See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/329092314

Experimental Economics: Method and Applications

Book · November 2018

DOI: 10.1017/9781107446786

CITATIONS READS

19 8,191

2 authors, including:

Nicolas Jacquemet
Ecole d'économie de Paris, French National Centre for Scientific Research
149 PUBLICATIONS 1,437 CITATIONS

SEE PROFILE

All content following this page was uploaded by Nicolas Jacquemet on 21 November 2018.

The user has requested enhancement of the downloaded file.

Cambridge University Press
978-1-107-06027-2 — Experimental Economics
Nicolas Jacquemet , Olivier L'Haridon
Frontmatter
More Information

Experimental Economics
Method and Applications

Over the past two decades, experimental economics has moved from a fringe activity
to become a standard tool for empirical research. With experimental economics now
regarded as part of the basic tool-kit for applied economics, this book demonstrates
how controlled experiments can be useful in providing evidence relevant to economic
research. Professors Jacquemet and L’Haridon take the standard model in applied econo-
metrics as a basis for the methodology of controlled experiments. Methodological
discussions are illustrated with standard experimental results. This book provides future
experimental practitioners with the means to construct experiments that fit their research
question, and newcomers with an understanding of the strengths and weaknesses of con-
trolled experiments. Graduate students and academic researchers working in the field of
experimental economics will be able to learn how to undertake, understand and criticise
empirical research based on lab experiments, and refer to specific experiments, results
or designs completed with case study applications.

Nicolas Jacquemet is a full professor at University Paris-1 Panthèon Sorbonne and the
Paris School of Economics. His research combines experimental methods and econo-
metrics to study discrimination, the effect of personality traits on economic behaviour,
the role of social pre-involvement in strategic behaviour and experimental game theory.
His research has been published in Econometrica, Management Science, Games and
Economic Behavior, Journal of Environmental Economics and Management, Journal of
Health Economics and Journal of Economic Psychology.
Olivier L’Haridon is a full professor at the University of Rennes 1. His research com-
bines experimental methods and decision theory, applied in the study of individual
decision-making as affected by uncertainty. His work has been published in Ameri-
can Economic Review, Management Science, Journal of Risk and Uncertainty, Theory
and Decision, Experimental Economics, Journal of Health Economics and Journal of
Economic Psychology.

© in this web service Cambridge University Press www.cambridge.org

Cambridge University Press
978-1-107-06027-2 — Experimental Economics
Nicolas Jacquemet , Olivier L'Haridon
Frontmatter
More Information

© in this web service Cambridge University Press www.cambridge.org

Cambridge University Press
978-1-107-06027-2 — Experimental Economics
Nicolas Jacquemet , Olivier L'Haridon
Frontmatter
More Information

Experimental Economics
Method and Applications

NICOLAS JACQUEMET
University Paris 1 Panthèon-Sorbonne and Paris School of Economics, France

O L I V I E R L’ H A R I D O N
Université de Rennes I, France

© in this web service Cambridge University Press www.cambridge.org

Cambridge University Press
978-1-107-06027-2 — Experimental Economics
Nicolas Jacquemet , Olivier L'Haridon
Frontmatter
More Information

University Printing House, Cambridge CB2 8BS, United Kingdom

One Liberty Plaza, 20th Floor, New York, NY 10006, USA
477 Williamstown Road, Port Melbourne, VIC 3207, Australia
314–321, 3rd Floor, Plot 3, Splendor Forum, Jasola District Centre, New Delhi – 110025, India
79 Anson Road, #06–04/06, Singapore 079906

Cambridge University Press is part of the University of Cambridge.

It furthers the University’s mission by disseminating knowledge in the pursuit of
education, learning, and research at the highest international levels of excellence.

www.cambridge.org
Information on this title: www.cambridge.org/9781107060272
DOI: 10.1017/9781107446786
© Cambridge University Press 2018
This publication is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without the written
permission of Cambridge University Press.
First published 2018
Printed in the United Kingdom by TJ International Ltd. Padstow Cornwall
A catalogue record for this publication is available from the British Library.
Library of Congress Cataloging-in-Publication Data
Names: Jacquemet, Nicolas, author. | L’Haridon, Olivier, author.
Title: Experimental economics method and applications / Nicolas Jacquemet,
Paris School of Economics, Olivier L’Haridon, Université de Rennes I, France.
Description: Cambridge, United Kingdom ; New York, NY, USA : Cambridge
University Press, [2018] | Includes bibliographical references and index.
Identifiers: LCCN 2018007008 | ISBN 9781107060272
Subjects: LCSH: Experimental economics.
Classification: LCC HB131 .J33 2018 | DDC 330.072/4–dc23
LC record available at https://lccn.loc.gov/2018007008
ISBN 978-1-107-06027-2 Hardback
ISBN 978-1-107-62977-6 Paperback
Cambridge University Press has no responsibility for the persistence or accuracy of
URLs for external or third-party internet websites referred to in this publication
and does not guarantee that any content on such websites is, or will remain,
accurate or appropriate.

© in this web service Cambridge University Press www.cambridge.org

Cambridge University Press
978-1-107-06027-2 — Experimental Economics
Nicolas Jacquemet , Olivier L'Haridon
Frontmatter
More Information

Contents

List of Figures page viii

List of Tables x
List of Illustrations xii
List of Focuses xiv
Abbreviations and Symbols xvi
Preface xxi

Part I What Is It? An Introduction to Experimental Economics 1

1 The Emergence of Experiments in Economics 3

1.1 The End of a Long-Standing Regretful Impossibility 4
1.2 Why Such a Change: Two Early Examples 6
1.3 The Research Programme: Three Examples 12
1.4 Experimental Economics Today: What Every Newcomer Must Know 22

2 A Laboratory Experiment: Overview 26

2.1 The Experiment 27
2.2 The Experimenter’s Role: The Game under Study 34
2.3 Experimental Second-Price Auction with Private Values 38
2.4 Case Study: Experimentally Designed Devices to Reduce Hypothetical
Bias 41

Part II Why? The Need for Experiments in Economics 49

3 The Need for Controlled Experiments in Empirical Economics 51

3.1 The Econometric Approach to Data Analysis 52
3.2 Estimating Causal Effects of Treatments 59
3.3 Identification Based on Observational Data 68
3.4 Inference Based on Controlled Experiments 72
3.5 From the Laboratory to the Field: An Overview of Controlled
Experiments in Economics 83

4 The Need for Experimental Methods in Economic Science 88

4.1 What Laboratory Experiments Aim For 88

© in this web service Cambridge University Press www.cambridge.org

Cambridge University Press
978-1-107-06027-2 — Experimental Economics
Nicolas Jacquemet , Olivier L'Haridon
Frontmatter
More Information

vi Contents

4.2 Experiments, Theory and Reality: How Experiments Achieve Their

Goals 90
4.3 Case Study. Deepening Understanding through Additional Controls and
Measures: The Dictator Game 95
4.4 How Experiments Interact with Theory: Testing Models 103
4.5 How Experiments Interact with Reality: Searching for Facts 110

Part III How? Laboratory Experiments in Practice 117

5 Designing an Experiment: Internal-Validity Issues 119

5.1 What Is an Experiment? How Is It Linked to Internal Validity? 119
5.2 The Incentive Structure of Experiments 132
5.3 Parameters and Experimental Treatments 147
5.4 The Perceived Experiment 159
5.5 Perceived Opponents and Learning 166
5.6 Case Study: Eliciting Beliefs 170

6 Conducting an Experiment 191

6.1 A Long, Long Time Beforehand: Setting Up an Experimental Laboratory 191
6.2 Two Months Before: The Basics 195
6.3 One Month Before: The Final Settings 204
6.4 One Week Before: Almost There 206
6.5 D-Day: Step-by-Step Proceedings 207
6.6 Case Study: Measuring Preferences in Choice over Time 208

7 The Econometrics of Experimental Data 229

7.1 Experimental Data 230
7.2 Estimation and Inference 243
7.3 Testing Procedures 256
7.4 Case Study: Eliciting Preferences under Risk 289

Part IV What For? What Laboratory Experiments Tell Us 321

8 The External Validity of Experimental Results 323

8.1 When and How Does External Validity Matter? 324
8.2 Is External Validity Testable? 336
8.3 Testing External Validity 339
8.4 Case Study: Replication: Enhanced Credibility Thanks to Accumulated
Evidence 352

9 More Accurate Theory and Better Public Policies: the Contributions of

Experimental Economics 361
9.1 Testing Theory: Drawing General Lessons from (Causal) Experimental
Evidence 362

© in this web service Cambridge University Press www.cambridge.org

Cambridge University Press
978-1-107-06027-2 — Experimental Economics
Nicolas Jacquemet , Olivier L'Haridon
Frontmatter
More Information

Contents vii

9.2 Case study: Rational Behaviour, Irrational Thinking: K-level Models 369
9.3 Test-Bedding Public Policies in the Laboratory: The Example of
Matching Markets 380
9.4 Whispering in the Ear of Princes: Behavioural Public Policy 385

References 398
Index 431
Index of Authors 441

© in this web service Cambridge University Press www.cambridge.org

Cambridge University Press
978-1-107-06027-2 — Experimental Economics
Nicolas Jacquemet , Olivier L'Haridon
Frontmatter
More Information

Figures

1.1 Trends in academic publishing in experimental economics page 5

1.2 Market equilibrium in the Chamberlin (1948) experiment 7
1.3 Observed behaviour in the Chamberlin (1948) experiment 8
1.4 Predicted and observed behaviour in the Smith (1962) replication 9
1.5 Table of payoffs in a non-cooperative game 13
1.6 Empirical behaviour in prisoners’ dilemma games 15
1.7 A simple four-moves sequential game 16
1.8 A six-moves centipede game 16
1.9 Payoff matrices of two zero-sum games 18
1.10 Empirical value functions 20
1.11 The actual use of information: informed players’ behaviour 21
2.1 Consent form 28
2.2 First screen: resale value in the first round 32
2.3 Second screen: bid in the first round of play 32
2.4 Third screen: results of the first round 33
2.5 The sixth round of the experiment: screen captures 34
3.1 The challenge of data analysis 54
3.2 The econometric approach to data analysis 56
3.3 Incentive effects of tournaments 81
3.4 Dispersion of efforts in tournaments 82
4.1 Meta-analysis results: the dictator game 97
4.2 The effect of social distance on dictators’ decisions 98
4.3 Offers in the dictator game with earned money 100
4.4 Donations from dictators who earned their position 101
4.5 Generous decisions by dictators are taken slowly 103
4.6 Cooperation in repeated games with different termination rules 108
4.7 Reciprocity in the field 114
5.1 Empirical free riding in VCM games 129
5.2 A typical display for an experimental quadratic scoring rule 176
5.3 A typical display for eliciting matching probabilities 183
6.1 Typical implementation of an experimental lab 192
6.2 An experimental lab: what it looks like 194
6.3 A basic experimental algorithm based on the dictator game 198

© in this web service Cambridge University Press www.cambridge.org

Cambridge University Press
978-1-107-06027-2 — Experimental Economics
Nicolas Jacquemet , Olivier L'Haridon
Frontmatter
More Information

List of Figures ix

6.4 A typical experimental session 206

6.5 An example of a time trade-off curve 228
7.1 Visual representations of data 234
7.2 Box plots for different distributions 235
7.3 Normal probability plots 236
7.4 A scatter plot 237
7.5 Anscombe’s quartet 239
7.6 Transformation functions and normality 243
7.7 An illustration of the central limit theorem 246
7.8 Confidence intervals on samples from a population with parameter θ 249
7.9 Critical values for common distributions: normal, t and χ 2 250
7.10 Critical values and rejection regions 259
7.11 Hypothesis testing 260
7.12 Power under different alternative hypotheses 262
7.13 The bracketing procedure used in L’Haridon and Vieider (2015) 296
7.14 The Binswanger (1980) method in Carpenter and Cadernas (2013) 298
7.15 Trade-off sequences and elicited utility under risk and uncertainty 307
8.1 The identification of heterogeneous treatment effects 332
8.2 Many very heterogeneous treatment effects 333
8.3 Social preferences when the monetary stakes are (very) high 343
8.4 Other-regarding behaviour in non-WEIRD populations 349
9.1 The chosen numbers in the Nagel (1995) guessing games 371
9.2 The distribution of behaviour over time in the guessing game 373
9.3 Early matches in the Kagel and Roth (2000) experiment 384
9.4 Matches by productivity type in the Kagel and Roth (2000) experiment 384
9.5 401(k) participation by tenure in Company A in Choi et al. (2004) 390
9.6 Individual welfare optima and consistent arbitrariness 394

© in this web service Cambridge University Press www.cambridge.org

Cambridge University Press
978-1-107-06027-2 — Experimental Economics
Nicolas Jacquemet , Olivier L'Haridon
Frontmatter
More Information

Tables

1.1 The choice sequence of the Allais paradox page 10

1.2 Observed continuation decisions in centipede games 17
1.3 Theoretical predictions in the non-revealing and fully revealing games 19
2.1 Empirical revelation properties of a second-price auction 39
3.1 Individuals, treatments and observations 63
4.1 Gift exchange in the field: donation patterns 114
5.1 Voluntary contributions without altruism 132
5.2 Smith (1982) precepts: three incentive-compatibility criteria 133
5.3 Outcome-based social preferences in the prisoners’ dilemma game 144
5.4 Multiple treatment variables: a 2 × 2 factorial design 158
5.5 A quadratic scoring rule 175
5.6 Examples of binary scoring rules 176
5.7 The constant-sum game in Nyarko and Schotter (2002) 188
6.1 Example of binary choices used by Tanaka et al. (2010) 209
6.2 The price list in Coller and Williams (1999) 216
6.3 The treatments in Coller and Williams 217
6.4 Four procedures to elicit indifference in choice over time 218
6.5 The convex time-budget method 224
6.6 The choice list in the direct-method elicitation 226
7.1 An example of experimental data based on second-price auctions 230
7.2 Descriptive statistics 238
7.3 Correlation measures and the Anscombe quartet 240
7.4 True data-generating process and decisions 258
7.5 Frequently used statistical tests 266
7.6 The ANOVA decomposition 280
7.7 A 2 × 2 table for independent samples 285
7.8 A 2 × 2 table for paired samples 289
7.9 Elicitation methods 291
7.10 An example of the bisection procedure 295
7.11 The bracketing procedure in Tversky and Kahneman (1992) 295
7.12 The payoffs and risk classification in Binswanger (1980) 297
7.13 The payoffs and risk classification in Eckel and Grossman (2008) 297
7.14 The ten paired lottery-choice decisions in Holt and Laury (2002) 299

© in this web service Cambridge University Press www.cambridge.org

Cambridge University Press
978-1-107-06027-2 — Experimental Economics
Nicolas Jacquemet , Olivier L'Haridon
Frontmatter
More Information

List of Tables xi

7.15 Lottery-choice decisions and the CRRA index 299

7.16 The ten paired lottery-choice decisions in Drichoutis and Lusk (2016) 302
8.1 In-lab versus online experiments: overview of experimental
comparisons 339
8.2 Calculation of the false-positive report probability 354
8.3 Replication versus robustness: a classification 357
9.1 Level classification in the control, graduate and computer treatments 379
9.2 The distribution of behaviour in the 11–20 Game 380
9.3 The Newcastle algorithm: a fictional example 383

© in this web service Cambridge University Press www.cambridge.org

Cambridge University Press
978-1-107-06027-2 — Experimental Economics
Nicolas Jacquemet , Olivier L'Haridon
Frontmatter
More Information

Illustrations

2.1 Second-price auctions as a preference revelation mechanism:

home-grown and induced values page 42
2.2 An experimental comparison of correction methods 46
3.1 Labour market effects of the minimum wage: a natural experiment 52
3.2 Incentives and performance: a ‘natural’ experiment 61
3.3 The need for assumptions on the data-generating process to achieve
inference (even) from experimental evidence 67
3.4 Incentives and performance: selection and incentive effects 73
3.5 Gender differences in competitiveness: experimental evidence from
exogenously chosen composition of groups 76
3.6 Piece rate: a field experiment 77
4.1 Whispering in the ears of antitrust authorities 91
4.2 Models as a reduction of reality: firms’ behaviour in collusion theory 93
4.3 Reciprocity at work: the fair-wage-effort hypothesis 104
4.4 Experimental evidence on the fair-wage-effort hypothesis 105
4.5 Trust: evidence from the lab 111
5.1 Endowment effects in market behaviour 123
5.2 Identified failures of internal validity: misconceptions about the
endowment effect 125
5.3 Saliency and coordination: experimental evidence based on the stag
hunt game 135
5.4 Evidence from non-incentivised behaviour: the status quo effect 137
5.5 The effect of incentives on experimental outcomes 138
5.6 Social preferences and strategic uncertainty: the ultimatum-bargaining
game 142
5.7 Altruism in the prisoners’ dilemma game 145
5.8 Outcome versus intention: an experiment on the nature of social
preferences 146
5.9 The effect of roles on behaviour: the Stanford prison (aborted)
experiment 148
5.10 Controlling for closeness: the inclusion-of-the-other-in-the-self scale 150
5.11 Individual consistency of social preferences: a within-subject design 154
5.12 Evidence of order effects: rationality spillovers 155

© in this web service Cambridge University Press www.cambridge.org

Cambridge University Press
978-1-107-06027-2 — Experimental Economics
Nicolas Jacquemet , Olivier L'Haridon
Frontmatter
More Information

List of Illustrations xiii

5.13 VCM: a 4 × 2 factorial design 157

5.14 Identified failures of internal validity: confusion in VCM games 160
5.15 Identified failures of internal validity: game form recognition in beauty
contest games 164
5.16 Belief elicitation and outcome behaviour in a VCM game 167
5.17 The effect of closeness and the ability to coordinate 168
5.18 The accuracy of self-reported expectation measures 172
6.1 Experimental instructions for a simple dictator game 197
6.2 Information provided to prospective participants in economic
experiments 202
6.3 Information provided to prospective participants in economic
experiments (continued) 203
6.4 Instructions for a time-preference-elicitation experiment 212
6.5 Eliciting indifferences via bisection 214
8.1 Reversed external validity: experimental evidence on the winner’s
curse in real auction markets 324
8.2 The measure of corruption from laboratory bribery behaviour 328
8.3 The external validity of gift exchange at work 330
8.4 Laboratory evidence of the external validity of declarative surveys 337
8.5 The predictive power of experimental time-preference measures 341
8.6 External validity of free riding in voluntary-contribution mechanisms 345
8.7 Overcoming coordination failures thanks to complexity 347
8.8 Self-selection in laboratory experiments 350
8.9 The winner’s curse with experienced bidders 352
9.1 Market size and collusion: ‘two are few and four are many’ 363
9.2 The hidden cost of incentives: motivation crowding out 364
9.3 The informational content of incentives: an experimental test 366
9.4 Preference reversal in a market situation 368
9.5 The market-entry game 374
9.6 Strategic thinking in the centipede game 376

© in this web service Cambridge University Press www.cambridge.org

Cambridge University Press
978-1-107-06027-2 — Experimental Economics
Nicolas Jacquemet , Olivier L'Haridon
Frontmatter
More Information

Focuses

2.1 Preference elicitation and policy-making: the hypothetical bias page 37

2.2 Preference elicitation: auctions, referenda and BDM mechanisms 44
3.1 Causal effects in theoretical analysis and empirical works 62
3.2 The programme evaluation approach and the structural approach 66
3.3 Incentives and performance: the confounding effect of self-selection 70
3.4 Two additional difference estimators and their identifying assumptions 74
4.1 On the use of response times to interpret observed behaviour in
experiments 102
5.1 Cold versus hot: available measures of outcome behaviour 121
5.2 Loss aversion: a behavioural foundation for the endowment effect 124
5.3 Equilibrium analysis of the VCM game 130
5.4 Incentive-compatible compensation of repeated choices: the random
incentive system 140
5.5 Intention-based social-preference models 143
5.6 Economics and psychology: an overview of the main methodological
disagreements 165
5.7 Prediction markets 174
5.8 Measuring beliefs over a continuous random variable 177
5.9 The binarised scoring rule 178
5.10 Risk aversion and hedging in experimental games 179
5.11 Using matching probabilities to test complex ambiguity models 182
5.12 Comparing elicitation methods 185
5.13 Experimental designs for ambiguity 186
6.1 The discounted-utility model 210
6.2 Behavioural foundations of the discounted-utility model 211
6.3 Accounting for non-linear utility 215
7.1 Censored and truncated data 232
7.2 Distance correlation as a measure of the degree of association 241
7.3 The exploratory analysis of treatment effects with odds ratios 242
7.4 Bayesian parameter estimation 247
7.5 Sample size and confidence intervals 252
7.6 Prediction intervals for a single observation 253
7.7 A five-step approach to hypothesis testing 257

© in this web service Cambridge University Press www.cambridge.org

Cambridge University Press
978-1-107-06027-2 — Experimental Economics
Nicolas Jacquemet , Olivier L'Haridon
Frontmatter
More Information

List of Focuses xv

7.8 Multiple test procedures 261

7.9 Sample-size determination 263
7.10 Bayes factors 265
7.11 The likelihood-ratio test 267
7.12 Testing for outliers 269
7.13 Goodness-of-fit tests and the normality hypothesis 274
7.14 Testing for randomness: the run test 275
7.15 Two-way and multi-way ANOVA 282
7.16 The balloon analogue risk task (BART) 292
7.17 Portfolio choice and the elicitation of risk attitudes 293
7.18 Incentives and repeated choice 300
7.19 Comparing standard-gamble methods 303
7.20 Survey questions and the measurement of risk attitudes 304
7.21 Comparing standard-gamble and value-equivalence methods 308
7.22 The basic prospect-theory model 310
7.23 Measuring loss aversion 311
7.24 Prospect theory with uncertainty and ambiguity 312
7.25 Probability weighting in choice under risk 313
7.26 Stochastic choice 315
8.1 The many different meanings of external validity in experimental
psychology 326
9.1 The cognitive-hierarchy model 375
9.2 An alternative theoretical model of strategic thinking: quantal-response
equilibrium 377
9.3 Designing a liberal and paternalistic choice architecture 388
9.4 Opt-in/opt-out versus active decisions: a non-liberal-paternalistic tool
to enhance enrolment in 401(k) without default 392
9.5 The malleability of consumer preferences: anchoring and consistent
arbitrariness 395

© in this web service Cambridge University Press www.cambridge.org

Cambridge University Press
978-1-107-06027-2 — Experimental Economics
Nicolas Jacquemet , Olivier L'Haridon
Frontmatter
More Information

Abbreviations and Symbols

Abbreviations
AD Aggregate Demand
ATE Average Treatment Effect
ATT Average Treatment on the Treated
BART Balloon Risk Analogue Task
BDM Becker-De Groot-Marschak
BMI Body Max Index
CADI Constant Absolute Decreasing Impatience
CDF Cumulative Distribution Function
CE Certainty Equivalence
CHM Cognitive-Hierarchy Model
CRDI Constant Relative Decreasing Impatience
CRRA Constant Relative Risk Aversion
DARA Decreasing Absolute Risk Aversion
DA Deferred Acceptance algorithm
DGP Data Generating Process
DM Dissonance Minimization
ECU Experimental Currency Unit
FPRP False Positive Report Probability
FR Fully-revealing game
FTC Federal Trade Commission
FW Fixed wage
HSD Honestly Significant Difference
IEC Institutional Ethics Committee
IOS Inclusion of the Other in the Self
IQR Interquartile Range
IRB Institutional Review Board
IV Induced Value
LHS Left-Hand Side
LSD Least Significant Difference
MARS Meta-Analysis Reporting Standards
MD Mean absolute Deviation
MLE Maximum Likelihood Estimator
MOOSE Meta-analysis of Observational Studies in Epidemiology

© in this web service Cambridge University Press www.cambridge.org

Cambridge University Press
978-1-107-06027-2 — Experimental Economics
Nicolas Jacquemet , Olivier L'Haridon
Frontmatter
More Information

Abbreviations and Symbols xvii

MPCR Marginal per Capita Return

MSE Mean Squared Error
MT Amazon’s Mechanical Turk
MT Mechanical Turk
MT Western Educated, Industrialized, Rich, and Democratic
NR Non-revealing game
OLS Ordinary Least Squares
PEEM Portable Extensions of Existing Models
PE Probability Equivalence
PGG Public Good Game
PRISMA Preferred Reporting Items for Systematic Reviews and Meta-Analyses
PR Piece-rate
Q-Q Quantile-Quantile
QRE Quantal-Response Equilibrium
RDU Rank-Dependent Utility
RHS Righ-Hand Side
RIS Random Incentive System
UBG Ultimatum Bargaining Game
VCM Voluntary Contribution Mechanism
WEIRD Western Educated, Industrialised, Rich, and Democratic
WTA Willingness to Accept
WTP Willingness to Pay
WVS World Value Survey
WVS World Values Survey

Symbols
ȳ sample average
 variation
δ exponential discount factor, parameter
ℓ effort
η decision error
θ̂ estimator
λ, γ parameters
E expectation
B bias
T test statistic
X matrix of individual observations, e.g observable characteristics
y vector of the observations on the outcome variable
I beliefs in bayesian estimation
L sampling distribution
N normal distribution
S state space
T treatment

© in this web service Cambridge University Press www.cambridge.org

Cambridge University Press
978-1-107-06027-2 — Experimental Economics
Nicolas Jacquemet , Olivier L'Haridon
Frontmatter
More Information

xviii Abbreviations and Symbols

X inputs
Y outputs
μ mean
variance-covariance matrix
ω() probability weighting function
B Binomial distribution
dCor distance correlation
dCov distance covariance
Fl , Fu critical values of the Fisher distribution
standard normal cumulative distribution
φ standard normal density
π profit
ε vector of error terms
ρ Pearson correlation coefficient
σ,ψ standard deviations
 parameter space
V variance
εi individual error terms
a, b, A, B general purpose parameters (actions, prizes, bids. . . )
bL lower bound of confidence interval
bU upper bound of confidence interval
c threshold in hypothesis testing
ce () cost of effort
d0 , d1 decisions in hypothesis testing
DR decision rule
e endowment
F(), f () functions
G() cumulative distribution function
g() density
h, i, j, k, s, t indexes
H0 , H1 , Ha statistical hypothesis
K number of samples, treatments, classes. . .
L() likelihood
LL() log-likelihood
m number of observable characteristics, median
N population size
n number of observations, sample size, number of modeling features
nX number of inputs
nY number of outputs
p, Pr probability
p(k) rank-ordered p-value
q, Q price, returns
r rank
rr rate of return

© in this web service Cambridge University Press www.cambridge.org

Cambridge University Press
978-1-107-06027-2 — Experimental Economics
Nicolas Jacquemet , Olivier L'Haridon
Frontmatter
More Information

Abbreviations and Symbols xix

S2 sample variance
SS sum of squares
T time, date, period
tα critical value of the Student t distribution
U(), V() preference functionals
u(), v() utility functions
w wage
X, Y random variables
x, y realization of random variables
Y(h) ordered value of Y (with order h)
Z dummy variable
zα critical value of the normal distribution
α Type I error
β Type II error
θ parameter(s)
E event
pτ tremble
R rejection region in hypothesis testing
W() event weighting function
xij observation for subject i and variable j
yi observation on the outcome variable for subject i

© in this web service Cambridge University Press www.cambridge.org

Cambridge University Press
978-1-107-06027-2 — Experimental Economics
Nicolas Jacquemet , Olivier L'Haridon
Frontmatter
More Information

© in this web service Cambridge University Press www.cambridge.org

Cambridge University Press
978-1-107-06027-2 — Experimental Economics
Nicolas Jacquemet , Olivier L'Haridon
Frontmatter
More Information

Preface

There is an experimental-economics paradox. Inside the community of researchers car-

rying out laboratory experiments, these latter are seen as no more and no less than a
tool for empirical research. From the outside, however, the method is often perceived
as part of a particular sub-field, behavioural economics, which applies insights from
both economics and psychology for the better understanding of economic behaviour.
Experimental economics is also usually taught this way in most programmes, as part of
behavioural-economics classes.
It has, however, long been recognised that experimental and behavioural economics
are not the same. Behavioural economics is a research programme with a clear ambition
and a well-defined objective: improving economic analysis using realistic psychological
assumptions about human behaviour. Experimental economics, on the contrary, is not,
per se, a research programme. Rather, it is a research method based on experimental
control, applied to the typical topics in economic analysis.
The aim of this textbook is to help close the gap between the perception and reality
of experimental methods in economics. We cover experimental economics, i.e. con-
trolled experiments used as a tool to provide empirical evidence that is relevant for
economic research. The structure of the textbook thus mimics the way many economet-
rics textbooks have been written for decades: the coverage focuses on applied statistical
methods, the use of which is illustrated with economic results.
There are, however, a number of (good) reasons for this confusion between
behavioural and experimental economics, which is at the heart of the experimental–
economics paradox. First, behavioural economics emerged partly from the use of
experiments – although the contribution of early experiments (such as the Allais para-
dox and the Chamberlin and Smith market experiments, described in Chapter 1) was
to both behavioural economics and mainstream economics (for instance, neoclassical
market analysis). Second, the experimental economics method is particularly suited for
the study of the phenomena of interest to behavioural economics. In a nutshell, con-
trol offers researchers a way of identifying departures from the neoclassical explanation
of behaviour. Third, not only behavioural economics but also experimental economics
owe a great deal to the accumulated knowledge in experimental psychology: controlled
experiments have been used for a long time in this field, and most methodologi-
cal discussions took place before they even appeared in economics. In addition, the

© in this web service Cambridge University Press www.cambridge.org

Cambridge University Press
978-1-107-06027-2 — Experimental Economics
Nicolas Jacquemet , Olivier L'Haridon
Frontmatter
More Information

xxii Preface

experimental method is taken as part of the psychology research toolkit across the whole
community of researchers.
The scope of this book has been greatly influenced by the place that experimental eco-
nomics occupies between neoclassical economics, behavioural economics, psychology
and statistics. First, our methodological discussion mainly focuses on the use of exper-
iments to understand economic behaviour. We complement this fairly standard view in
applied economics by regularly devoting space to insights from, and some discrepan-
cies with, psychology. We also cover a number of standard experimental results that are
generally seen as part of behavioural economics.
Second, we mainly focus on laboratory experiments rather than field experiments or
randomised controlled trials (see Chapter 3, Section 3.5 for the definition of these). This
restriction reflects at least three factors. First, one textbook cannot suffice to embrace
the large literature on methods for both laboratory experiments and randomised con-
trolled trials. Second, this restriction also comes from our own limitations in expertise.
Last, but not least, laboratory experiments are a convenient step in the study of con-
trolled experiments in economics. Laboratory experiments can be seen as an extreme
case of controlled experiments; they allow the accurate identification of behavioural
phenomena, but at the cost of a highly artificial environment. Due to this artificiality,
laboratory experiments provide answers that are sometimes hard to interpret – and are
often challenged by non-experimentalists. Other kinds of experiment offer a way of
loosening these limitations by implementing the same empirical method in less artifi-
cial contexts. We thus believe that laboratory experiments are a good starting point for
anyone who wants to learn about controlled experiments in economics. Many of the
discussions in this textbook aim to clarify the most appropriate cases for each type of
empirical method; for example, whether observational or experimental data are required
and, if it is experimental data, how close to the field the experiment should be.

Structure of the book

This textbook is not the first experimental-economics book by a long way, with respect
to both methods and applications. Our predecessors can be split into two groups. First,
textbooks/handbooks written for students and academics provide extensive surveys of
experimental results. This applies to the textbook of Friedman and Sunder (1994) and
the two seminal handbooks edited by Plott and Smith (2008) and Kagel and Roth (1995).
In the same spirit, a number of books propose reviews of existing results from laboratory
experiments with more specialised perspectives: Camerer (2003) contrasts behaviour in
the lab with predictions from game theory, Cartwright (2011) and Chaudhuri (2009)
mainly focus on social preferences and behavioural economics, and Angner (2012) pro-
vides a detailed overview of laboratory experiments regarding decision problems. These
are all required reading for anyone wanting to learn more about experimental results.
On the other hand, a few advanced books on the methodology of experiments have
recently appeared. These are state-of-the-art collections of papers, written mainly for

© in this web service Cambridge University Press www.cambridge.org

Cambridge University Press
978-1-107-06027-2 — Experimental Economics
Nicolas Jacquemet , Olivier L'Haridon
Frontmatter
More Information

Preface xxiii

academics working in the field. This is the case for Guala (2005), Bardsley et al. (2009)
and Fréchette and Schotter (2015).
This textbook is an attempt to build a bridge between these two kinds of reference:
it provides a detailed presentation of the methodological aspects of economic experi-
ments for readers (students, academics and professionals) who want to enter the field.
To this end the book inverses the usual way of presenting the material, as the experi-
mental results are used to illustrate methodological issues – rather than spreading out
the methodological discussions over the presentation of various experimental designs.
The content of the book is set out at the end of Chapter 1. We are aware that ‘Method-
ology, like sex, is better demonstrated than discussed, though often better anticipated
than experienced’ (Leamer, 1983, p. 40). Mimicking the approach in applied economics
and econometrics textbooks, the concrete applications of the method that constitute the
core material in existing textbooks are here introduced as illustrations of the main mate-
rial. To this end, the book contains three types of side material describing particular
experiments, results or designs: case studies, illustrations and focuses.

• Case studies are sections devoted to the detailed presentation of a particular strand of
experiments. They seek to illustrate the methodological discussions provided in the
corresponding chapter – identified as such in the table of contents.
• Illustrations are boxes providing a presentation of one particular experiment or
result, to illustrate the point discussed in the text. Illustrations are often provided
in sequences, showing how the literature has evolved according to the different
dimensions discussed in the text.
• Focuses are boxes providing a more detailed and/or formal presentation of a point
discussed in the text.

These together provide examples of most of the applications or results that are gen-
erally seen as essential in the field – as described in Section 1.4. To help readers bring
together all of the information on one particular topic, they appear as specific index
headers (see p. 431).

Audience

There are three natural audiences for this book. Its first purpose is as part of a graduate
course, describing methods in experimental economics. The organisation of the book
closely follows the typical outline of an 8 × 3-hour course. Chapters 1–4 cover the
material that would serve as an introductory lecture to laboratory experiments. These
chapters describe the main objectives of laboratory experiments and provide exam-
ples. Chapters 5 and 8 provide core methodological insights that would best be split
in two lectures each. Longer classes could include a discussion of the statistical analysis
of experimental data based on Chapter 7 and a discussion of the insights drawn from
behavioural economics in Chapter 9, and/or use case studies to devote some lectures to
applications that illustrate the main material. In particular, a thorough methodological

© in this web service Cambridge University Press www.cambridge.org

Cambridge University Press
978-1-107-06027-2 — Experimental Economics
Nicolas Jacquemet , Olivier L'Haridon
Frontmatter
More Information

xxiv Preface

course would probably feature some lectures devoted to risk preferences (Section 7.4),
time preferences (Section 6.6) and belief-elicitation methods (Section 5.6).
Second, the book more generally seeks to provide future experimental practitioners
with a broad picture of the toolkit that they will need. By providing the rationale for the
general method and setting out in detail each particular choice of design feature, we hope
that readers will be able to construct experiments that fit their research question well. A
good understanding of the methodological challenges is also an important requirement
for becoming an informed reader: this book may help to interpret the results from lab-
oratory experiments or the writing of referee reports on papers using the experimental
method. Third, we hope the community of academics who are new to this literature will
find it a useful summary of the current state of the art about what experimental eco-
nomics can tell us, and under which conditions it provides valuable answers to research
questions in economics.

Acknowledgements

The book was written using the course material for PhD/master 2 courses in a number
of different places, and in particular at our home institutions. We are more than grateful
to the students who attended these classes for their commitment, remarks, scepticism
and enthusiasm. We gratefully acknowledge the support from the Institut Universitaire
de France.
It is likely that the book would never have reached its final stage without the encour-
agement, help and remarks from, and discussions with, Jay Shogren. The writing process
took such a long time that we will certainly omit many people whose contributions at
earlier stages were much appreciated. This also meant that we have worked with many
research assistants, whose help very often exceeded what was expected. Our thanks to
Lisa Simon and Solene Delecourt for their work on early drafts of some of the chap-
ters; Sophie Cottet for producing the graphs and figures; and Alberto Prati, Guillaume
Royer and Shaden Shabayek for their work on some of the boxes. Last, an incredi-
ble number of PhD students and colleagues spent a great deal of time reading the first
drafts of different parts of the book and provided us with invaluable feedback. We grate-
fully thank Arthur Attema, Aurélien Baillon, Han Bleichrodt, Aurélie Bonein, Elias
Bouacida, Béatrice Boulu-Reshef, Arthur Charpentier, Paolo Crosetto, Laurent Denant-
Boémont, Antoine Hémon, Justine Jouxtel, Antoine Malézieux, Elven Priour, Kirsten
Rohde, Angelo Secchi, Benoit Tarroux and Adam Zylbersztejn.
While the field of behavioural and experimental economics is sometimes described as
over-competitive, it is also one in which researchers from all over the world cooperate
on methodological and bibliographic issues, thanks to the ESA discussion group: the
discussions there provided us many insights and ideas for which we gratefully thank
all contributors. Our gratitude also goes to Sandra Freeland and Andrew Clark for their
thorough proofreading of the manuscript, and the editorial team at Cambridge Univer-
sity Press, Phil Good, Neil Ryan and Chris Harrison, for their continuous support and
outstanding work.

© in this web service Cambridge University Press www.cambridge.org

View publication stats