restrain their harvesting in attempts to preserve the resource. Communities composed of such individuals do not overharvest the resource even when there are benefits from overusage and monitoring is nearly nonexistent. It is not correct to assume that preferences are fixed and exogenous, especially in relation to the environment. People making voluntary contributions to public goods cannot be explained solely by pure self-​interest. Individuals do not follow a purely self-​interested strategy, but rather strike a balance between self and group interests

Governing the commons  181

7.3  Internal moral constraints and environmental sustainability Preferences are subjective. There is no expectation that others would or should feel the same way. Therefore, there may be individual variation in the degree that people feel as preferences from the resource conservation. People’s sense of morality may differ from their social or moral preferences. On the other hand, people have a moral conviction that their moral standards ought to apply as universal ones that others also share. Some people may gravitate toward subjective beliefs that allow them to pursue their self-​interest. If morality is from preferences, they will behave less morally. But if morality is from an internal constraint, they still believe in certain moral rules. They adopt morality from different reasons. This section develops the idea of a Kantian approach to moral behavior in an environmental context. When people have some propensity to act morally, they will cut back consumption of an environmentally harmful good. Where does the propensity to act morally come from? In Kant’s moral philosophy, certain actions are prescribed or forbidden as matters of duty, regardless of the consequences of the action. These duties are based on the categorical imperative, the principle that is used to judge plans of action. Duty must be followed out of duty itself, not out of expectations of favorable outcomes for oneself or others. Following Kant’s categorical imperative, moral duties are seen as absolute that completely supersede any consideration of utility. Individuals recognize that acting morally means they will suffer a loss of personal utility. The identity with internal moral constraints follows a moral rule imposing an extra moral cost on the resource extraction. The identity with internal moral constraints abstains from engaging in the harvesting activity, by imposing an extra moral cost per unit activity in the resource extraction defined by m, where 0 < m < 1. It is interpreted as the degree that the identity with internal moral constraints abstains from harvesting by following a moral rule. Let UMC denote the payoff for the identity with internal moral constraints. The identity with internal moral constraints, guided by a moral rule, is defined as follows: Definition 7.3: An individual holds the identity with internal moral constraints if the utility function is given by: UMC = b(e) -​c(e) -​me, (7.8) where m is a moral cost per unit activity in the resource extraction. In this section, there are two different identities that an individual can adopt: the identity with moral preferences (Definition 7.2) and an agent with moral constraints (Definition 7.3).

182  Governing the commons

The timing of the situation is the following: [Stage 1] The individual adopts either the identity with moral preferences (UPE) or the identity with internal moral constraints (UMC). [Stage 2] The individual chooses an effort e in the resource extraction. At Stage 1, each individual gets an opportunity to adopt either the identity with moral preferences or the identity with internal moral constraints. Let y be the individual’s subjective belief about choosing the identity with moral preferences, where 0 < y < 1.With probability y, the individual adopts the identity with moral preferences; with probability 1 -​y, he or she adopts the identity with internal moral constraints. At Stage 2, each individual chooses his or her action at Stage 2, taking into account his or her sense of identity. Using (7.2) and (7.8), the individual’s expected utility at Stage 2 is written as: (1 -​y)UMC + yUPE = (1 -​y){b(e) -​c(e) -​me} + y{b(e) + k(1 -​e) -​c(e)} = b(e) -​c(e) + yk(1 -​e) -​(1 -​y)me. (7.9) Individual identity is composed as a weighted sum of two different types: the identity with moral preferences and the identity with internal moral constraints. The individual chooses the optimal level of effort in the resource extraction to maximize his or her utility (7.9). The problem that he or she faces at Stage 2 is: maxeb(e) -​c(e) + yk(1 -​e) -​(1 -​y)me. The first proposition states the existence of an equilibrium: Proposition 7.4: If 1 < (1/​n)F, there exists a unique equilibrium where an agent chooses a positive effort in the resource extraction. Otherwise, there exists no equilibrium in the interval ((1/​n)F, 1] and there may not always exist an equilibrium in the interval [0, (1/​n)F]. Proof. Suppose that 1 < (1/​n)F. Since e < 1, e < (1/​n)F holds. Then, b(e) = e from (7.1).Thus, from (7.9), the agent’s payoff is given by (1 − y)UMC + yUPE = e − c(e) + yk(1 − e) -​(1 -​y)me. By differentiating this payoff with respect to e, the first-​order condition for an interior solution is obtained as follows: 1 -​m -​y(k -​m) = c′(e). (7.10) First, we show that the left-​hand side of (7.10) is a positive constant. 1 -​m -​y(k -​m) > 1 -​m -​y(1 -​m) = (1 -​m)(1 -​y) > 0 since 0 < k < 1, 0 < y < 1, and 0 < m < 1.Thus, the left-​hand side of (7.10) is a positive constant. Next, since c′(e) > 0 and c′′(e) > 0 for e > 0, the right-​hand side of (7.10) is strictly increasing; by the assumption (lime→1c′(e) = ∞), c′(e) goes through the positive real line 1 -​m -​y(k -​m) as e goes from 0 to 1. Thus, there exists a unique value of e that satisfies (7.10) in the interval [0, 1]. This unique equilibrium satisfies the second-​order condition for a maximum. Now suppose that (1/​n)F < 1. Consider e in the interval ((1/​n)F, 1]. Then, b(e) = (1/​n)F from (7.1). Thus, from (7.9), the individual’s payoff is given by

Governing the commons  183

(1 − y)UMC + yUPE = (1/​n)F − c(e) + yk(1 − e) -​ (1 -​ y)me. Differentiating this payoff with respect to e yields -​yk -​ m(1 -​ y) = c′(e). Since the right-​hand side is strictly increasing, there exists no equilibrium in the interval ((1/​n)F, 1]. Consider e in the interval [0, (1/​n)F], where (1/​n)F < 1. Then, b(e) = e from (7.1). The agent’s payoff is given by (1 − y)UMC + yUPE = e − c(e) + yk(1 − e) -​(1 -​ y)me from (7.9). Differentiating this payoff with respect to e yields (7.10). As shown above, there is a unique equilibrium in the interval [0, 1]. However, an equilibrium may not always exist in the interval [0, (1/​n)F]. ■ For 1 < (1/​n)F in Figure 7.4, the curve c′(e) intersects the real line, 1 -​ m -​ y(k -​ m). Thus, if the resource stock is sufficiently large compared with the number of individuals, there is a unique equilibrium in the interval [0, 1]. On the other hand, for (1/​n)F < 1 (i.e., the resource stock is sufficiently small compared with the number of individuals), Figure 7.5 depicts the case in which the curve c′(e) does not intersect the real line, 1 -​m -​y(k -​m), in the interval [0, (1/​n)F]. For a steeper c′(e), there may exist an equilibrium in the interval [0, (1/​n)F] (Figure 7.6). However, as shown in Figure 7.5, it is not always true that an equilibrium exists in the interval [0, (1/​n)F]. Next, it is shown that holding the identity with moral preferences does not necessarily reduce the resource extraction. The subjective belief about the identity adoption affects the resource extraction and conservation. A further departure from conventional models of the commons is that the amount of effort expended on the resource extraction depends on the degree of moral preferences, k, and that of moral constraints, m, for the resource conservation.

c′ (e) c′ (e)

1 – m – y(k – m)

O FIGURE 7.4  An


equilibrium: 1 –​m –​y(k –​m) and 1 < (1/​n)F



184  Governing the commons c′ (e) c′ (e)

1 – m – y(k – m)

O FIGURE 7.5  No




equilibrium: 1 –​m –​y(k –​m) and (1/​n)F < 1

c′ (e) c′ (e)

1 – m – y(k – m)

O FIGURE 7.6  An



equilibrium: 1 –​m –​y(k –​m) and (1/​n)F < 1


Governing the commons  185

Proposition 7.5: Suppose that 1 < (1/​n)F. An equilibrium effort in the resource extraction, e, is strictly decreasing in the probability that adopts the identity with moral preferences, y, if k > m. However, e is strictly increasing in y if k < m. Proof. As shown in Proposition 7.4, if 1 < (1/​n)F, there exists a unique positive value of e that satisfies (7.10). Implicit differentiation of (7.10) with respect to y yields: de/​dy = −(k − m)(1/​c′′(e)).(7.11) Since c′′(e) > 0 for e > 0, the sign of de/​dy in (7.11) is negative if k > m; and it is positive if k < m. ■ Proposition 7.5 shows how the equilibrium value of e varies with the parameter y, depending on the relative magnitude of k (moral preferences) and m (moral constraints). The equilibrium resource extraction varies with the probability that the individual holds the identity with moral preferences. If moral preferences are greater in magnitude than moral constraints, the harvesting effort e decreases as y increases: de(y)/​dy < 0 if k > m. For a relatively large degree of moral preferences, the more the individual’s exposure to y, the more the effort in the resource conservation. As the probability that the individual holds the identity with moral preferences increases, he or she is expected to provide a lower activity level to the resource extraction. On the other hand, if moral preferences are less in magnitude than moral constraints, the harvesting effort e increases as y increases: de(y)/​ dy > 0 if k < m. For a relatively small degree of moral preferences, the more the individual’s exposure to y, the more the effort in the resource extraction. Conversely, as the probability that the individual holds the identity with internal moral constraints, 1 -​y, increases, he or she is expected to provide a lower activity level to the resource extraction. Thus, we can analyze the implications for the relative magnitude of moral preferences and moral constraints in the conservation of a common-​pool resource. The model is set in discrete time, and time is denoted by t = 0, 1, 2, … Resource extraction and conservation occur at discrete points in time. This section also considers a renewable resource with common-​pool characteristics, whose stock at any point in time t is given by F(t). If F(t + 1) -​ F(t) > 0 for all t > 0, resource sustainability can be achieved. The conservation activity regrows the renewable resource, while effort in the extraction activity depletes the resource. The interaction between these components governs the dynamics of the model. We consider whether a threshold value, e*, exists such that F(t + 1) -​ F(t) > 0 holds for any e < e*. As shown in Proposition 7.5, if 1 < (1/​n)F, for a relatively large (small) degree of moral preferences, an equilibrium effort e is strictly decreasing (increasing) in the probability that the individual holds the identity with moral preferences, y. Thus, when k > ( ( 0. Let e(y*) = 1/​2. For k > m, F(t + 1) -​F(t) > 0 if and only if y > y*. On the other hand, for k < m, F(t + 1) -​F(t) > 0 if and only if y < y*. Proof. If 1 < (1/​n)F(t), there is a positive solution e(t) > 0 to (7.10) for t > 0. From (7.7), F(t + 1) -​F(t) = n(1 -​e(t)) -​ne(t) = n(1 − 2e(t)). Consider y* such that e(y*) = 1/​2. Since sign{de/​dy} in (7.11) is negative for k > m, 1/​2 > e(y) for y > y*. Therefore, for any t, F(t + 1) -​F(t) > 0 if and only if y > y*. On the other hand, since sign{de/​dy} in (7.11) is positive for k < m, 1/​2 > e(y) for y < y*. Therefore, for any t, F(t + 1) -​F(t) > 0 if and only if y < y*. ■ There exists no overexploitation of common-​ pool resources over time if F(t + 1) -​F(t) > 0 for any t. We consider whether a threshold value, e*, of resource extraction exists such that F(t + 1) -​ F(t) > 0 holds for any e < e*. For a relatively large degree of moral preferences, an equilibrium harvesting effort e strictly decreases if the individual exposes more to the identity with moral preferences (by a higher probability y > y*). On the other hand, for a relatively small degree of moral preferences, an equilibrium harvesting effort e strictly decreases if the individual exposes more to the identity with internal moral constraints (by a higher probability 1 -​y > 1 -​y* or a lower probability y < y*). Proposition 7.6 shows that resource sustainability can be attained by exposing more to the identity with moral preferences if moral preferences are greater in magnitude than moral constraints and more to the identity with internal moral constraints if moral preferences are less in magnitude than moral constraints. Many communities have been able to manage their common-​pool resources. However, a community that has a sufficiently number of individuals who hold the identity with moral preferences does not necessarily attain resource sustainability. For a relatively small degree of moral preferences, if individuals who hold the identity with moral preferences constitute a large share of the population, the community will not be successful in resource sustainability.

7.4  From Homo economicus to Homo moralis The metaphor of Homo economicus, the idea that people are self-​interested utility maximizers, represents a hard-​core assumption of neoclassical economics. Homo economicus is a single-​minded, utility-​maximizing automaton, who does not take into account other people. People making voluntary contributions to common good, such as blood donations, cannot be explained solely by self-​interest. The self-​ interest fails to explain the diversity in human behavior. People often perform pro-​ environmental actions at their own cost. The Homo-​economicus type of preferences fails to explain pro-​environmental behavior. Behavioral economics offers an alternative explanation of human action based on recognizing limited self-​interest.7 It

Governing the commons  187

has provided a great amount of evidence against the Homo-​economicus model of individual behavior. Subjects in the ultimatum game experiments deviate from the self-​interested choice: the vast majority of the offers to the responder are between 40% and 50% of the available surplus. They exhibit some form of other-​regarding preferences. An individual may be concerned about the payoffs other people receive. Behavior not complying with the standard model can be categorized as other-​regarding behavior when motives like fairness, reciprocity, altruism, and self-​ identity concerns (such as reputation and self-​respect) affect decisions. The prevalence of the Homo-​economicus model in economics has crowed out consideration of important noneconomic aspects of human nature, especially, the moral dimensions of human thought and conduct. Without honesty, trust, and goodwill, our economic life would grind to a halt. The morality accepted by individuals influences how they behave. By ethical behavior, we understand the urge to judge human actions as either good or bad. Human beings are fundamentally moral creatures and have moral values. An emerging alternative guiding metaphor in economics is Homo moralis. Homo moralis is a person, deeply embedded in social life, who takes into consideration moral issues and is not just driven by self-​interest. Human beings care about the feelings that fill the hearts of others. According to Adam Smith’s Theory of Moral Sentiments, a natural sense of fellow-​feeling is the thread that weaves the social fabric together. If our sentiments are in harmony with those of the person we observe, that sense of sentiments is what Smith’s Theory of Moral Sentiments refers to as sympathy. Our sympathy is the standard by which we assess the propriety (or impropriety) of others’ sentiments. Individuals may have various preferences, some reflecting self-​interest and others’ moral concerns. Although people usually prefer more wealth to less, moral considerations constrain this preference. In conventional economic theory, an agent has fixed, stable preferences, given constraints. Empirical evidence shows that individuals exhibit heterogeneity in their preferences (Falk et al., 2018). For example, people care for equitable outcomes, which is referred to as fairness. Reciprocity exists as a type of social preference. People react more cooperatively in response to friendly actions; they respond noncooperatively to hostile actions. Furthermore, as altruistic traits, people help others while making sacrifices. Third-​party punishment is the sanctioning of a wrongdoer that involves the action of peers not directly by the consequences of the rule violation.8 People may be willing to punish rule violators even in one-​shot interactions, because they derive nonpecuniary benefits from punishing.9 Moral sentiments often motivate people to incur costs that they could avoid. Cultural and social influences can undeniably affect human morality. The recent work on the evolution of cooperation suggests that human morality is first and foremost altruistic (e.g., Gintis et al., 2003). In the context of behavioral evolution of altruism, a person’s altruism is thought of not in terms of any act alone, but rather in terms of a pattern of altruistic acts extended over time. Even though an individual act of altruism is costly to the individual, an overall altruistic pattern may be highly valuable. The evolutionary game theory shows that cooperative behavior

188  Governing the commons

can survive in the long run. Preference evolution views preferences not as fixed and exogenous, but rather as subject to selection according to their fitness in an underlying context. In a model of preference evolution, Alger and Weibull (2013; 2016) show that a Homo-​moralis type of preference arises endogenously as the most favored by evolution.They identify an ‘evolutionarily stable preference’ by applying the definition of evolutionary stability to preference stability. The Homo-​moralis preferences emerge as evolutionarily stable under assortative matching. There exists a kind of negative synergy between economic incentives and moral behavior (Gneezy and Rustichini, 2000). Policies that appeal to economic self-​ interest may diminish ethical reasons for contributing to the common good. In various cases, people behave as if they were intrinsically motivated rather than stimulated by any monetary reward. Motivation refers to the reason on which one acts. People may behave altruistically because they are genuinely altruistic. They may do so because they are concerned about their social reputation and self-​respect. Then, the presence of monetary rewards spoils the reputational value of good deeds. Therefore, extrinsic, economic incentives may crowd out altruistic behavior. Various types of intrinsic motivation exist. Enjoyment derived from the activity may be conceptualized as intrinsic motivation. For example, people may enjoy a warm glow from contributing to a public good (Andreoni, 1989; 1990). While a pure altruist values the well-​being of other people, an impure altruist derives some utility from contributing, labeled warm glow. Instead, intrinsic motivations are evoked from social preferences such as guilt and shame. Lopez et al. (2012) investigate the relative effectiveness of the Guilt treatment and the Shame one in a public good field experiment in Colombia. In the experiment, each fisherman decides whether to help clean the beaches and wharves. In the Guilt treatment, individuals are privately reminded of their action and its impact on others. In the Shame treatment, on the other hand, information about an audited individual’s contribution decision is publicly revealed to the group. According to Lopez et al. (2012), while the Guilt treatment is not effective in improving outcomes, the Shane treatment significantly increases contributions and is more effective than the treatment with formal regulations. On the other hand, there exist circumstances under which actions ought to be guided by moral rules, regardless of the utility consequences. Roemer (2010; 2015) investigates what an equilibrium would look like in an economy where people are motivated by more cooperative concerns. Roemer describes Kantian equilibrium as a cooperative solution concept deriving from Kant’s categorical imperative. The distinction between altruism and cooperation based on Kantian ethics should be made forcefully. According to Roemer (2010, p. 18, original emphasis): We often attempt to instill Kantian ethics in children: “Don’t litter—​how would you like it if everyone else littered as you are doing?” Magical thinking—​ “If I donate, all others like me will also donate”—​although irrational, is perhaps quite common and will have the same effect. It is interesting that in these cases we do not ask individuals to pretend they were less selfish people, in

Governing the commons  189

the sense of taking on others’ preferences; rather, we ask them to evaluate from their own [selfish] viewpoint the consequences of non-​cooperative behavior. Intrinsic motivation may involve following a moral rule. This motivation stems from the duty that one must follow a particular moral rule.The goal is to act appropriately. Moral agents then make decisions on the basis of reason. Kant’s conception of practical reason holds that reason must determine ethical principles of conduct apart from considerations of inclination and interest. The agents are subject to moral judgment because of the maxims that they set for themselves. For Kantian ethics, social preferences are not truly moral. Kant would think sympathetic feelings unreliable. Kant famously begins the Groundworks of the Metaphysics of Morals by claiming that the only good without qualification is a good will. The good will cannot be grounded in particular desires, which are only conditionally good. His aim was to establish an a priori basis for ethical conduct, which is applicable in all circumstances. Kant is well known for emphasizing universality as a criterion of conformity to reason.10 An action has moral worth when it is done from the motive of duty, and its moral worth is determined by its underlying principle, rather than its results or intended consequences. There is an important distinction to be made between moral rules and moral principle in Kantian moral theory. The categorical imperative is not itself a moral rule—​it is an abstract formal principle.The first step of moral deliberation is the formulation of the rule that governs the specific action. Once the rule governing the action has been found, the agent then compares this formula with the requirements stipulated by the categorical imperative that the rule displays the characteristics of universality and necessity. If there is a supreme principle of morality, it is the categorical imperative. Whether an action is morally good depends on how it relates to this principle. A commitment to the principle is deeply embedded in Kantian moral philosophy. If the required formal characteristics of universality and necessity are discovered in the rule, it is judged by pure practical reason to be moral. If a duty results from application of the categorical imperative, it is moral. A society in which nobody has other-​regarding preferences or incorporates internal moral constraints is unlikely to be sustainable no matter how many laws exist. It is not correct to assume that people only cooperate for fear of external constraints such as fines. They appear to cooperate out of a greater sense of justice. People have the capacity to care about the morality of their actions, and their decisions are accordingly influenced by moral factors. For Hume and Smith, appropriate moral action is necessarily based on moral sentiments (the motivational importance of moral sentiments); for Kant, however, the emotional feelings that derive moral behavior are to be distrusted, and moral action consists of dutiful adherence to moral rules (a commitment to the principle of morality).The problem of motivation could be resolved by the fact that an agent has the psychological capacity such as sympathy to which moral considerations are related. However, this is done by appeal to the particular inclinations or purposes of agents. On Kant’s view, hypothetical imperatives direct an agent based on a particular desire.

190  Governing the commons

Inasmuch as people are bound by a moral rule, the moral rule must be a principle of reason. Consistent cooperation may arise from following a moral rule. The tendency to follow a moral rule may come from understanding one’s internal moral constraints. Thus, there are two moral components: moral preferences and internal moral constraints. Both moral preferences and internal moral constraints provide us with a better idea of why people cooperate. In a situation, if people can associate others who share similar moral preferences, the total amount of cooperation can increase. On the other hand, if people who are morally constrained follow moral rules, cooperation can be elicited. Appropriate moral action is based on the relative magnitude of moral preferences and moral sense of duty. A society can be constructive and sustainable to the degree that its members embody moral preferences and regulate themselves by moral rules.

Notes 1 More precisely, according to Ostrom et al. (1999), common-​pool resources are defined as natural or human-​constructed resource systems in which (1) exclusion of beneficiaries through physical and institutional means is especially costly and (2) exploitation by one user reduces resource availability for others. 2 In the voluntary contribution mechanism by Isaac and Walker (1988), each member of a group receives an endowment of money and then the members simultaneously decide to contribute some proportion of their endowment to the group account. These contributions benefit all group members.The endowment left over goes to the individual’s private account.The dominant strategy is then to place all of one’s endowment in the private account. However, for the social optimum to be attained, all individuals need to contribute their entire endowment to the group account. The percentage of the endowment placed in the group account is interpreted as a measure of cooperation in this mechanism. 3 The practice of community-​based natural resource management focuses on the engagement of local communities in natural resource decision-​making and use, which may include forms of ownership and community tenure. One of its basic premises is that local communities may more sustainably manage natural resources than centralized governments. 4 This section is in part based on Teraji’s (2019) model. 5 Sethi and Somanathan (1996) study the evolution of three strategies, namely defecting, cooperating, and enforcing, among agents interacting in a common-​pool resource game. The evolutionary process is described by a replicator dynamics equation, and differential survival is proportional to the relative performance value of each strategy. 6 It would be more reasonable to describe individuals who are endowed with different behavioral characteristics. Then one could express heterogeneity in strategies through different effort levels. Olson (1965) finds that certain types of inequality have a positive effect on collective action. As Olson (1965) argues, an increase in heterogeneity will make rich users internalize more the consequences of their actions, and they will invest less to use the collective good. However, on the other hand, this increase will diminish the propensity of poor users to take into account the impact on the resource that they may cause. 7 For example, individuals exhibit inequality aversion, as in Fehr and Schmidt (1999). Further, Rabin (1993) develops the concept of fairness equilibrium: it is a set of strategies and beliefs on other players’ intentions such that (i) each player plays the best reply to the

Governing the commons  191

other players’ strategies given his or her beliefs on others’ intentions and (ii) those beliefs are actually verified. Rabin (1993) shows that once we allow for reciprocal motivations on the part of the players, it is possible to think of the public good game as a coordination problem with full contribution being an efficient equilibrium and full free-​r iding an inefficient equilibrium with other equilibrium in between. 8 Third-​party punishment is defined in contrast to second-​party punishment, where the sanctioning individual is directly harmed by the wrongdoing (Gintis, 2000). 9 For example, in a public good experiment by Fehr and Gächter (2002), 84.3% of subjects use costly punishment (with 74.2% of punishment cases targeted at below-​average contributors) even though participants know they will never interact again with the same individuals. 10 Kant can be construed as claiming that life should be understood from two different perspectives. Human life is, in one viewpoint, understood as belonging to the sensible world. From this perspective, the individual operates within the context of the laws of nature. Human beings, in the other viewpoint, conceive of themselves as free from external. From this perspective, the individual uses the power of reason to discern the concepts that put the ideas of sense under a meaningful set of rules.


Afriat, S. N. (1967). “The construction of utility functions from expenditures data.” International Economic Review 8: 67–​77. Akerlof, G. A. (1970). “The market for ‘lemons’: quality uncertainty and the market mechanism.” Quarterly Journal of Economics 84: 488–​500. Akerlof, G. A. (1980). “A theory of social customs, of which unemployment may be one consequence.” Quarterly Journal of Economics 94: 49–​75. Akerlof, G. A. (1982). “Labor contracts as partial gift exchange.” Quarterly Journal of Economics 97: 543–​569. Akerlof, G. A., and Kranton, R. E. (2000). “Economics and identity.” Quarterly Journal of Economics 115: 715–​753. Alexander, R. D. (1987). The Biology of Moral Systems. New York: Aldine De Gruyter. Alger, I., and Weibull, J. W. (2013). “Homo moralis—​preference evolution under incomplete information and assortative matching.” Econometrica 81: 2269–​2302. Alger, I., and Weibull. J. W. (2016). “Evolution and Kantian morality.” Games and Economic Behavior 98: 56–​67. Allais, M. (1953). “ Le comportement de l’homme rationnnel devant le risque: critique des postulats et axioms de l’ecole Americaine.” Econometrica 21: 503–​546. Allcott, H. (2011). “Social norms and energy conservation.” Journal of Public Economics 95: 1082–​1095. Allcott, H., and Taubinsky, D. (2015). “ Evaluating behaviorally motivated policy: experimental evidence from the lightbulb market.” American Economic Review 105: 2501–​2538. Allingham, M. G., and Sandmo, A. (1972). “Income tax evasion: a theoretical analysis.” Journal of Public Economics 1: 323–​338. Alm, J., and Torgler, B. (2006). “Culture differences and tax morale in the United States and in Europe.” Journal of Economic Psychology 27: 224–​246. Altman, M. (2000). “A behavioral model of path dependency: the economics of profitable inefficiency and market failure.” Journal of Socio-​Economics 29: 127–​145. Altman, M. (2005). “The ethical economy and competitive markets: reconciling altruistic, moralistic, and ethical behavior with the rational economic agents and competitive markets.” Journal of Economic Psychology 26: 732–​757.

Bibliography  193

Andreoni, J. (1989). “Giving with impure altruism: applications to charity and Ricardian equivalence.” Journal of Political Economy 97: 1447–​1458. Andreoni, J. (1990). “Impure altruism and donations to public goods: a theory of warm-​glow giving.” Economic Journal 100: 464–​477. Andreoni, J. (1995). “Cooperation in public-​goods experiments: kindness or confusion?” American Economic Review 85: 891–​904. Andreoni, J., and Bernheim, B. D. (2009). “Social image and the 50-​50 norm: a theoretical and experimental analysis of audience effects.” Econometrica 77: 1607–​1636. Arrow, K. (1950). “A difficulty in the concept of social welfare.” Journal of Political Economy 58: 328–​346. Arthur, W. B. (1994). Increasing Returns and Path Dependency in the Economy. Ann Arbor, MI: University of Michigan Press. Ashraf, N., Camerer, C. F., and Loewenstein, G. (2005). “Adam Smith, Behavioral economist.” Journal of Economic Perspectives 19: 131–​145. Atkinson, A. B. (1995). Public Economics in Action: The Basic Income/​Flat Tax Proposal. Oxford: Clarendon Press. Aumann, R. J. (1974). “Subjectivity and correlation in randomized strategies.” Journal of Mathematical Economics 1: 67–​96. Aumann, R. J. (1987). “Correlated equilibrium as an expression of Bayesian rationality.” Econometrica 63: 1161–​1180. Axelrod, R. (1984). The Evolution of Cooperation. New York: Basic Books. Bagnoli, M., and Watts, S. G. (2003). “Selling to socially responsible consumers: competition and the private provision of public goods.” Journal of Economics and Management Strategy 12: 419–​445. Baron, D. P. (2001). “Private politics, corporate social responsibility, and integrated strategy.” Journal of Economics and Management Strategy 10: 7–​45. Baron, D. P. (2008). “Managerial contracting and corporate social responsibility.” Journal of Public Economics 92: 268–​288. Baumol,W. J. (1990). “Entrepreneurship: productive, unproductive, and destructive.” Journal of Political Economy 98: 893–​921. Becker, G. S. (1968). “Crime and punishment: an economic approach.” Journal of Political Economy 76: 169–​217. Becker, G. S. (1976). “Altruism, egoism, and genetic fitness: economics and sociobilology.” Journal of Economic Literature 4: 817–​826. Becker, G. S., and Murphy, K. M. (1988). “A theory of rational addiction.” Journal of Political Economy 96: 675–​700. Bénabou, R., and Tirole, J. (2004). “Willpower and personal rules.” Journal of Political Economy 112: 848–​886. Bénabou, R., and Tirole, J. (2006). “Incentives and prosocial behavior.” American Economic Review 96: 1652–​1678. Benartzi, S., and Thaler, R. (2004). “Save more tomorrow: using behavioral economics in increase employee savings.” Journal of Political Economy 112: 164–​187. Bentham, J. ([1781] 1988). The Principles of Morals and Legislation. NewYork: Prometheus Books. Bergson, A. (1938). “A reformulation of certain aspects of welfare economics.” Quarterly Journal of Economics 52: 310–​334. Bergstrom, T., Blume, L., and Varian, H. (1986). “On the private provision of public goods.” Journal of Public Economics 29: 25–​49. Bernheim, B. D. (2009). “Behavioral welfare economics.” Journal of the European Economic Association 7: 267–​319.

194 Bibliography

Bernheim, B. D. (2016). “The good, the bad, and the ugly: a unified approach to behavioral welfare economics.” Journal of Benefit-​Cost Analysis 7: 12–​68. Besley, T. J., and Ghatak, M. (2007). “Retailing public goods: the economics of corporate social responsibility.” Journal of Public Economics 91: 1645–​1663. Bhattacharya, C. B., and Sen, S. (2003). “Consumer–​company identification: a framework for understanding consumers’ relationships with companies.” Journal of Marketing 67: 76–​88. Bicchieri, C. (2006). The Grammar of Society: The Nature and Dynamics of Social Norms. Cambridge, MA: Cambridge University Press. Blau. P. M. (1964). Exchange and Power in Social Life. New York: John Wiley. Bowen, H. R. (1953). Social Responsibilities of the Businessman. New York: Harper and Brothers. Bowles, S. (1998). “Endogenous preferences: the cultural consequences of markets and other economic institutions.” Journal of Economic Literature 36: 75–​111. Bowles, S. (2016). The Moral Economy:Why Good Incentives Are Not Substitutes for Good Citizens. London:Yale University Press. Bowles, S., and Hwang, S.-​H. (2008). “Social preferences and public economics: mechanism design when social preferences depend on incentives.” Journal of Public Economics 92: 1811–​1820. Boyer, R., and Orléan, A. (1992). “How do conventions evolve?” Journal of Evolutionary Economics 2: 165–​177. Boyd, R., and Richerson, P. J. (1989). “The evolution of indirect reciprocity.” Social Networks 11: 213–​236. Brown, T. J., and Dacin, P. A. (1997). “The company and the product: corporate associations and consumer product responses,” Journal of Marketing 61: 68–​84. Buchanan, J. M. (1966). “Joint supply, externality and optimality.” Economica 33: 404–​415. Camerer, C. F. (2003). Behavioral Game Theory: Experiments in Strategic Interaction. Princeton, NJ: Princeton University Press. Caplin, A., Dean, M., and Martin, D. (2011). “Search and satisficing.” American Economic Review 101: 2899–​2922. Carpenter, C. (2007). “Heavy alcohol use and crime: evidence from underage drunk-​driving laws.” Journal of Law and Economics 50: 539–​557. Charness, G., and Rabin, M. (2002). “Understanding social preferences with simple tests.” Quarterly Journal of Economics 117: 817–​869. Chaudhuri, A. (2011). “Sustaining cooperation in laboratory public goods experiments: a selective survey of the literature.” Experimental Economics 14: 47–​83. Cialdini, R. B., Reno, R. R., and Kallgren, C. A. (1990). “A focus theory of normative conduct: recycling the concept of norms to reduce littering in public places.” Journal of Personality and Social Psychology 58: 1015–​1026. Cialdini, R. B., and Trost, M. B. (1998). “Social influence; social norms, conformity and compliance.” In: Gilbert, G., Fiske, S., and Lindzey, G. (eds.), The Handbook of Social Psychology, vol. 2. New York: McGraw-​Hill, pp. 151–​192. Coase, R. H. (1960). “The problem of social cost.” Journal of Law and Economics 3: 1–​44. Coleman, J. (1990). Foundations of Social Theory. Cambridge, MA: Harvard University Press. Commission of the European Communities (2001). Green Paper: Promoting a European Framework for Corporate Social Responsibility. Brussels: Commission of the European Communities. Cyert, R. M., and March, J. G. (1963). A Behavioral Theory of the Firm. Englewood Cliffs, NJ: Prentice Hall. Dasgupta, P. (2001). Human Well-​Being and the Natural Environment. Oxford: Oxford University Press.

Bibliography  195

de Jong, J. P., and Marsili, O. (2015). “The distribution of Schumpeterian and Kirznerian opportunities.” Small Business Economics 44: 19–​35. Diamond, P. A. (1973). “Consumption externalities and corrective imperfect pricing.” Bell Journal of Economics 4: 526–​538. Dreibelbis, R., Kroeger, A., Hossain, K., Venkatesh, M., and Ram, P. K. (2016). “Behavior change without behavior change communication: nudging handwashing among primary school students in Bangladesh.” International Journal of Environmental Research and Public Health 13(1): 129. Easterlin, R. A. (1995). “Will raising the incomes of all increase the happiness of all?” Journal of Economics Behavior and Organization 27: 35–​47. Ellsberg, D. (1961). “Risk, ambiguity, and the Savage axioms.” Quarterly Journal of Economics 75: 643–​669. Engelman, D., and Fischbacher, U. (2009). “Indirect reciprocity and strategic reputation building in an experimental helping game.” Games and Economic Behavior 67: 399–​407. Evans, A. T., Peters, E., Strasser, A. A., Emery, L. F., Sheerin. K. M., and Romer, D. (2015). “Graphic warning labels elicit affective and thoughtful responses from smokers: results of a randomized clinical trial.” PloS One 10: e0142879. Evans, J. S. B. T. (1989). Bias in Human Reasoning: Causes and Consequences. London: Erlbaum. Falk, A., Becker, A., Dohmen, T., Enke, B., Huffman, D., and Sunde, U. (2018). “Global evidence on economic preferences.” Quarterly Journal of Economics 133: 1645–​1692. Fehr, E., and Fischbacher, U. (2004). “Third party punishment and social norms.” Evolution and Human Behavior 25: 63–​87. Fehr, E., and Fischbacher, U. (2005). “Human altruism—​proximate patterns and evolutionary origins.” Analyse Kritik 27: 6–​27. Fehr, E., and Gächter, S. (2000).“Cooperation and punishment in public goods experiments.” American Economic Review 90: 980–​994. Fehr, E., and Gächter, S. (2002). “Altruistic punishment in humans.” Nature 415: 137–​140. Fehr, E., Gächter, S., and Kirchsteiger, G. (1997). “Reciprocity as a contract enforcement device: experimental evidence.” Econometrica 65: 833–​860. Fehr, E., Kirchsteiger, G., and Reidl, A. (1993). “Does fairness prevent market clearing? An experimental investigation.” Quarterly Journal of Economics 108: 437–​460. Fehr, E., and Schmidt, K. M. (1999). “A theory of fairness, competition, and cooperation.” Quarterly Journal of Economics 114: 817–​868. Foss, N. J., and Klein, P. G. (2012). Organizing Entrepreneurial Judgment: A New Approach to the Firm. Cambridge: Cambridge University Press. Freeman, E. (1984). Strategic Management: A Stakeholder Approach. Boston: Pitman Publishing. Frey, B. S. (2003). “Deterrence and tax morale in the European Union.” European Review 11: 385–​406. Frey, B. S., and Oberholzer-​Gee, F. (1997). “The cost of price incentives: an empirical analysis of motivation crowding-​out.” American Economic Review 87: 746–​755. Friedman, M. (1962). Capitalism and Freedom. Chicago: University of Chicago Press. Friedman, M. (1970). “The social responsibility of business is to increase its profits.” New York Times Magazine 13: 32–​33, 122–​126. Gigerenzer, G., and Selten, R. (eds.). (2001). Bounded Rationality: The Adaptive Tool-​Box. Cambridge, MA: The MIT Press. Gigerenzer, G.,Todd. P. M., and the ABC Research Group. (1999). Simple Heuristics That Make Us Smart. Oxford: Oxford University Press. Gigerenzer, G., Hoffrage, U., and Goldstein, D. G. (2008). “Fast and frugal heuristics are plausible models of cognition: reply to Dougherty, Franco-​Watkins, and Tomas (2008).” Psychological Review 115: 230–​239.

196 Bibliography

Giné, X., Karlan, D., and Zinman, J. (2010). “Put your money where your butt is: a commitment contract for smoking cessation.” American Economic Journal: Applied Economics 2: 213–​235. Gintis, H. (2000). “Beyond Homo economicus: evidence from experimental economics.” Ecological Economics 35: 311–​322. Gintis, H. (2009). The Bounds of Reason: Game Theory and the Unification of the Behavioral Sciences. Princeton, NJ: Princeton University Press. Gintis, H., Bowles, S., Boyd, R., and Fehr, E. (2003). “Explaining altruistic behavior in humans.” Evolution and Human Behavior 24: 153–​172. Gintis, H., Bowles, S., Boyd, R., and Fehr, E. (eds.). (2005). Moral Sentiments and Material Interests:The Foundations of Cooperation in Economic Life. Cambridge, MA: The MIT Press. Gneezy, U., and Rustichini, A. (2000). “ A fine is a price.” Journal of Legal Studies 29: 1–​17. Goldstein, N. J., Cialdini, R. B., and Griskevicius,V. (2008). “A room with a viewpoint: using social norms to motivate environmental conservation in hotels.” Journal of Consumer Research 35: 472–​482. Gruber, J., and Köszegi, B. (2001). “Is addiction ‘rational’? Theory and evidence.” Quarterly Journal of Economics 116: 1261–​1303. Grüne-​Yanoff, T. (2012). “Old wine in new casks: libertarian paternalism still violates liberal principles.” Social Choice and Welfare 38: 635–​645. Gul, F., and Pesendorfer,W. (2001).“Temptation and self-​control.” Econometrica 69: 1403–​1435. Gul, F., and Pesendorfer,W. (2004).“Self-​control and the theory of consumption.” Econometrica 72: 119–​158. Gul, F., and Pesendorfer, W. (2007). “Welfare without happiness.” American Economc Review 97: 471–​476. Güth, W., Schmittberger, R., and Schwarze, B. (1982). “An experimental analysis of ultimatum bargaining.” Journal of Economic Behavior and Organization 3: 367–​388. Hallsworth, M., List, J., Metcalfe, R., and Vlaev, I. (2017). “The behavioralist as tax collector: using natural field experiments to enhance tax compliance.” Journal of Public Economics 148: 14–​31. Hardin, G. (1968). “The tragedy of the commons.” Science 162: 1243–​1248. Harsanyi, J. C. (1955). “Cardinal welfare, individualistic ethics, and interpersonal comparisons of utility.” Journal of Political Economy 63: 309–​321. Hausman, D. M. (2012). Preferences, Value, Choice, and Welfare. Cambridge: Cambridge University Press. Hayek, F. A. (1948). Individualism and Economic Order. Chicago: University of Chicago Press. Hayek, F. A. (1973). Law, Legislation, and Liberty, Vol. 1, Rules and Order. Chicago: University of Chicago Press. Henrich, J. (2000). “Does culture matter in economic behavior? Ultimatum game bargaining among the Machiguenga of the Peruvian Amazon.” American Economic Review 90: 973–​979. Henrich, J., Boyd, R., Bowles, S., Camerer, C., Fehr, E., and Gintis, H. (2004). Foundations of Human Sociality: Economic Experiments and Ethnographic Evidence From Fifteen Small-​scale Societies. Oxford: Oxford University Press. Henrich, J., Boyd, R., Bowles, S., Camerer, C., Fehr, E., Gintis, H., and McElreath, R. (2001). “In search of Homo economicus: behavioral experiments in 15 small-​scale societies.” American Economic Review 91: 73–​78. Hossain, T., and List, J. (2012). “The behavioralist visits the factory: increasing productivity using simple framing manipulations.” Management Science 58: 2151–​2167. Houthakker, H. (1950).“Revealed preference and the utility function.” Economica 17: 159–​174.

Bibliography  197

Huberman, G., Iyengar, S., and Jiang, W. (2007). “Defined contribution pension plans: determinants of participation and contributions rates.” Journal of Financial Services Research 31: 1–​32. Hume, D. ([1739–​1740] 1983). A Treatise of Human Nature. Oxford: Clarendon Press. Isaac, R. M., and Walker, J. M. (1988). “Group size effects in public goods provision: the voluntary contribution mechanism.” Quarterly Journal of Economics 103: 179–​200. Jaffe, A. B., and Stavins, R. N. (1994). “The energy-​efficiency gap: what does it mean?” Energy Policy 22: 804–​810. Jevons, W. S. ([1871] 1965). The Theory of Political Economy, 5th ed. New York: A. M. Kelley. Johnson, E. J., and Goldstein, D. (2003). “Do defaults save lives?” Science 302: 1338–​1339. Kahneman, D. (2003). “Maps of bounded rationality: psychology for behavioral economics.” American Economic Review 93: 1449–​1475. Kahneman, D. (2011). Thinking, Fast and Slow. New York: Farrar, Straus and Giroux. Kahneman, D., Knetsch, J. L, and Thaler, R. H. (1990). “Experimental tests of the endowment effect and the Coase theorem.” Journal of Political Economy 98: 1325–​1348. Kahneman, D., and Tversky, A. (1972). “Subjective probability: a judgement of representativeness.” Cognitive Psychology 3: 430–​454. Kahneman, D., and Tversky, A. (1973). “On the psychology of prediction.” Psychological Review 80: 237–​251. Kahneman, D., and Tversky, A. (1974). “Judgment under uncertainty: heuristics and biases.” Science 185: 1124–​1131. Kahneman, D., and Tversky, A. (1979). “Prospect theory: an analysis of decision under risk.” Econometrica 47: 263–​291. Kahneman, D., Wakker, P., and Sarin, R. (1997). “Back to Bentham? Explorations of experienced utility.” Quarterly Journal of Economics 112: 375–​406. Kant, I. (1785). Grundlegung zur Metaphysik der Sitten. Cambridge: Cambridge University Press. Translated by Mary J. Gregor: Groundwork of the Metaphysics of Morals, 1988. Kerr, J,Vardhan, M., and Jindal, R. (2012). “Prosocial behavior and incentives: evidence from field experiments in rural Mexico and Tanzania.” Ecological Economics 73: 220–​227. Kihlstrom, R. E., and Laffont, J.-​J. (1979). “A general equilibrium entrepreneurial theory of firm formation based on risk aversion.” Journal of Political Economy 87: 719–​748. Kirzner, I. M. (1973). Competition and Entrepreneurship. Chicago: University of Chicago Press. Kirzner, I. M. (1979). Perception, Opportunity, and Profit. Chicago: University of Chicago Press. Kirzner, I. M. (1981). “Mises on entrepreneurship.” Wirtschaftspolitische Blätter 28: 51–​57. Kirzner, I. M. (1992). The Meaning of Market Process: Essays in the Development of Modern Austrian Economics. London: Routledge. Kirzner, I. M. (1997). “Entrepreneurial discovery and the competitive market process: an Austrian approach.” Journal of Economic Literature 35: 60–​85. Kirzner, I. M. (1999). “Creativity and/​or alertness: a reconsideration of the Schumpeterian entrepreneur.” Review of Austrian Economics 11: 5–​17. Knight, F. H. (1921). Risk, Uncertainty, and Profit. Boston: Houghton Mifflin. Knudsen,T. (2002).“Economic selection theory.” Journal of Evolutionary Economics 12: 443–​470. Koppel, H., and Regner,T. (2014). “Corporate social responsibility in the workplace: experimental evidence from a gift-​exchange game.” Experimental Economics 17: 347–​370. Lachmann, L. M. (1970). The Legacy of Max Weber. London: Heinnemann. Lachmann, L. M. (1976). “From Mises to Shackle: an essay on Austrian economics and the Kaleidic society.” Journal of Economic Literature 14: 54–​62. Laffont, J.-​J. (1975). “Macroeconomic constraints, economic efficiency and ethics: an introduction to Kantian economics.” Economica 42: 430–​437.

198 Bibliography

Laibson, D. (1997). “Golden eggs and hyperbolic discounting.” Quarterly Journal of Economics 112: 443–​478. Lange, O. (1942). “The foundations of welfare economics.” Econometrica 10: 215–​228. Langlois, R. N. (2005). “The entrepreneurial theory of the firm and the theory of the entrepreneurial firm.” Journal of Management Studies 44: 1107–​1124. Leibenstein, H. (1966). “Allocative efficiency vs. ‘X-​efficiency.’ ” American Economic Review 56: 392–​415. Leibenstein, H. (1978). General X-​Efficiency Theory and Economic Development. Oxford: Oxford University Press. Leibenstein, H. (1979). “A branch of economics is missing: micro-​micro theory.” Journal of Economic Literature 17: 477–​502. Leibenstein, H. (1987). Inside the Firm:The Inefficiencies of Hierarchy. Cambridge, MA: Harvard University Press. Lewis, D. (1969). Convention: A Philosophical Study. Cambridge, MA: Harvard University Press. Lopez, M. C., Murphy, J. J., Spraggon, J. M., and Stranlund, J. K. (2012). “Comparing the effectiveness of regulation and prosocial emotions to enhance cooperation: experimental evidence from fishing communities in Colombia.” Economic Inquiry 50: 131–​42. Lourenço, J. S., Ciriolo, E., Almeida, S. R., and Troussard, X. (2016). “Behavioural insights applied to policy.” European Report 2016. EUR 27726 EN. https://​​10.2760/​ 903938. Luca, D. L., Owens, E., and Sharma, G. (2015). “Can alcohol prohibition reduce violence against women?” American Economic Review 105: 625–​629. Luttmer, E. F. P., and Singhal, M. (2014). “Tax morale.” Journal of Economic Perspectives 28: 149–​168. Madrian, B. C., and Shea, D. F. (2001). “The power of suggestion: inertia in 401(k) participation and savings behavior.” Quarterly Journal of Economics 116: 1149–​1187. March, J. G., and Simon, H. A. (1958). Organization. New York: John Wiley. Markose, S. M. (2005). “Computability and evolutionary complexity: markets as complex adaptive systems (CAS).” Economic Journal 115: F156–​F192. McAdams, R., and Rasmusen, E. B. (2007). “Norms in law and economics.” In: Polinsky, M., and Shavell, S. (eds.), Handbook of Law and Economics, vol. 2. Amsterdam: Elsevier, pp. 1573–​1618. Milgrom, P. R., and Roberts, J. (1992). Economics of Organization and Management. Englewood Cliffs, NJ: Prentice Hall. Mill, J. S. ([1869] 1991). On Liberty. London: Routledge. Mirrlees, J. (1971). “An exploration in the theory of optimum income taxation.” Review of Economic Studies 38: 175–​208. Mises, L. ([1949] 1966). Human Actions: A Treatise on Economic. Chicago: Henry Regnery. Mohr, L. A., Webb, D. J., and Harris, K. E. (2001). “Do consumers expect companies to be socially responsible? The impact of corporate social responsibility on buying behavior.” Journal of Consumer Affairs 35: 45–​72. Mullainathan, S., and Thaler, R. H. (2000). “Behavioral economics.” NBER Working Paper 7948. Myerson, R. B. (1982). “Optimal coordination mechanisms in generalized principal-​agent problems.” Journal of Mathematical Economics 10: 67–​81. Naylor, R. (1989). “Strikes, free riders, and social customs.” Quarterly Journal of Economics 104: 771–​785. Nelson, R. R., and Winter, S. (1982). An Evolutionary Theory of Economic Change. Cambridge, MA: Harvard University Press.

Bibliography  199

Newell, A., and Simon, H. A. (1972). Human Problem Solving. Englewood Cliffs, NJ: Prentice Hall. Nowak, M. A., and Sigmund, K. (1998). “The dynamics of indirect reciprocity.” Journal of Theoretical Biology 194: 561–​574. O’Donoghue, T., and Rabin, M. (2006). “Optimal sin taxes.” Journal of Public Economics 90: 1825–​1849. Olson, M. (1965). The Logic of Collective Action: Public Goods and the Theory of Groups. Cambridge, MA: Harvard University Press. Ostrom, E. (1990). Governing the Commons. New York: Cambridge University Press. Ostrom, E. (2000). “Collective action and the evolution of social norms.” Journal of Economic Perspective 14: 137–​158. Ostrom, E. (2005). Understanding Institutional Diversity. Princeton, NJ: Princeton University Press. Ostrom, E., Burger, J., Field, C. B., Norgaard, R. B., and Policansky, D. (1999). “Revisiting the commons: local lessons, global challenges.” Science 284: 278–​282. Pigou, A. C. (1920). The Economics of Welfare. London: McMillan. Polanyi, M. (1969). In: Grene, M. (ed.), Knowing and Being: Essays by Michael Polanyi. Chicago: University of Chicago Press. Posner, R. A. (1997). “Social norms and the law: an economic approach.” American Economic Review 87: 365–​369. Rabin, M. (1993). “Incorporating fairness into game theory and economics.” American Economic Review 83: 1281–​1302. Rabin, M. ([1995] 2019). “Moral preferences, moral constraints, and self-​serving biases.” Working Paper, University of California at Berkeley. Rawls, J. (1971). A Theory of Justice. Cambridge, MA: Harvard University Press. Rizzo, M. J. (2013). “Foundations of The Economics of Time and Ignorance.” Review of Austrian Economics 26: 45–​52. Rizzolatti, G., and Craighero, L. (2005).“Mirror neurons: a neurological approach to empathy.” In: Changeux, J.-​P., Demasio, A. R., Singer, W., and Cgristen, Y. (eds.), Neurobilology of Human Values. Berlin: Springer, pp. 107–​123. Robbins, L. (1932). An Essay on the Nature and Significant of Economic Science. NewYork: NewYork University Press. Roemer, J. E. (2010). “Kantian equilibrium.” Scandian Journal of Economics 112: 1–​24. Roemer, J. E. (2015). “Kantian optimization: a microfoundation for cooperation.” Journal of Public Economics 127: 45–​57. Roth, A. E., Prasnikar,V., Okuno-​Fujiwara, M., and Zamir, S. (1991). “Bargaining and market behavior in Jerusalem, Ljubljana, Pittsburgh, and Tokyo: an experimental study.” American Economic Review 81: 1068–​1095. Samuelson, P. A. (1937). “A note on measurement of utility.” Review of Economic Studies 4: 155–​161. Samuelson, P. A. (1938). “A note on the pure theory of consumer’s behaviour.” Economica 5: 61–​71. Samuelson, P. A. (1947 [1983]). Foundations of Economic Analysis. Cambridge, MA: Harvard University Press. Samuelson, P. A. (1954). “The pure theory of public expenditure.” Review of Economics and Statistics 36: 387–​389. Samuelson, W., and Zeckhauser, R. J. (1988). “Status quo bias in decision making.” Journal of Risk and Uncertainty 1: 7–​59. Schelling, T. C. (1960). The Strategy of Conflict. Cambridge, MA: Harvard University Press.

200 Bibliography

Schumpeter, J.A. ([1912] 1934). The Theory of Economic Development. Cambridge, MA: Harvard University Press. Sen, A. K. (1973). “Behavior and the concept of preferences.” Economica 40: 241–​259. Sen, A. K. (1982). Choice,Welfare and Measurements. Oxford: Basil Blackwell. Sen, A. K. (1993). “Internal consistency of choice.” Econometrica 61: 495–​521. Sen, A. K. (1997). “Maximization and the art of choice.” Econometrica 65: 745–​779. Sen, A, K. (2002). Rationality and Freedom. Cambridge, MA: Harvard University Press. Sethi, R., and Somanathan, E. (1996). “The evolution of social norms in common property resource use.” American Economic Review 86: 766–​788. Simon, H. A. (1955). “A behavioral model of rational choice.” Quarterly Journal of Economics 69: 99–​118. Simon, H. A. (1956). “Rational choice and the structure of the environment.” Psychological Review 63: 129–​138. Simon, H. A. (1957). Models of Man: Social and Rational. New York: Wiley. Simon, H. A. (1959). “Theories of decision-​making in economics and behavioral science.” American Economic Review 49: 253–​283. Simon, H. A. (1969). The Science of the Artificial. Cambridge, MA: The MIT Press. Simon, H. A. (1976). “From substantive to procedural rationality.” In: Latsis, S. J. (ed.), Method and Appraisal in Economics. Cambridge, MA: Cambridge University Press, pp. 129–​148. Simon, H. A. (1978). “Rationality as process and as product of thought.” American Economic Review 68: 1–​16. Simon, H. A. (1985). “Human nature in politics: the dialogue of psychology with political science.” American Political Science Review 79: 293–​304. Simon. H. A. (1990). “Invariants of human behavior.” Annual Review of Psychology 41: 1–​19. Simon, H. A. (1993). “Altruism and economics.” American Economic Review 83: 156–​161. Simon, H. A. (1997). Models of Bounded Rationality, Vol. III: Empirically Grounded Economic Reason. Cambridge, MA: The MIT Press. Sliwka, D. (2007). “Trust as a signal of a social norm and the hidden costs of incentive schemes.” American Economic Review 97: 999–​1012. Smith, A. ([1759] 1981). The Theory of Moral Sentiments. Indianapolis: Liberty Fund. Smith, V. L. (2008). Rationality in Economics: Constructivist and Ecological Forms. Cambridge, MA: Cambridge University Press. Spence, M. (1973). “Job market signaling.” Quarterly Journal of Economics 87: 355–​374. Stigler, G. J. (1961). “The economics of information.” Journal of Political Economy 69: 213–​225. Stiglitz, J. E. (1975). “The theory of ‘screening,’ education, and the distribution of income.” American Economic Review 65: 283–​300. Strotz, R. H. (1955). “Myopia and inconsistency in dynamic utility maximization.” Review of Economic Studies 23: 165–​180. Sugden, R. (1985). “Consistent conjectures, and voluntary contributions to public goods.” Journal of Public Economics 27: 117–​124. Sugden, R. (1989). “Spontaneous order.” Journal of Economic Perspectives 3: 85–​97. Sugden, R. (1995). “A theory of focal points.” Economic Journal 105: 533–​550. Teraji, S. (2007). “Morale and the evolution of norms.” Journal of Socio-​Economics 36: 48–​57. Teraji, S. (2009).“A model of corporate social performance: social satisfaction and moral conduct.” Journal of Socio-​Economics 38: 926–​934. Teraji, S. (2013). “A theory of norm compliance: punishment and reputation.” Journal of Socio-​ Economics 44: 1–​6. Teraji, S. (2019). “Identity switching and conservation on the commons.” International Review of Economics 66: 101–​113.

Bibliography  201

Thaler, R. H. (1981). “Some empirical evidence on dynamic inconsistency.” Economic Letters 8: 201–​207. Thaler, R. H. (1988). “Anomalies: the ultimatum game.” Journal of Economic Perspectives 2: 195–​206. Thaler, R. H. (1991). Quasi Rational Economics. New York: Russell Sage Foundation. Thaler, R. H., and Sunstein, C. R. (2003). “Libertarian paternalism.” American Economic Review 93: 175–​179. Thaler, R. H., and Sunstein, C. R. (2008). Nudge: Improving Decisions About Health,Wealth and Happiness. New York: Simon and Schuster. Tirole, J. (1996). “A theory of collective reputation (with applications to the persistence of corruption and to firm quality).” Review of Economic Studies 63: 1–​22. Titmuss, R. M. (1970). The Gift Relationship: From Human Blood to Social Policy. NewYork:Allen and Unwin. Tomer, J. F. (2007). “What is behavioral economics?” Journal of Socio-​Economics 36: 463–​479. Trivers, R. (1971). “The evolution of reciprocal altruism.” Quarterly Review of Biology 46: 35–​57. Tuomala, M. (1990). Optimal Income Tax and Redistribution. Oxford: Oxford University Press. Tversky, A., and Kahneman, D. (1971). “Belief in the law of small numbers.” Psychological Bulletin 76: 105–​110. Tversky, A., and Kahneman, D. (1974). “Judgment under uncertainty: heuristics and biases.” Science 185: 1124–​1131. Tversky, A., and Kahneman, D. (1981). “The framing of decisions and the psychology of choice.” Science 211: 453–​458. Tversky, A., and Kahneman, D. (1983). “Extensional versus intuitive reasoning: the conjunction fallacy in probability judgment.” Psychological Review 90: 293–​315. Vanderschraaf, P. (1995). “Conventions as correlated equilibrium.” Erkenntnis 42: 65–​87. Varian, H. (1982). “The nonparametric approach to demand analysis.” Econometrica 50: 945–​973. Vogel, D. (2005). The Market for Virtue:The Potential and Limits of Corporate Social Responsibility. Washington, DC: Brookings Institution Press. von Neumann, J., and Morgenstern, O. (1947). The Theory of Games and Economic Behavior. Princeton, NJ: Princeton University Press. Webster, Jr. F. E. (1975). “Determining the characteristics of the socially conscious consumer.” Journal of Consumer Research 2: 188–​196. White, M. D. (2013). The Manipulation of Choice: Ethics and Libertarian Paternalism. New York: Palgrave Macmillan. Zelmer, J. (2003). “Linear public goods experiments: a meta-​analysis.” Experimental Economics 6: 299–​310.


action 91 addiction 60, 102 administrative costs 53, 55 agency 64, 134 alertness 91, 93, 107 Allais paradox 40 altruism 15, 17, 21, 110, 115–​22 anchoring 94 ancillary conditions 34–​5 anomalies 9, 94 attitude 9–​10, 29, 38, 41, 74, 130, 134–​5, 166 Austrian economics 140 autonomy 99, 101, 171, 180 authority 66–​7, 82 awareness 75, 92, 129–​32, 149, 162 bargaining: Coasean 5; power 163 behavior: adaptive 82, 86; failure 9, 21, 33, 49, 73, 111; regularity of 138, 142, 147 behavioural insights team (BIT) 12 belief 29–​30, 50, 63, 74, 87, 89, 94, 115, 135, 141 bias 7, 14, 49, 73, 98 bottom-​up 180 brand loyalty 128 burden 67 capability 45–​6, 63, 165 categorical imperative 22–​3, 167–​8, 181, 188–​9 certainty effect 41

cheating 18, 68, 123 choice: architecture 14, 74, 99–​100; function 27–​9; set 13, 27, 32–​3 choreographer 148–​9 cognitive dissonance 174 climate change 4, 72, 171 collaboration 130–​2 commitment 23, 42, 106, 124, 128–​30, 153, 189 common-​pool resource 172–​5, 179–​80, 185–​6 common prior 148 commons 171 compensation 4, 21, 64, 116, 130–​2, 134 competition 12, 83, 92, 171 completeness 32, 39 conflict 61, 141–​2 conformity 143, 147, 152 consumer protection laws 2 contract 62–​6 contribution 3, 16, 20, 22, 62, 69, 74, 88, 93, 112, 123, 125, 140, 150–​1, 166, 180, 186, 188 convention 84, 119, 142–​3, 146–​8 cooperation 18, 115, 123–​4, 141–​2, 150–​1, 163–​5, 171, 173, 180, 187–​8, 190 coordination 138–​41 corporate social performance 127 corporate social responsibility 125 correlation device 148–​9 crowding-​out 109–​10 culture 161–​2

Index  203

decision-​making 7, 81–​2, 84, 86, 89, 98 default 7, 13, 33, 100–​1 deterrence 66–​7, 151 deviation 10–​12, 73, 93–​4, 113, 153, 157, 159, 168 dignity 115 discounting: exponential 104–​5; quasi-​hyperbolic 61, 105 discovery 65, 79, 90, 92–​3 discretion 84, 106, 121–​2 distortion 1, 4, 46, 56, 59, 70 docility 87 donation 13, 19, 21–​2, 110, 115, 125, 129 dual process models 11, 100 duty 22, 166–​7, 181 efficiency 1, 5, 8, 52, 56, 70–​4, 84, 91, 134, 150 effort variability 85, 106, 122 emergence 15, 133, 136, 140, 143, 145, 147 enforcement 3, 66–​9, 99, 151, 168, 173, 180 energy paradox 8, 72 entrepreneurship 88–​92 environment 70–​1, 74, 79, 83, 85–​7, 106, 129, 131, 134 envy 115–​22 equilibrium: correlated 147–​9; Kantian 168, 188; morale 117, 121–​2; Nash 117, 143, 145, 147–​8, 168; sub-​game perfect 15, 163–​4; sub-​optimal 106, 121–​2; Walrasian competitive 1 error 93–​5 ethics 22, 36, 69, 129, 133, 167–​8, 188–​9 evolution 11, 87, 116–​17, 121–​2, 155, 162, 187–​8, 190 externalities 58, 70 extraction 172, 174–​7, 179–​83, 185–​6 expectations 22, 81, 89, 108, 111, 129, 131, 134–​6, 139–​42, 145–​7, 149, 152, 170, 181 expected utility theory 38, 40–​1, 73, 78, 97 fairness 53, 112, 114–​15 fear 67, 96, 110, 124, 151, 168, 189 fellow-​feeling 16, 187 firm 2, 53, 82–​4, 126–​8, 130–​5 first best 4, 6, 53 focal point 145, 147, 170 framing 9–​10, 12–​13, 32–​4, 96, 100–​1 freedom 12–​14, 46, 61, 74, 99, 101 free riding 3, 19, 123, 137, 149, 151, 191 frequency 94, 111, 143–​5, 164 functionings 45–​6

gain 18, 95–​8 gamble 40–​1, 67–​8 game theory 142, 145, 147, 163, 187 generalized axiom of revealed preference 26 gift exchange 14, 66, 128, 136 goodwill 128, 187 government intervention 1, 3 greedy 74, 150 green 74, 129, 162 guilt 113, 188 habit 59, 102, 139, 142 happiness 46–​50 hedometer 36 heterogeneity 6, 89, 93, 113, 187, 190 heuristics 86, 93–​4, 107 history 37–​8, 106, 146–​7, 150, 161 Homo economicus 14–​15, 186–​7 Homo moralis 186–​8 honesty 187 hypothetical imperative 167, 189 identification problems 164 identity: pro-​environmental 173, 175–​7, 179–​80; selfish 172, 175–​6, 179 ignorance 38, 50, 92–​3 image 111, 150 impartial spectator 16–​17 incentive: compatibility 55, 65; economic 20–​1, 65, 71, 108–​12; social 22, 108, 110, 112, 138, 171 increasing returns 119–​21 indirect reciprocity 19, 152, 170 inequality aversion 113–​14, 190 inertia 14, 84, 135, 153, 155 information: asymmetric 2, 63–​4; imperfect 2, 62–​6, 72 internal moral constraints 165–​9, 181–​2 interaction 63, 87, 110, 117, 122–​3, 129–​30, 134–​5, 138–​52, 160–​4, 166, 172–​3, 179, 185, 187 inter-​personal 175 intra-​personal 175 introspection 30, 36 irrationality 78 judgment 16–​17, 32, 42, 44, 48, 70, 88–​9, 166–​7 justice 22, 44–​5, 75, 109, 114, 168, 189 Kantian ethics 22, 167, 188–​9 knowledge: common 146, 149; division of 62, 140

204 Index

leadership 89 learning 78, 82, 92–​3, 99, 117, 138, 162–​3 libertarian paternalism 14, 74, 99, 101 lock-​in 121 loss 95–​8 loss aversion 41, 73, 96 manager 126–​35 man within the breast 16 marginal revolution 35 market: disequilibrium 91–​2; failure 9, 70–​1, 73, 98; system 1, 21, 126 mechanism 3–​4, 14, 16, 19, 22, 52, 59, 65, 71, 85–​7, 101, 111, 115, 118, 120, 123–​4, 129, 141, 145, 147–​8, 151, 160, 162, 173–​4, 190 mental states 27, 30–​1, 36–​7, 45 menu-​dependence 31 measurements 36 messages 7, 65, 74, 102, 151 mistakes 9, 34, 70, 77, 91, 98 monitoring 173, 180 moral: hazard 64–​5; obligations 22; rule 22, 167, 169, 181, 189–​90; sentiments 15, 17, 67, 166, 173, 187, 189 motivation: extrinsic 20, 65–​6, 108–​9; intrinsic 21, 66, 69, 108–​11, 188–​9 multiple equilibria 120, 142, 147 natural monopoly 2, 8, 64, 98 nature 148 non-​linear tax schedule 55 norm: compliance 151–​4, 159–​60; descriptive 142; injunctive 142; social 69, 75, 120–​1, 141–​2, 148–​9 nudge 13, 74, 99–​101 obesity 57–​9, 75, 105, 108 opportunities 88, 90–​3, 153 optimal taxation theory 53, 70 opt-​in 13 opt-​out 13–​14 order 139 organization 82–​5 other interest 15, 20 Pareto efficiency (also Pareto optimality) 1 path dependence 121 patience 32 pattern 69–​71, 130, 139, 162, 187 payoff 15, 32, 112–​14, 122–​3, 143, 150, 175, 177 penalty 67–​8, 109, 111, 154

perception 79–​80, 88, 91, 93–​4, 128–​31, 138, 141, 145, 150 philanthropy 125, 129 pledge 106 policymaker 50, 100 pollution 2–​3, 5, 46, 53, 71 polycentricity 171 positivism 25 precedent 146–​7 preferences: changes 32, 104; extended 38; moral 168–​9, 173, 175, 181–​3, 185–​6, 190; self-​regarding 18, 116; other-​regarding 10, 15, 116, 187–​9; revealed 25–​6, 28–​9, 30–​3, 37; social 14, 20, 22, 56, 65, 108, 112–​13, 129–​31, 135, 165–​6, 173, 180, 187–​9; true 8, 31, 33–​4, 70, 77, 98 prediction 11, 15, 71, 94, 123, 141, 145 preferences 25–​35 principal-​agent 64, 134 prisoner’s dilemma 17–​18 procrastination 10 property rights 173 prospect theory 41, 73, 93, 97–​8 public goods 2–​3, 65, 112, 122–​5, 129, 135–​7, 150–​1 punishment 123–​4, 151–​4, 156–​60, 187 quasi-​hyperbolic 61, 105 rationality: bounded 7, 10, 71, 77–​9, 81, 85–​7; constructivist 87, 107; ecological 85–​7, 107; global 78–​9; procedural 79; selective 84; substantive 79 reasoning 10, 27, 87, 146–​7, 167 reciprocity 5, 17–​19, 115, 141, 150, 170, 187 redistribution 56 reference group 112, 115 resource sustainability 179–​80, 185–​6 reference point 41, 73, 97 reflection effect 41 regularity 138, 142, 145–​7 regulation 1, 74 reinforcement 60 reputation 19, 22–​3, 111, 160 revelation 65 reward 20–​2, 32, 54, 61–​2, 65, 68, 92, 102–​3, 108–​11, 150, 154, 188 risk 88 routine 82–​3 rule following 139 rules of thumb 7, 93

Index  205

salience 14, 145, 147 Samuelsonian condition 3 sanction 19–​21, 66, 111, 123–​4, 142, 151, 167, 173 satisficing 80–​2 scarcity 25 screening 63–​4 second best 6 search 79–​83, 92–​3 self-​command 17 self-​control 102–​6 self-​enforcement 173 self-​governance 173 self-​image 22, 74, 150, 161 self-​interest 10, 14–​15, 18–​20, 109, 112, 115–​16, 129, 137, 141, 161, 180–​1, 186–​7 self-​organization 161 self-​respect 187–​8 shareholder 126–​7, 134 shame 22, 69, 74, 188 signaling 19, 63, 148 social: category 161, 174; harm 168–​9; surplus 169 socialization 116–​2, 129, 142 stakeholder 126–​8 status-​quo 7 St. Petersburg game 38 strategy 18, 33, 47, 79–​81, 116, 130, 134, 143, 145–​9, 168, 180 strong axiom of revealed preference 26 subjectivism 89, 140 subsidy 5, 64 substitutes 20, 47, 65, 99 sugar-​sweetened beverages 58 surprise 92–​3 sustainability 173, 179–​81, 185–​6 sympathy 15–​17, 20, 166, 187

talent 90, 107 task 1, 71, 79, 85–​6, 109 tax: commodity 56; compliance 66–​70; design 52, 70; evasion 62, 66–​70; emission 4, 70; income 52–​6; lump-​sum 43, 45, 55, 65; morale 69; sin 59–​61 team 96 temptation 18, 32, 59, 62, 103 threat 3, 152, 160 tit-​for-​tat 18 time inconsistency 10, 61, 105 top-​down 173 tradition 11, 36, 140, 142 transaction costs 5, 70, 101 transformation function 43 transitivity 32, 39 trust 128 ultimatum game 15, 163, 166, 187 uncertainty 88–​9, 139 utilitarianism 36, 42, 50 utility: cardinal 25, 35–​7, 41, 44, 47; comparison 25, 36–​7, 41, 44, 54, 115; ordinal 1, 25, 35, 37, 41 value function 97–​8 veil of ignorance 38 violation 32, 95, 151, 153–​4, 160, 187 virtue 17, 133 von Neumann-​Morgenstern utility function 68, 169 wage 54, 63–​4, 115 warm glow 20, 125, 129, 188 weak axiom of revealed preference 26 welfare 41–​6 well-​being 35, 37, 45–​51 willpower 10–​11 X-​efficiency (also X-​inefficiency) 84