ebook img

Neuroeconomics. Decision Making and the Brain PDF

575 Pages·2014·57.117 MB·English
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview Neuroeconomics. Decision Making and the Brain

NEUROECONOMICS SECOND EDITION NEUROECONOMICS Decision Making and the Brain SECOND EDITION Edited by P W. G AUL LIMCHER NewYorkUniversity,NewYork,NY,USA E F RNST EHR UniversityofZurich,Switzerland AMSTERDAM(cid:129)BOSTON(cid:129)HEIDELBERG(cid:129)LONDON NEWYORK(cid:129)OXFORD(cid:129)PARIS(cid:129)SANDIEGO SANFRANCISCO(cid:129)SINGAPORE(cid:129)SYDNEY(cid:129)TOKYO AcademicPressisanimprintofElsevier Academic Press isan imprint of Elsevier 32 JamestownRoad, LondonNW1 7BY, UK 225 Wyman Street, Waltham, MA 02451, USA 525 B Street, Suite1800,San Diego, CA 92101-4495, USA Copyright r 2014, 2008 Elsevier Inc. All rights reserved No part of this publication may be reproduced, stored in a retrieval system ortransmitted in any form or by any means electronic, mechanical, photocopying, recording or otherwisewithout the prior writtenpermission ofthe publisher. Permissions may besought directly from Elsevier’s Science&Technology RightsDepartment in Oxford, UK:phone (144)(0) 1865843830;fax (144)(0) 1865 853333; email: permissions@elsevier. com. Alternatively, visit theScienceand Technology Books website atwww.elsevierdirect.com/ rights for furtherinformation. Notice No responsibility isassumed by the publisherfor anyinjury and/ordamage to persons or property as amatter of productsliability, negligence or otherwise, orfrom anyuse or operation ofany methods, products, instructions or ideas contained in thematerial herein. Because ofrapid advancesin themedical sciences, in particular,independent verificationof diagnosesanddrug dosages should bemade. British Library Cataloguing-in-Publication Data Acataloguerecordfor thisbook is availablefrom the BritishLibrary Library ofCongress Cataloging-in-Publication Data Acatalog record for this book isavailablefrom theLibraryof Congress ISBN : 978-0-12-416008-8 For information on allAcademic Press publications visit our website at elsevierdirect.com Typesetby MPS Limited, Chennai, India www.adi-mps.com Printed and bound in United States ofAmerica 14 15 16 17 10 9 8 7 6 5 4 3 2 1 Preface USING THE BOOK Since the publication ofthe first edition of this book, there has been an increasing institutional recognition of the importance of Neuroeconomics in the future of Oneofthecriticalchallengesfacinganyoneinterested neuroscience, economics, and psychology. At the time in Neuroeconomics isthe interdisciplinary nature of the that the first edition was published, just a handful of science. To be a neuroeconomist one must be fluent in academic institutions included scientists engaged in the languages of Economics, Psychology, Neuroscience, neuroeconomic research, but the field has matured and (to a lesser degree) Primate Anthropology. We at an astonishing rate over the past 5 years. Today recognize that very few scholars or students will come scholars at nearly a hundred institutions worldwide tothisbookwithabackgroundinalloftheseareas,and are at work on neuroeconomic problems and courses the first section of the book is designed to address that on neuroeconomics are now commonplace at both the constraint before the later sections of the book are graduate and undergraduatelevels. encountered. At the time that the first edition of this volume was published, it was also true that very little was known about the biological mechanism of human and animal Section 1 decision making. Accordingly, the first edition was primarily designed to provide scholars interested in Accordingly, the first section of this book is broken beginning to undertake neuroeconomic research with a into4components.Chapters1and2describethetheory strong interdisciplinary background in the area. Today, and methods of experimental economics. For someone however, the landscape is quite different. Over the past trained in economics these chapters would be entirely 5 years the field of neuroeconomics has matured intel- superfluous. They are intended for neurobiologists and lectually as well as institutionally. Thanks to the work psychologists who are trying to get a handle on eco- of hundreds of cutting-edge scholars, we now know nomic thought as a preparation for the second through quite a lot about how and where decisions are made in fifth sections of the book. For someone who knows a the brain, and that is reflected in the structure of this little economics but who has not been formally trained secondedition. in economics, these two chapters are essential. Chapters As we did in the first edition, we continue to believe 3 and 4 (along with the appendix) provide a survey that a strong interdisciplinary background is important of the psychology of Judgment and Decision Making. forscholarsinthisarea,butwenowconfinethatreview These chapters should be of particular value to econo- to the first section of the book. The second section pre- mists and neuroscientists unfamiliar with that tradition. sents what is known to date about the neural structure Chapters 5 and 6 are designed to provide basic literacy of preferences ranging from risk attitudes to social pre- in the fundamental methodologies of neuroscience for ferences, and to inter-temporal choice. The third section non-neuroscientists. Social scientists who hope to make focusesonthelearningofvaluesandtheneuralsystems sense of the empirical chapters that follow are urged centraltoourunderstandingoftheneuralrepresentation to take particular care in reading Chapter 5 which pro- of subjective value. Section 4 examines what is known vides a primer on basic neuroscience and Chapter 6 about the choice process itself; the mechanism inter- which describes the many methods of neuroscientific posed between the valuation processes described in research. Even for those non-neuroscientists familiar Section 3 and behavior. Section 5 expands on social with basic neuroscience, Chapter 6 should provide studies of decision making, an important frontier in a valuable source for understanding the limitations of neuroeconomic research today. The book concludes methods employed in neuroscientific research that can with an appendix describing the prospect theory of be consulted as one reads the rest of the book. Finally, Kahneman and Tversky in detail. It provides important Chapter 7 provides a useful introduction to the study practical information on the use of prospect theory in of non-human animals in decision making. For those neuroeconomicexperiments. whose work, or studies, have focused exclusively on ix x PREFACE human decision makers, this chapter should provide a dopamine. This is an area where tremendous progress clear motivation for understanding the studies of has been made in computational neuroscience and non-human decision makers provided throughout this the chapter provides a clear summary of the mathe- volume. matical and empirical bases for understanding how For instructors using this volume as a textbook in “values” are learned by the mammalian brain. This a graduate or undergraduate class, the first section will be new material for many economist readers, but should typically not be presented in its entirety. If the even for neurobiologists familiar with studies of the course is, for example, being presented in a depart- neurotransmitter dopamine, this chapter should pro- ment of psychology as an advanced elective, Chapters vide an important computational foundation. The sec- 1 and 2 may be all that is required for most students. ond chapter of the section, Chapter 16, extends these If the course is for students from many backgrounds, ideas with a review of advanced topics in reinforce- Section 1 may be an appropriate object for self-study ment learning. Together Chapters 15 and 16 should early in the class sequence. provide an essential starting point for anyone inter- ested in the neural basis of valuation. Chapter 17 builds on these ideas, and requires some familiarity Section 2 with the material in Chapter 15. It discusses advanced topics in value encoding in other brain areas and pre- The second section of the book presents core concepts sents data on current research frontiers in reinforce- that guide much neuroeconomic research. The section ment learning. The section concludes with Chapter 18 begins with a first chapter that deals with the basic which presents alternative views of many of the ideas neural foundations of subjective value in simple binary presented in the first three chapters of this section (cid:1) choicesituationsandisfollowedbychaptersthatengage including important challenges to the core theories centralnotionsin neuroeconomicresearch.Severalof the presented in the preceding chapters. This chapter will chapters of this section focus on the notion of preferences; beof particularinterestto behavioral economists. theydescribeacoreideainthestudyofdecisionmaking at both a behavioral and a neuroscientific level of analy- sis. Thus Chapter 9 in this section presents a detailed Section 4 account of the notion of risk-preference. It describes eco- nomic and psychological models of risk preferences that The fourth section presents an overview of the have guided neuroscientific research and then provides choice process: the neural mechanism that takes value a detailed review of current neuroeconomic research in or sensory evidence as an input and triggers action this subarea. Other chapters in this section develop this as a behavioral output. It begins with two chapters same theme for intertemporal preferences, social prefer- that examine the two main neuroscientific threads in ences, and for the impact of emotion on preferences. the study of choice: perceptual decision making and Chapter 13 describes a wealth of research suggesting value-based decision making. These are followed by a that neural activity encodes the value of goods and series of chapters that examine advanced topics in action in a single common neural currency (cid:1) an idea this area. Chapter 21 describes evidence that multiple closelyrelatedtotheeconomicnotionofutility.Overthe neural systems interact during the choice process. last decade, this idea has emerged as a central theme in Chapter 22 examines cutting-edge research on the neuroeconomic research and this chapter reviews those multiple brain systems that integrate costs and bene- important findings. The section closes with a review of fits in the generation of choice. Chapter 23 describes what might be called the chemistry of choice (Chapter 14). neuron-level modeling that is beginning to reconcile It examines an emerging thrust of neuroeconomics: the perceptual, value-based, and cost-related decision study of how pharmacologic agents like the hormone making in a single framework. The final chapter in oxytocin influence choice behavior. To achieve that goal this section examines well-known violations of tradi- it provides a basic review of neuropharmacology that tional choice theory and explores the emerging notion will be of particular interest to non-neuroscientific that neurobiological constraints may underlie many readers. of these phenomena. Section 3 Section 5 The third section of the book focuses on how we This section concludes the volume with an overview learn and represent value. The first chapter in this of social decision making, within the framework of section (Chapter 15) provides a much-needed overview game theory. Building explicitly on the material pre- of the neurobiology of reinforcement learning and sented in Chapter 11 and in Chapter 2, this section of xi PREFACE the book explores the neurobiology of social decision prospect theory has played a key role in the develop- making in some detail. It begins with an overview of ment of neuroeconomics, but the theory is more behavioral game theory in Chapter 25 at the behavioral complicated than is generally realized. This chapter and neural levels. It then proceeds to studies of the thus serves two goals. First, it should provide a neurobiology of game theoretic behaviors in non- detailed and highly valuable how-to guide explain- human primates, an area of very active research at this ing how prospect theory should, and should not, time. It concludes with a review of neurobiological be used in the laboratory. Second, it provides an studies of empathy and the theory of mind, which overview of current neuroeconomic research on the guidesneuroeconomicstudiesofsocialbehavior. foundations of prospect theory. Slightly longer than most of the other chapters in the book, it should Appendix provide an invaluable hands-on reference for anyone planning prospect theoretic research. The volume concludes with a very detailed appen- PaulGlimcher and Ernst Fehr dix on prospect theory. Kahneman and Tversky’s Acknowledgments We would like to thank the many people whose hard Finally, we wish to express our thanks to those work made this volume possible. First and foremost who made this book possible. To Johannes Menzel we want to acknowledge the hard work of the many of Elsevier who was the editor of the first edition, to hundreds of independent researchers whose work is April Graham who served as the book’s associate edi- represented in this volume. The authors of the book, tor at Academic Press and to Mica Haley at Academic anecessarilysmallsubsetofthosescientists,haveworked Press who was our editor. Finally, and most impor- hard to encapsulate the staggering accomplishments of tantly, we want to express our truly undying grati- theirpeers.Theyhavedonewhatwetaketobeanamaz- tude to one of most important people in the field ing job, although we acknowledge that even this huge of Neuroeconomics, Samanta Shaw. Samanta is one of volumeisincompletebecausethefieldofneuroeconomics the great heroes of the birth of neuroeconomics, grows every day. We ask the forgiveness of the many although she is little known outside the core of the scholarswhoseworkhasreceivedinadequatecoverage. field. For the last 5 years she has served as the admin- We would also like to express particular thanks to istrative director of the Society for Neuroeconomics the authors of the first edition of this volume. The first and in that capacity she has probably done more to edition was much more a survey of neuroeconomics further the field than anyone else. As the de facto editor and much less a textbook than is this volume. In order of this volume she has prepared the second edition (as to make it more of a text, we were forced to signifi- she did the first edition) through submission, revision, cantly reduce the number of authors and to sharpen revision again, production, and marketing. We, and all the focus of the book. The authors of the first edition of neuroeconomics, owe her an immeasurable debt. were gracious in allowing us to reuse material from Thanks, Sam. that edition where it seemed appropriate and we here express our gratitude to them for their contribution. PaulW.Glimcher and Ernst Fehr xiii List of Contributors Adele Diederich Jacobs University, Bremen gGmbH, Joshua I. Gold University of Pennsylvania School of Bremen,Germany Medicine,Philadelphia,PA,USA AndrewCaplin NewYorkUniversity,NewYork,USA Karolina M. Lempert New York University, New York, Anita Tusche Max Planck Institute for Human Cognitive NY,USA andBrainSciences,Leipzig,Germany Kenji Doya Okinawa Institute of Science and Technology, Antonio Rangel CaliforniaInstituteofTechnology,Pasadena, Okinawa,Japan CA, USA; California Institute of Technology, Pasadena, CA, Kent C. Berridge University of Michigan, Ann Arbor MI, USA USA BenedettoDeMartino University College London, London, KenwayLouie NewYorkUniversity,NewYork,NY,USA UK; California Institute of Technology Pasadena, USA KevinMcCabe MasonUniversity,Arlington,VA,USA ChristianC.Ruff UniversityofZu¨rich,Zu¨rich,Switzerland LaurieR.Santos YaleUniversity,NewHaven,CT,USA Colin F. Camerer California Institute of Technology, Matthew F. S. Rushworth University of Oxford, Oxford, Pasadena,CA,USA UK Craig R. Fox UCLA Anderson School of Management, Los Michael C. Dorris Queen’s University, Kingston ON, Angeles,CA,USA Canada; Chinese Academy of Sciences, Shanghai, China DaeyeolLee YaleUniversityNewHaven,CT,USA Michael L. Platt Duke University, Durham, NC, USA; DanielHouser GeorgeMasonUniversity,Fairfax,VA,USA DukeUniversity,Durham,NC,USA Elizabeth A. Phelps New York University, New York, NY, Minoru Kimura Tamagawa University, Machida Tokyo, USA Japan ElkeU.Weber ColumbiaUniversity,NewYork,NY,USA Molly J. Crockett University College London, London, UK; EricJ.Johnson ColumbiaUniversity,NewYork,NY,USA UniversityofZu¨rich,Zu¨rich,Switzerland ErnstFehr UniversityofZu¨rich,Zu¨rich,Switzerland NathanielD.Daw NewYorkUniversity,NewYork,USA HaukeR.Heekeren FreieUniversita¨tBerlin,Berlin,Germany PaulW.Glimcher NewYorkUniversity,NewYork,USA Hilke Plassmann INSEAD, Fontainebleau, France; Ecole Phillipe N. Tobler University of Zu¨rich, Zu¨rich, NormaleSupe´rieure,INSERM,Paris,France Switzerland IanKrajbich UniversityofZu¨rich,Zu¨rich,Switzerland RogerRatcliff OhioStateUniversity,OH,USA JeromeR.Busemeyer IndianaUniversity,IN,USA Russell A. Poldrack University of Texas at Austin, Austin, TX,USA John A. Clithero California Institute of Technology, Pasadena,CA,USA ScottA.Huettel DukeUniversity,Durham,NC,USA John P. O’Doherty California Institute of Technology, Tania Singer Max Planck Institute for Human Cognitive andBrainSciences,Leipzig,Germany Pasadena,CA,USA Jonathan D. Wallis University of California at Berkeley, ToddA.Hare UniversityofZu¨rich,Zu¨rich,Switzerland Berkeley,CA94720,USA Xiao-Jing Wang New York University, New York, NY, Joseph W. Kable University of Pennsylvania, Philadelphia, USA;NYUShanghai,Shanghai,China PA,USA YaelNiv PrincetonUniversity,Princeton,NJ,USA xv Introduction: A Brief History of Neuroeconomics Paul W. Glimcher and Ernst Fehr With significant material from the first edition by Colin F.Camererand RussellA. Poldrack Neuroeconomics has its origins in two places, in Beginninginthe1930sagroupofeconomists(cid:1)most events following the neoclassical economic revolution famously, Samuelson, Arrow, and Debreu (cid:1) began to of the 1930s and in the birth of cognitive neuroscience investigate the mathematical structure of consumer during the 1990s. So we begin this brief history with a choice and behavior in markets. Rather than simply review of the neoclassical revolution and the birth of building models that incorporated a set of parameters cognitive neuroscience. thatmight,onaprioripsychologicalgrounds,bepredic- tive of choice behavior, this group of theorists began to investigate what mathematical structure of choices might result from simple, more “primitive,” assump- NEOCLASSICAL ECONOMICS tions on preferences. Many of these models (and the styleofmodelingthatfollowed)hadastrongnormative1 The birth of economics is often traced to Adam flavor,inthesensethatattentionwasmostimmediately Smith’s publication of The Wealth of Nations in 1776. focused on idealized choices and efficient allocation With this publication begins the classical period of of resources (as opposed to necessarily seeking to economic theory. Smith described a number of phe- describe how people choose, as psychologists do, and nomena critical to understanding choice behavior and howmarketswork). the aggregation of choices into market activity. These To better understand this approach, consider what were, in essence, psychological insights. They were is probably the first and most important aspect of these relatively ad hoc rules that explained how features of simple models. Instead of assuming that an unknown the environment influenced the behavior of a nation internal property that we may call “preference” or of consumers and producers. “utility” causeschoices,this approachtakes only choices What followed the classical period was an interval as primitives and subsequently asks whether choices during which economic theory became very heteroge- can be represented by some form of preference or utility nous. A number of competing schools with different function. In other words, preferences and utility func- approaches developed. Many economists of the time tions have no independent existence from choices. They (Edgeworth, Ramsey, Fisher) dreamed about tools to are just a way of representing or redescribing the object infer value from physical signals, through a “hedoni- of interest (cid:1) choices. To see what this means, assume meter” for example, but these early neuroeconomists thatanindividualwhocanchoosebetweenanappleand did not have such tools (Colander, 2007). an orange, chooses the orange. Then we can represent Oneschoolofthought,duetoJohnMaynardKeynes, this choice by assigning a higher number to the orange, was based on the view that regularities in consumer say 10 for the orange and 9 for the apple and by adding behavior could (among otherthings) provide a basis for the assumption that higher numbers represent preferred fiscalpolicytomanageeconomicfluctuations.Manyele- objects. We can also call this a utility function, meaning mentsinKeynes’theory suchasthe“propensitytocon- that U(orange)510 and U(apple)59. Clearly this utility sume” or entrepreneurs’ “animal spirits” that influence function represents the individual’s choice but so does their investment decisions were based on psychological any other function that assigns a higher number to the concepts.ThisframeworkdominatedUnitedStatesfiscal orange. It is thus transparent that utility functions only policyuntilthe1960s. represent choice and have no independent explanatory 1Inthelanguageofeconomicsonetypicallydistinguisheswhatarecalledpositiveandnormativetheories.Positivetheoriesarethose whichseekonlytopredictfuturebehavior,eitherbyindividualsorbyeconomies.Normativetheoriesarethosethatseektomake statementsaboutwhatchoicesare“best,”orinthelanguageofeconomicswelfaremaximizing.Positivetheoriesarethusdescriptive, andnormativetheoriesprescriptive. xvii xviii INTRODUCTION:ABRIEFHISTORYOFNEUROECONOMICS power. They are merely a convenient tool for capturing theessenceofthings....Purepoliticaleconomyhastherefore adecision-maker’schoices. agreatinterestinrelyingaslittleaspossibleonthedomainof psychology. The most important consequence of this approach is Busino(1964;p.xxiv) that it made economics independent from psychology (and neuroscience) because all economists need to The revealed preference revolution in economics study is choices while questions about the neural and achieved Pareto’s goal on the basis of clear conceptual psychological processes behind choices are irrelevant foundations and it took several decades until econom- as long as human decision making obeys some simple ics returned to psychology as an important source of consistency axioms such as the Weak Axiom of Revealed insights. Preference (or WARP) or the Generalized Axiom of What followed the development of WARP and Revealed Preferences (GARP). Both of these axioms are GARP were a series of additional theorems of this type discussed in greater detail in Chapter 1 of this book. which extended the scope of revealed-preference theory GARP essentially means that an individual’s choices to choices with uncertain outcomes whose probabilities are transitive while WARP means that if an individual are known (von Neumann and Morgenstern’s expected chooses an orange over an apple when both are utility theory, EU (1944); Chapter 1) or subjective (or available in situation 1, then the individual should not “personal,” in Savage’s subjective EU theory; SEU; choose the apple over the orange in another situation Chapter 9), and in which outcomes may be spread over where again both are available. Neoclassical econo- time (discounted utility theory; Chapter 10). What is mists such as Samuelson and Houthakker have shown mostinterestingaboutthesetheoriesisthattheydemon- that if individuals obey these consistency axioms pow- strate, amongst other things, that a chooser who obeys erful implications follow. In particular, their behavior these axioms must behave both “as if” he has a utility can be represented by the maximization of some utility function that constitutes a functional relation between function, implying that they behave as if they maxi- outcomes and their utilities and “as if” his actions were mize some utility function. In addition, the consistency aimed atmaximizingtotal utility. In their seminal book, axioms are directly testable at the behavioral level; the von Neumann and Morgenstern also laid the founda- existence of utility functions can be subjected to rigor- tionsformuchofgametheorywhichtheysawasaspecial ous empirical tests. For example, if individuals violate problem in utility theory, in which outcomes are gener- WARP then their behavior cannot be rationalized by atedbytheinteractingchoicesofmanyplayers. anyutility function. At the end of this period, neoclassical economics The revealed preference approach, which forms seemed incredibly powerful. Starting with a few sim- much of the subject matter of Chapters 1 and 2, thus ple assumptions which fully described a new theory starts from a set of assumptions called axioms which (for example, expected utility theory), the neoclassi- encapsulateatheoryofsomekind(oftenaverylimited cists developed a framework for thinking about and one) in formal language. The poetry in the approach predicting choice. These theories of consumer choice (what distinguishes a beautiful theory from an ugly would later form the basis for the demand part of one) is embodied in the simplicity of the axioms, and the Arrow(cid:1)Debreu theory of competitive “general” the degree to which surprisingly simple axioms make equilibrium, a system in which prices and quantities sharp predictions about what kind of choice patterns of all goods were determined simultaneously by should and should not be observed. Finally, it is matching supply and demand. This is an important critical to note that what the theory predicts is what tool because it enables the modeler to anticipate all new choices could possibly follow from an observed consequences of a policy change: for example, impos- set of previous choices (including choices that respond ing a luxury tax on yachts might increase crime in a to policy and other changes in the environment such shipbuilding town because of a rise in unemployment as responsesto changes inprices, taxes, or incomes). there. This sort of analysis is unique to economics and It cannot be emphasized enough how much the partly explains the broad influence of economics in revealed-preference view suppressed interest in the psy- regulation and policymaking. chologicalnatureofpreference,becausecleveraxiomatic Whilethe“asif”approachhasathoroughconceptual systemscouldbeusedtoinferpropertiesofunobservable and mathematical foundation, and testable predictions, preference from observable choice (Bruni and Sugden, in the realm of revealed preference theory, Milton 2007). Before the neoclassical revolution, Pareto noted in Friedman extended the “as if” argument to other 1897that: domains in a rather questionable way, and without a rigorous foundation. In the 1950s, Friedman wrote an It is an empirical fact that the natural sciences influential book, The Methodology of Positive Economics. haveprogressedonlywhentheyhavetakensecondaryprinci- Friedman argued that assumptions underlying a ples as their point of departure, instead of trying to discover

See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.