14th TIBER Symposium
on Psychology and Economics
August 27th 2015, Tilburg University
Download this program on our website:
www.tilburguniversity.edu/tiber14
When tweeting about TIBER 14, please use #tiber14
Index
3-5
Schedule
7 - 30
Abstracts of presentations in parallel sessions
Confidence, Beliefs, & Knowledge
Cooperation
Social Distance & Social Preferences
Trust & Fairness
Environment & Sustainable Behavior
Bias, Redistribution & Scarcity
Reference Points & Numbers
Status & Prevention Behavior
Pro-social Behavior
Saving & Borrowing
Regret
Cheating
7-8
9 - 10
11 - 12
13 - 14
15 - 16
17 - 18
19 - 20
21 - 22
23 - 24
25 - 26
27 - 28
29 - 30
31 - 33
34
Abstracts of poster presentations
Notes
2
Schedule
Time
Place
8.30 - 9.15
Foyer
Dante
DZ1
Opening of the symposium
DZ1
KEYNOTE Armin Falk: Malleability of moral behavior
9.15 - 9.20
9.20 - 10.20
10.30 - 12.10
SESSION 1
10.30 - 10.55
Event
Registration
/ Coffee or tea
Parallel Sessions 1-4
DZ1
Confidence, Beliefs, & Knowledge
Mira Fischer - Investment in learning and beliefs about knowledge and talent: Experimental evidence on the effects of two
dimensions of confidence
Patrick Gerhard - The effect of past performance framing on investors’ belief updating
Marret Noordewier - Curiosity and time: From not knowing to almost knowing
Joël Van der Weele - Deception and self-deception
SESSION 2
10.30 - 10.55
10.55 - 11.20
11.20 - 11.45
11.45 - 12.10
DZ3
Cooperation
Ayala Arad - Analyzing team communication to study multi-dimensional reasoning
Pippa Beston - Name & shame: The power of social punishment
Jörg Gross - Building the Leviathan – Voluntary centralisation of punishment power sustains cooperation
Shaul Shalvi - Corrupt collaboration
SESSION 3
10.30 - 10.55
10.55 - 11.20
11.20 - 11.45
11.45 - 12.10
DZ4
Social Distance & Social Preferences
Geoffrey Castillo - Preference reversals in the social domain: The effect of social distance on choice and valuation
Fadong Chen - Cognitive processes of prosocial and antisocial punishment
Matthias Greiff - How small differences in context affect the measurement of social value orientation
Maximilian Hoyer - Stand by your man - The role of shared history in investment decisions
SESSION 4
10.30 - 10.55
10.55 - 11.20
11.20 - 11.45
11.45 - 12.10
DZ5
Trust & Fairness
Adrian Chadi - Forced to be generous - Experimental evidence on the behavioral effects of minimum wages outside the laboratory
Uyanga Turmunkh - Not all talk is cheap: Communication and cooperation in a high stakes TV game show
Jeroen van Baar - Kind beyond belief: Fairness norms predict trust game reciprocity independently of second-order expectations
Yilong Xu - Social information and selfishness
10.55 - 11.20
11.20 - 11.45
11.45 - 12.10
3
12.10 - 13.50
Foyer
13.50 - 15.05
SESSION 5
13.50 - 14.15
14.15 - 14.40
Parallel Sessions 5-8
DZ1
Environment & Sustainable Behavior
Marko Milovanovic - Big Brother is watching you
Eline van der Heijden - Going green: Framing effects in a dynamic coordination game experiment
Arianne van der Wal - Temporal myopia in sustainable behavior when facing uncertainty
DZ3
Bias, Redistribution & Scarcity
Joshua Miller - Surprised by the gambler’s and hot hand fallacies? A truth in the law of small numbers
Nickolas Gagnon - With who you work can matter: an experiment on social distance and redistribution preferences in small groups
Jeroen Nieboer - Popularity breeds contempt: Effects of popularity feedback in a speed-dating experiment
DZ4
Reference Points & Numbers
Alessandro Saia - Wishing for bad weather: demand shocks and labor supply of Bologna pizza vendors
Adriaan Soetevent - P(l)aying at the pump: Behavioral biases in buying fuel
Tong Wang - Number Preferences
DZ5
Status & Prevention Behavior
Samuel Franssens - Dishonest conspicuous consumption leads to social costs imposed by social equals
Patty Jansen - Measuring prevention performance through behavorial self-reports: an application of the one-dimensional Rasch
model
Wojtek Przepiorka - The production of a reputation premium: Bargain-hunting and herding in eBay auctions
14.40 - 15.05
SESSION 6
13.50 - 14.15
14.15 - 14.40
14.40 - 15.05
SESSION 7
13.50 - 14.15
14.15 - 14.40
14.40 - 15.05
SESSION 8
13.50 - 14.15
14.15 - 14.40
14.40 - 15.05
15.05 - 15.25
LUNCH + POSTER SESSION
Foyer
Coffee + tea
4
15.25 - 16.40
Parallel Sessions 9-12
SESSION 9
15.25-15.50
15.50-16.15
16.15-16.40
DZ1
Pro-social Behavior
Rene Bekkers - Altruism, warm glow, and generosity: a national experiment
David Reiley - Hassle costs and charitable giving: A field experiment with google employees
Paul Smeets - Why do investors hold socially responsible mutual funds?
SESSION 10
15.25-15.50
15.50-16.15
16.15-16.40
DZ3
Saving & Borrowing
Job Krijnen - An importance heuristic in decision deferral
Benjamin Marx - Default bias in borrowing: evidence from a field experiment on federal student loans
Anett John - When commitment fails - evidence from a regular saver product in the Philippines (Cancelled)
SESSION 11
15.25-15.50
15.50-16.15
16.15-16.40
DZ4
Regret
Andreas Nicklisch - Adversity is a school of wisdom: Experimental evidence on cooperative protection against stochastic losses
Philippe Van de Calseyde - How doubt affects the experience of regret
Koen van der Swaluw - The weight loss lottery as a commitment device. Who will play, and what will they feel?
SESSION 12
15.25-15.50
15.50-16.15
16.15-16.40
DZ5
Cheating & Fairness
Celia Blanco - License to cheat under loss aversion
Jan Stoop - Do cheaters in the lab also cheat in the field?
Sara Rezaei - Towards a theory of marginal fairness concerns
16.40- 17.00
Foyer
Coffee + tea
17.00 - 18.00
DZ1
KEYNOTE Maya Bar-Hillel: Position effects in simultaneous choice - A triple conundrum solved
18.30 - 22.00
Faculty
Club
Optional: Dinner at the Faculty Club (for those who registered)
5
6
Oral presentations - Abstracts
10.30 - 11.20: parallel session 1: Confidence, Beliefs, & Knowledge
10.30-10.55
10.55-11.20
Speaker: Mira Fischer - University of Cologne
Additional Author: Dirk Sliwka
Speaker: Patrick Gerhard - Maastricht University
Additional Authors: Arvid O. I. Hoffmann and Thomas Post
Investment in learning and beliefs about knowledge and talent: Experimental
evidence on the effects of two dimensions of confidence
Motivational beliefs are held to be an important determinant of success in
education. Whereas in educational settings it is often claimed that holding
overly optimistic beliefs about own ability affects learning positively,
economists have mostly pointed at the detrimental effects of overconfidence.
The purpose of this paper is to analyze the separate role of confidence about
prior knowledge and confidence in the ability to learn for the incentives to
make human capital investments. We study a real effort experiment in which
agents receive randomly biased feedback about their performance in a
knowledge and in a memory test before making a decision to acquire
information that helps in a third test covering the same dimension as the
knowledge test. We identify a positive effect of confidence in learning and a
negative effect of confidence in knowledge on investments. Both effects are
driven by the better half of participants in the knowledge dimension. Our
results show that generalized statements about the role of confidence can be
misleading and confidence should be viewed as a multidimensional concept.
This leads to insights not only for the education of children but also for the
design of feedback policies in firms. While raising confidence in the ability to
acquire a certain skill can be beneficial, raising confidence in the skill itself
can be detrimental.
The effect of past performance framing on investors’ belief updating
We analyze how past performance framing impacts investors’ belief
updating. Belief updates drive investment decisions. Larger updates lead to
more frequent trading, which reduces investment performance. Hence, we
are interested in reducing the magnitude of investors’ belief updating. To do
so, we present 377 experimental subjects with past performance information
and subsequently measure updates of their beliefs (return expectations and
risk perceptions) over multiple rounds. We employ three different frames,
varying the default information horizon (annual, monthly, daily). An
important distinction compared to previous work is that subjects can easily
and without costs opt out of the default and obtain past performance
information on each of the three horizons. Building on research on default
options, framing, and belief updating, we expect that a longer default
information horizon reduces belief updating. However, we find that, across
all subjects, the default option does not have an impact on the magnitude of
subjects’ belief updating. Dividing the sample into subjects staying in versus
opting out of the default, however, uncovers an important pattern of results.
While showing returns over a longer information horizon reduces the
magnitude of belief updates for subjects staying in this default, we find the
opposite effect for subjects opting out of this default. Financial literacy is
positively associated with the tendency to opt out of the default. Our results
show that default options potentially influence individual investors’ decision
making, but when opting out of the default is easy, they need to be carefully
selected to have a beneficial effect.
7
11.20 - 12.10: parallel session 1: Confidence, Beliefs, & Knowledge
11.20 - 11.45
11.45-12.10
Speaker: Marret Noordewier - Leiden University
Additional Author: Eric van Dijk
Speaker: Joël J. van der Weele - Universiteit van Amsterdam
Additional Author: Peter Schwardmann
Curiosity and time: From not knowing to almost knowing
How does it feel to be curious? We reasoned that there are two sides to
curiosity: not knowing something (i.e., information-gap) and almost knowing
something (i.e., anticipation of resolution). In different experiments, we
showed that time affects the relative impact of these two components: When
people did not expect to close their information-gap soon (big time-gap), the
anticipation of the resolution was weaker and not knowing affected the
experiential content of curiosity more strongly than when they expected to
close their information-gap quickly (small time-gap). As such, people
experienced less positive affect, more discomfort, and more annoyance with
lack of information in a big than a small time-gap situation. Moreover, when
time in the big time-gap passed, the anticipation of the resolution became
stronger, positive affect increased, and discomfort and annoyance with lack
of information decreased. Time is thus a key factor in the experience of
curiosity. Theoretical and practical implications will be discussed, as well as
follow-up studies in the context of teaser advertising.
Deception and self-deception
Overconfidence has been documented in countless studies, but given the
costs of false beliefs, it is still a puzzle why people are so often
overconfident. We experimentally investigate the determinants of
overconfidence and test the hypothesis that overconfidence serves to more
effectively persuade or deceive others. After performing a cognitively
challenging task, subjects report beliefs about their performance relative to
others in a private elicitation with incentives for truthful reporting. Across
treatments we vary the possibility for subjects to earn money by persuading
others about their relative performance in highly structured face-to-face
interactions. We find that the shadow of future interactions increases
overconfidence by about 50% and reduces responsiveness to new
information about performance. In addition, individuals who self-report
having a dominant personality engage in more self-deception, regardless of
the experimental condition. In the ensuing interactions, we find that
confidence increases earnings more than does actual performance. When
their counterparts have been trained in lie detection, this effect reverses and
it is actual performance that determines earnings. These results suggest thus
that self-deception is prominent and may be beneficial in daily interactions
where the receiver is not trained or focused on detecting real ability or
performance.
8
10.30 - 11.20: parallel session 2: Cooperation
10.30-10.55
10.55-11.20
Speaker: Ayala Arad - Tel Aviv University
Additional Author: Stefan Penczynski
Speaker: Pippa J. Beston - Bangor University
Additional Author: Erin A. Heerey
Analyzing team communication to study multi-dimensional reasoning
We experimentally study individuals’ decision procedures in various complex
interactions, such as multi-object auctions and resources allocation games.
In our experiment, a team of two individuals plays as one entity against other
teams, and teammates communicate through text messages before
choosing their team’s strategy. Analysis of the messages reveals that players
think in terms of dimensions (or features) of strategies rather than strategies
per se, as suggested in Arad and Rubinstein (2012). It appears that players
make a decision for each dimension separately and then pick a strategy that
is consistent with those choices. We further investigate this multidimensional reasoning by identifying the perceived dimensions of a strategy,
which are context-dependent, and the common decision rules used within
each dimension.
Name & shame: The power of social punishment
Cooperation in social games is an economic puzzle. This strategy reduces
individual payoffs and is therefore not rational. Yet, cooperation is widely
observed in such games, particularly when the option to punish free-riders is
present. Research in laboratory settings suggests monetary punishments are
superior in maintaining cooperation to social/reputation-based mechanisms.
This is surprising as humans are inherently social beings, finely attuned and
inclined to respond to social feedback. Here, we investigated the premise
that monetary punishments reform uncooperative behaviour more effectively
than social sanctions. Four participants played either an anonymous webbased public goods game in a campus computer lab, or the same game in a
face-to-face social environment. We tested the effects of three punishment
conditions: monetary (return reduced by 50%), social (lowest contributor
named), and combined (monetary + social punishments) on participants’
contributions. In the anonymous setting, we replicated the common finding
that monetary punishments improve investments over social punishments.
However, results reversed in the social setting. Here, “name and shame”
punishments enhanced investments more than monetary sanctions.
Independent ratings of group positive affect (shared laughter/smiles) from
videos predicted average contribution on the next trial, such that
investments increased with positivity. In the real world, individuals often
make economic and cooperative decisions within social environments.
Evidence suggests that they integrate social and monetary feedback to form
a unitary estimate of reward value. We therefore suggest that rewards
provided by interaction partners drive cooperation by enhancing perceived
outcomes, making social rewards and punishments a critical aspect of the
economic environment.
9
11.20 - 12.10: parallel session 2: Cooperation
11.20-11.45
Speaker: Jörg Gross - University van Amsterdam
Additional Authors: Zsombor Z. Meder, Sanae Okamoto-Barth, and Arno
Riedl
Building the Leviathan – Voluntary centralisation of punishment power
sustains cooperation
The prevalence of cooperation among humans is puzzling because
cooperators can be exploited by free riders. Peer punishment has been
suggested as a solution to this puzzle, but cumulating evidence questions its
robustness in sustaining cooperation. Amongst others, punishment fails
when it is not powerful enough, or when it elicits counter-punishment.
Existing research however has ignored that the distribution of punishment
power can be the result of social interactions. We introduce a novel
experiment in which individuals can transfer punishment power to others.
We find that while decentralised peer punishment fails to overcome free
riding, the voluntary transfer of punishment power enables groups to sustain
cooperation. This is achieved by non-punishing cooperators empowering
those who are willing to punish in the interest of the group. Our results show
how voluntary power centralisation can efficiently sustain cooperation, which
could explain why hierarchical power structures are widespread among
animals and humans.
11.45-12.10
Speaker: Shaul Shalvi - Ben Gurion University
Additional Author: Ori Weisel
Corrupt collaboration
Cooperation is essential for completing tasks individuals cannot accomplish
alone. While the benefits of cooperation are clear, little is known about its
possible negative aspects. Introducing a novel sequential dyadic die-rolling
paradigm, we show that collaborative settings provide fertile ground for the
emergence of corruption. In the main experimental treatment the outcome
of the two players is perfectly aligned. Player A privately rolls a die, reports
the result to player B, who then privately rolls and reports the result as well.
Both players are paid the value of the reports if, and only if, they are identical
(e.g., if both report 6, each earns €6). Since rolls are truly private, players can
inflate their profit by misreporting the actual outcomes. Indeed, the
proportion of reported doubles was 489% higher than the expected
proportion if participants were honest, 48% higher than when individuals
rolled and reported alone, and 96% higher than when lies only benefited the
other player. Breaking the alignment in payoffs between player A and player B
reduced the extent of brazen lying. Despite the central role of player B in
determining whether a double was reported or not, modifying the incentive
structure of either player A or player B had nearly identical effects on the
frequency of reported doubles. Our results highlight the role of
collaboration—particularly on equal terms—in shaping corruption. Together,
these findings fit a functional perspective on morality, revealing that when
different moral sentiments—to be honest vs. to join forces in
collaboration—clash, people opt for corrupt behavior which serves both their
own and their partner's interests.
10
10.30 - 11.20: parallel session 3: Social Distance & Social Preferences
10.30-10.55
10.55-11.20
Speaker: Geoffrey M. D. Castillo - CeDEx, School of Economics, University of
Nottingham
Speaker: Fadong Chen - University of Konstanz, Germany
Additional Author: Urs Fischbacher
Preference reversals in the social domain: The effect of social distance on
choice and valuation
This paper investigates the influence of social distances on choices and
valuations. We consider a decision-maker who discounts payoff provided to
others as a function of the social distance. This is supported by empirical
work (Jones and Rachlin, 2006) and theories in psychology (Aron and Aron,
1986; Aron et al., 2001; Trope and Liberman, 2010). We predict that, keeping
everything else constant, choices should be affected by variations in social
distances. To test this idea experimentally, we use the preference reversal
paradigm and import two types of social distances with two conditions:
Faculty memberships and Charities. We use tools developed in social
psychology—the Inclusion of the Other in the Self (IOS) scale and the
Inclusion of the Ingroup in the Self (IIS) scale—to control for and measure
social distances. We find that subjects’ propensity to choose an allocation
depends on the social distance between this subject and the recipient of the
allocation. We also find evidence of preference reversals in the social domain
but only when using Faculty members as recipients of the allocations. This
stems from differences in valuations between the two conditions: subjects in
the Faculty condition disregard the social distance dimension and base their
valuation purely on the money dimension. This suggests that preference
reversals occur because of differences in attention drawn to the attributes.
This interpretation is further reinforced by the fact that the IOS and IIS
measures show that social distances were perceived to be less important in
the Faculty condition.
Cognitive processes of prosocial and antisocial punishment
Prosocial punishment is crucial while antisocial punishment is detrimental
for the maintenance of cooperation. To promote cooperation efficiently, it is
important to understand the cognitive mechanism underlying prosocial and
antisocial punishment. This paper investigates this issue by examining
subjects’ decisions and associated response times in a repeated public
goods game and a social value orientation (SVO) task. In order to create
different decision situations for subjects, we vary the cost and the effect of
punishment at the punishment stage of the public goods game. We find that
subjects who have higher SVO are less likely to punish others and they take
longer time to make punishment decisions. The cognitive processes of
decisions in both prosocial and antisocial situations comply with an evidence
accumulation model since the cognitive conflict between different motives
leads to longer response times, the response time increases with the
difficulty of the decisions and the response time is longest when the
frequency of punishment is 50%.
11
11.20 – 12.10: parallel session 3: Social Distance & Social Preferences
11.20-11.45
11.45-12.10
Speaker: Matthias Greiff - Justus-Liebig-University Giessen
Additional Authors: Kurt A. Ackermann and Ryan O. Murphy
Speaker: Maximilian O. Hoyer - Universiteit van Amsterdam
Additional Author: Frans A. A. M. van Winden
How small differences in context affect the measurement of social value
orientation
Individual differences in social value orientation (SVO) can explain prosocial
behavior across a wide range of social situations, such as helping behavior
and charitable giving. However, a „pure“ measure of SVO may be hard to
achieve since the measurement of SVO is affected by the context in which
the measurement occurs. Using a laboratory experiment we analyze the
sensitivity of SVO to different contexts. Using a within-subjects design we
use four different contexts and measure SVO in each: (1) resource allocation
tasks with role certainty, (2) resource allocation tasks with role uncertainty,
(3) decomposed games, and (4) matrix games. Although on the aggregate
level, the distribution of SVO is similar for all four contexts, there are large
differences on the individual level because minor variations in context (role
certainty vs. role uncertainty) and even variations in how information about
payoffs is presented (decomposed vs. matrix game) have non-negligible
effects on subjects’ behavior.
Stand by your man - The role of shared history in investment decisions
We investigate the relationship between an investor and a project manager.
Project managers choose from a pool of projects, the success probabilities of
which are uncertain. Information about the future success probability of a
project is gained by observing its outcome. Investors can change projects,
but also have to change project managers if they want to do so. An
additional joint project or a voluntary transfer precedes their interaction. We
hypothesize that investors favor projects that are managed by project
managers with whom they have shared positive experiences in the past,
which should lead to a lower rate of project change than in comparable
situations with shared negative experiences. The role of this social element is
isolated using a control treatment in which the role of the project manager
does not exist. Interaction through voluntary transfers plays a clear and
significant role in the investors’ decision making, whereas the influence of
merely sharing a positive or negative experience proves more complex than
anticipated.
12
10.30 - 11.20: parallel session 4: Trust & Fairness
10.30-10.55
10.55-11.20
Speaker: Adrian Chadi – IAAEU, University of Trier
Additional Authors: Mario Mechtel and Vanessa Mertins
Speaker: Uyanga Turmunkh - Erasmus University Rotterdam
Additional Authors: Martijn J. van den Assem and Dennie van Dolder
Forced to be generous - Experimental evidence on the behavioral effects of
minimum wages outside the laboratory
More than a hundred students assisted in a university research project by
conducting telephone interviews for half a day and received an expense
allowance of 30 Euro. After getting emails about a potential conflict between
the level of pay and the new minimum wage legislation in Germany, the
project coordination decided to increase expenditures for the second part of
the project, a follow-up online-survey. By earning 15 Euro for the survey
alone, the average hourly wage of each interviewer for the entire project
increased to a level above the minimum wage. To use this situation for an
investigation into the behavioural effects of minimum wages, we randomly
varied the level of information on why the project coordination was willing to
pay such a large sum of money.
Our findings stand in contrast to lab evidence, according to which a
comparatively high wage level loses its positive signal when an employer is
'forced to be generous'. Mentioning the minimum wage as the reason for
paying extraordinary wages did not affect fairness perceptions, nor did it
reduce effort in a vignette task that was part of the online-survey. Most
remarkably, only mentioning that the employer was forced to be generous by
others (without reference to the minimum wage) increased effort and
reduced reservation wages. Instead of interpreting involuntarily wage hikes
as a less positive counterpart to pure gift-giving, it seems that employees
may actually show solidarity when they see that an exogenous force requires
their employer to pay more.
Not all talk is cheap: Communication and cooperation in a high stakes TV
game show
We investigate the credibility of non-binding communication about
cooperative behavior, using data from a game show with high stakes.
Finalists in the British TV show Golden Balls play a variant of the classical
one-shot prisoners’ dilemma, in which they simultaneously choose to either
split or steal an amount that on average exceeds $20,000 (see also van den
Assem, van Dolder & Thaler, Management Science 2012). Prior to their
decisions, contestants are given a brief opportunity to communicate face-toface. We classify all statements made during these conversations in terms of
their malleability to ex-post justification, and hypothesize that people who
defect will prefer to make statements that allow them – ex-post – to deny the
fact that they have been lying. We find that contestants who state their
intention to split in an explicit and unconditional manner (“I will split”) are
more likely to split than those who state their intention only implicitly (“I
came here to split”) or conditionally (“I split if you split”). For implicit
statements we find that these are more informative when the contestant
refers to her character (“It’s not in my nature to steal”) than when she refers
to her preferences or feelings (“I want to choose split”). Seemingly empty
statements that are both implicit and conditional foreshadow stealing. Our
results reject the notion that cheap talk cannot be credible when incentives
do not align, and support the growing evidence that lying is accompanied by
an aversive experience that can have profound implications for the credibility
of cheap talk. What people say and how they say it can predict their
subsequent choice of action.
13
11.20 – 12.10: parallel session 4: Trust & Fairness
11.20-11.45
Speaker: Jeroen M. van Baar - Donders Institute for Brain, Cognition and
Behaviour, Radboud University Nijmegen
Additional Authors: Luke J. Chang and Alan G. Sanfey
Kind beyond belief: Fairness norms predict trust game reciprocity
independently of second-order expectations
A well-functioning society depends heavily on trust and reciprocity. Up to
now, reciprocity has been explained in behavioral economics by use of guilt
aversion models (Battigalli & Dufwenberg, 2007; Chang, Smith, Dufwenberg
& Sanfey, 2011). In these models, guilt is conceptualized as a deviation from
second-order expectations that yields negative utility. Apart from secondorder expectations, however, there might be other norms driving reciprocity
behavior. The role of an agent’s own fairness norms, for example, has not yet
been considered in models of reciprocity decision-making, in part because
fairness norms in the classic trust game are often collinear to second-order
expectations. We present a modified trust game, in which fairness norms
and second-order expectations prescribe different behavior. We find that
both fairness norms and second-order expectations are predictive of
reciprocity behavior, and that these two motivations assume weights that
vary between participants. We propose a novel computational model of trust
game reciprocity that takes into account both second-order expectations and
the trustee’s own fairness norms.
11.45-12.10
Speaker: Yilong Xu - Tilburg University
Additional Author: Jan J. M. Potters
Social Information and selfishness
We study whether and how participants react to social information regarding
the decisions of other participants in a twice-repeated dictator game, both
with and without “earned” property rights of being the dictator. We find that
anticipating allocations to be publicly revealed makes dictators more
generous in the first round allocation, while learning the first round
allocations of others makes dictators become more self-regarding in the
second round. In fact, 32% of the dictators in our experiment decrease their
offers by half or more after learning others’ allocations, whereas this is only
15% without social information. The most self-regarding allocations are
found in the second period after reviewing other dictators’ allocations in case
property right is earned.
14
13.45 - 14.40: parallel session 5: Environment & Sustainable Behavior
13.50-14.15
14.15-14.40
Speaker: Marko Milovanovic - University of Groningen
Additional Authors: Linda E.M. Steg and Russel Spears
Speaker: Eline C. van der Heijden - Tilburg University
Additional Author: Reyer Gerlagh
Big Brother is watching you
In a social dilemma, individual and group interests are at odds. Nowhere is
this tension more evident then in the environmental domain. The collective
interest is evident, yet so are the sacrifices individuals have to make to
ensure it. In this paper, we present two theoretical perspectives that take a
fundamentally different stance in explaining intrinsic motivation in the
context of pro-social behaviour. A traditional perspective, in which group
interests are perceived as extrinsic to the self, and a self-categorisation
perspective, which integrates the concepts of ‘group’ and ‘self’. Out aim is to
contrast these two fundamentally different perspectives and examine their
implications for how best to foster pursuit of the collective good in a
sustainable way. We present evidence from an experimental paradigm in
which we manipulated the context (surveillance vs autonomy) in which
effortful pro-social behavior was performed (producing energy on a rowing
machine). Results show that participants invested less effort in the pro-social
task after surveillance was removed, but for those that identified highly with
the task initiators, this loss in performance was significantly weakened. This
suggests that self-categorisation with social groups can counteract the
negative motivational effects of extrinsic reinforcement. Furthermore, we
found that high identifiers reported higher levels of intrinsic motivation after
the task, even if they were previously exposed to surveillance. These findings
show that social influence can be processed as intrinsic motivation, and
suggest that collective interest can be processed as self-interest, provided
self-categorisation with the group.
Going green: Framing effects in a dynamic coordination game experiment
We design a dynamic stag-hunt game, which captures some features of a
transition from an economy using dirty technologies to a green economy.
Groups consisting of three subjects can transit from a benchmark stable
equilibrium to another stable equilibrium with higher payoffs, but the
transition is slow and costly. Coordination is required to implement the
transition while minimizing costs. In the experiment, games are repeated five
times, which enables groups to learn to coordinate over time. We compare a
neutral language treatment with a ‘green framing’ treatment, in which
meaningful context is added to the instructions. We find the green framing
to significantly increase the number of profitable transitions, but also to
inhibit the learning from past experiences, and thus it reduces coherence of
strategies. Consequently, payoffs in both treatments are similar even though
the green framing results in twice as many green technologies. The
experimental results suggest general support for ‘going green’, but we also
find evidence for anchoring of beliefs by green framing; proponents and
opponents stick to their initial strategies.
15
14.40 – 15.05: parallel session 5: Environment & Sustainable Behavior
14.40-15.05
Speaker: Arianne J. van der Wal - VU University
Additional Authors: Amir Grinstein and Femke van Horen
Temporal myopia in sustainable behavior when facing uncertainty
Environmental concerns increase rapidly due to the growing world
population, consumption and globalization. The need to enhance
sustainable behavior is therefore of paramount importance. However, we
argue that sustainable behavior is particularly negatively influenced by
uncertainty. Building upon life history theory, the rationale is that uncertainty
leads consumers to focus on immediate outcomes more than future
outcomes. Due to the fact that sustainable behavior is strongly associated
with future benefits, we propose that matching the temporal framing of
sustainable behavior with the priorities of consumers during uncertainty—
emphasizing the immediate benefits instead of future benefits—should
buffer against the negative effect of uncertainty on sustainable behavior. In
three experimental studies we provided evidence for our line of reasoning.
The first experiment showed that uncertainty leads indeed to less sustainable
behavior compared to certainty. The second experiment found that people’s
dispositional temporal discounting moderated this effect, showing that
people who display lower levels of dispositional temporal discounting were
less affected by the negative effect of uncertainty on sustainable behavior.
The third experiment revealed that emphasizing the immediate benefits of
sustainable products during uncertainty could buffer against the negative
effect of uncertainty on sustainable behavior. Hence, business and
government that wish to promote sustainable (consumer) behavior need to
emphasize the immediate benefits of sustainable behavior during uncertain
times.
16
13.45 - 14.40: parallel session 6: Bias, Redistribution & Scarcity
13.50-14.15
14.15-14.40
Speaker: Joshua B. Miller – Bocconi University
Speaker: Nickolas Gagnon - Tilburg University
Surprised by the gambler’s and hot hand fallacies? A truth in the law of small
numbers
We discover a subtle but substantial bias in a measure of serial dependence
in sequential data. This measure has been used prominently in the literature
investigating biased beliefs in sequential data. When correcting for the bias
in an analysis of data from prominent studies in the literature, the results are
reversed. The bias also provides a learning-based foundation for the law of
small numbers, and has implications for the study of the perception of
randomness.
With who you work can matter: An experiment on social distance and
redistribution preferences in small groups
What if preferences for income redistribution depend on to which income
group one is socially closer in their day-to-day life? This paper investigates
this possibility in small groups with a game where groups of four have to
complete a task for a wage and then vote on an earnings redistribution rate.
This is done in a classroom experiment. It looks if completing a real-effort
task for a wage with minimal cooperation with somebody with a different
wage changes one's preferences for redistribution compared to completing
the same task with somebody who has the same wage. It finds no effect on
low wage participants, but high wage participants support significantly more
earnings redistribution when they work with a participant with a low wage
than when they work with a participant with a high wage. The effect is
especially strong in preventing high wage participants to vote for no
redistribution at all.
17
14.40 – 15.05: parallel session 6: Bias, Redistribution & Scarcity
14.40-15.05
Speaker: Jeroen Nieboer - London School of Economics and Political Science
Additional Author: Nikhil Masters
Popularity breeds contempt: Effects of popularity feedback in a speed-dating
experiment
An important part of the process of mate selection is the allocation of
resources, such as time and attention, to the pursuit of potential partners. In
many settings however, people do not have access to information that could
facilitate this process. We investigate one such setting, speed-dating, in
which the resource allocation question is summed up in a simple yes/no
decision about each potential partner. We experimentally manipulate the
conditions of an actual speed-dating event by providing participants
feedback on the popularity of potential partners. The results of our withinsubject comparisons show that popularity feedback motivates a substantial
number of the participants to revise some of their “yes” decisions to “no”.
Furthermore, the likelihood that a potential partner is dropped is related to
their popularity: more popular people are more likely to be dropped.
18
13.45 - 14.40: parallel session 7: Reference Points & Numbers
13.50-14.15
Speaker: Alessandro Saia - University of Bologna and OECD
Wishing for bad weather: Demand shocks and labor supply of Bologna pizza
vendors
Recent influential papers analyse labor supply behavior of New York taxi
drivers (Camerer et al., 1997; Farber, 2008; and Crawford and Meng, 2011)
and provide evidence that is in sharp contrast with the prediction of the
standard model of labor supply but is consistent with alternative theories
that explain labor supply decisions with Reference-Dependence Preferences
(Koszegi and Rabin, 2006). Models of Reference-Dependence Preferences
predict that labor supply depends not only on earnings opportunities, as in
standard neoclassical theory, but also on a target (reference point). In the
spirit of previous works I study the labor supply behavior of Bologna Pizza
Delivery Vendors, who, like taxi drivers, face transitory variations in their
earnings opportunities and they are free to set their effort by accepting (or
not accepting) orders. Differently than previous papers, however, I am able
to use weather shocks as instrument for earnings opportunities. By
exploiting variation in demand due to adverse weather conditions, I find that
the target component plays no role in vendors' labor supply decisions. When
truly exogenous demand shifters are used, the behavior of pizza vendors is
perfectly consistent with the predictions of the standard model of labor
supply. Moreover, I show that using the identification strategies of previous
papers (in particular, Crawford and Meng, 2011) pizza vendors would look
like taxi drivers in terms of labor supply behavior. This demonstrates that
results in the previous literature are driven by the (wrong) way in which
demand shocks are identified.
14.15-14.40
Speaker: Adriaan R. Soetevent - University of Groningen
Additional Author: Gert-Jan Romensen
P(l)aying at the pump: Behavioral biases in buying fuel
Do car drivers who fill their tank use round amounts or a round number of
liters as reference points and what drives this behavior? We answer these
questions by the on-site recording of about 2,000 transactions by reading the
pump displays and linking these with data on car and driver characteristics.
Targeting a specific sum of money is three times more popular, with 10.8%
of drivers paying exactly €XX.00 and 25.4% paying an amount with the
decimals between 98 and 02. Males and younger drivers engage in this
behavior more often. In line with the idea that round amounts are convenient
when paying cash, our data show more bunching at stations with the
opportunity to pay cash (39.3%). This however does not explain why 24.6%
of all drivers buy a round value of fuel at stations where only debit card
payments are possible. We investigate the importance of budgets constraints
by distinguishing between private and corporate drivers and by considering
transactions before and after pay day. We also look at the explanatory power
of other possible reasons to target round amounts, such as a savings and
fun motive.
19
14.40 – 15.05: parallel session 7: Reference Points & Numbers
14.40-15.05
Speaker: Tong Wang - Erasmus School of Economics, Erasmus University
Rotterdam
Additional Authors: R. J. D. Potter van Loon, M. J. van den Assem, and D.
van Dolder
Number preferences
We explore people’s preferences for numbers in two types of lottery games.
For the first game, we analyse 5 million choice combinations of 6 different
numbers ranging from 1 to 45 for 175 Lotto draws over a two-and-a-half year
period. For the second game, we analyse over 500 thousand choice
combinations of 4 numbers ranging from 0 to 36 for two lottery events
organized by a casino company. In the two games we observe similar
preferences for specific numbers and for combinations of numbers. Players
show a tendency to pick numbers that have a specific meaning to them, such
as their age and the numbers appearing in their date of birth. They tend to
combine numbers that form an arithmetic sequence. The visual presentation
of numbers also matters: numbers in the center are more popular and there
is a tendency to select combinations that form a visual pattern. Lotto players
respond to prior draws by avoiding winning numbers from the latest draw
but chasing numbers that occurred frequently in a longer series of recent
draws. For the casino lottery, we observe that numbers are more likely to be
selected when they appear in the voucher code that is required for
participation. Our findings support the growing literature showing that
simple labels can influence choice and that the presentation of choice
options matters.
20
13.45 - 14.40: parallel session 8: Status & Prevention Behavior
13.50-14.15
14.15-14.40
Speaker: Samuel A. Franssens - KU Leuven
Additional Author: Siegfried Dewitte
Speaker: Patty C. P. Jansen - TU/e
Additional Authors: Chris C. P. Snijders and Martijn C. Willemsen
Dishonest conspicuous consumption leads to social costs imposed by social
equals
Conspicuous consumption yields social benefits because it is perceived as a
signal of status. People may therefore be tempted to signal more status than
they truly have by conspicuously consuming. We hypothesize that people will
assign less status to individuals who exaggerate their status by
conspicuously consuming dishonestly. This hypothesis is based on the
following reasoning. Status is a zero-sum game: Whenever one person gains
status, at least one other person loses status. Whenever one person
dishonestly gains status, other people may then feel that their status is
threatened and may react by bringing down the dishonest signaller. In other
words, we posit that disregard for dishonest conspicuous consumption is a
reaction with the goal of safeguarding one’s own status. Now, status threat
is higher for observers from the same social level as the signaller than for
observers from a different social level than the signaller, because the former
risk being surpassed on the social ladder by the signaller whereas the latter
do not. If disregard for dishonest conspicuous consumption is indeed driven
by status threat, it should then be strongest when observers are at the same
social level as the dishonest signaller. Three scenario-based experiments find
evidence for the hypothesis that dishonest conspicuous consumption leads
to low regard. Correlational (Experiment 1 and 2) and experimental results
(Experiment 3) show that regard for dishonest conspicuous consumption is
lower for people with the same social status as the dishonest conspicuous
consumer than for people with a different social status.
Measuring prevention performance through behavorial self-reports: An
application of the one-dimensional Rasch model
The few studies in the risk prevention domain that have investigated risk
behaviors across subdomains (e.g. fire risk and transport accidents) typically
used multiple factors and did not consider the possibility that a single
general disposition might be underlying the multitude of prevention
behaviors. There are however good reasons to believe that a onedimensional prevention scale could be developed, as similar onedimensional scales have proven to be successful in other domains with a
similarly large set of subdomains and behaviors (Byrka & Kaiser, 2012;
Kaiser, 1998). These studies used the Rasch model to map a range of
behavioral self-reports on a one-dimensional scale, by ordering the behaviors
on the basis of their “difficulty level”. To test the one-dimensionality of the
prevention behaviors, we used existing survey data (n=3,700) of an insurance
company that consisted of 48 prevention behavioral self-reports (e.g. “do you
have a smoke detector?”) across several subdomains. The results confirm
that prevention behaviors across subdomains can be mapped on a onedimensional scale. Furthermore, the Rasch model allowed us to construct a
smaller scale with 27 items with similar reliability and validity. We conclude
that prevention behaviors, across subdomains, can be mapped on a onedimensional scale when ordering the behaviors according to their difficulty.
This indicates that people are consistent regarding their prevention behavior
across subdomains, as long as the behaviors have similar difficulty levels. An
important implication is that to measure someone’s general disposition
towards risk prevention not all items of all subdomains have to be included.
21
14.40 – 15.05: parallel session 8: Status & Prevention Behavior
14.40-15.05
Speaker: Wojtek Przepiorkava – Utrecht University
The production of a reputation premium: Bargain-hunting and herding in
eBay auctions
Reputation matters in online auction markets. Sellers with more positive and
fewer negative customer ratings receive higher bids and attain higher sales.
It has been shown theoretically and empirically that buyers produce this
reputation premium by trading off prices against the uncertainty they face
with regard to sellers’ trustworthiness. However, finding a good bargain at a
low risk of being cheated (or otherwise dissatisfied) is time consuming and
therefore costly. Buyers may thus resort to “cheaper” strategies and simply
join auctions which have received a bid already. Although such herding will
necessarily lead to higher bidding competition, it may be rational if the
presence of other bidders is a sign of good item quality and seller reputation.
Based on a large sample of eBay auctions of a homogenous good (N ≈
91,000), we investigate whether buyers’ herd behaviour could undermine
eBay’s reputation mechanism. First, we find evidence for herding. The lower
the initial price of an item and the earlier an item receives its first bid, the
more bidders bid on that item driving up the item’s final price. Second, the
herding we observe might be rational because sellers with a better reputation
set lower initial prices and receive their first bids earlier. Finally, even after
controlling for herding in our statistical analyses, we find a significant and
substantial effect of a seller’s reputation on an item’s final price.
22
15.25 - 16.15: parallel session 9: Pro-social Behavior
15.25-15.50
15.50-16.15
Speaker: Rene Bekkers - VU University Amsterdam
Additional Authors: Mark Ottoni-Wilhelm and Dave Verkaik
Speaker: David H. Reiley - Pandora Media, Inc., and UC Berkeley
Additional Authors: Jessan Hutchison-Quillian and Anya Samek
Altruism, warm glow, and generosity: A national experiment
In a symbiosis between work from economics and social psychology, a field
experiment using a nationally representative sample is conducted to answer
one of the most fundamental questions in the science of philanthropy: To
what extent does altruism (versus warm glow) influence donations to
charity? Using real money and a specific charity, we study to what extent
different levels of ‘giving by others’ influences private giving (crowding out),
to what extent inducing empathy and the moral principle of care increase
giving and whether these inductions affect altruism and warm glow
motivations for giving and hence the level of crowding out. In the experiment
we study first to what extent giving by others influences private giving
(crowding-out), and second to what extent the induction of empathy and the
moral principle of care increase giving. Our third question is to what extent
the induction of empathy and the moral principle of care change the
motivation for giving and hence the level of crowding-out. To the best of our
knowledge, our experiment is the first test of empathic and moral motivation
for altruism as a motive for philanthropic behavior in a large population.
Hassle costs and charitable giving: A field experiment with google employees
Hassle costs play an important role in consumer decision making, from
redemption of rebates and coupons to enrollment in retirement savings or
poverty-aid programs. We propose that the presence of hassle costs also
plays an important role in decisions to donate to charity, and we test our
hypothesis with field experiments in a workplace giving campaign at Google.
While the control group continued to have the ability to donate by visiting
the charity’s website and entering their credit card information, the treatment
group was instead given the option to donate via payroll deduction with the
click of a button, thereby decreasing the ‘hassle cost’ of giving. We found
that reducing hassle costs had a large and significant effect on the decision
to donate to the charity, doubling the amount of money raised in each of two
separate randomized experiments.
23
16.15 – 16.40: parallel session 9: Pro-social Behavior
16.15-16.40
Speaker: Paul M. A. Smeets - Maastricht University
Additional Author: Arno Riedl
Why do investors hold socially responsible mutual funds?
We investigate why investors hold socially responsible (SRI) mutual funds.
We use administrative investor data and link them to behavior in controlled
experiments and survey responses. We find that intrinsic social preferences
play an important role in the decision of investors to hold socially
responsible mutual funds. Investors who are more generous in an
incentivized experiment are more likely to hold an SRI fund and invest a
larger fraction of their portfolio in SRI funds. Our results also suggest that
selfish investors hold SRI funds for social signaling reasons, as they more
often talk about their investments to others. Moreover, we find that SRI is a
complement rather than a substitute for donations to charity. Socially
responsible investors donate a larger sum of money than conventional
investors.
24
15.25 - 16.15: parallel session 10: Saving & Borrowing
15.25-15.50
Speaker: Job M. T. Krijnen - Tilburg University
Additional Authors: Marcel Zeelenberg and Seger M. Breugelmans
An importance heuristic in decision deferral
We examined and found in a series of experiments that people defer
important decisions more than unimportant decisions, also in situations
where deferral is sub-optimal, has potential disadvantages, and even in
situations where deferral has no material benefits and is financially costly. In
a second series of experiments, we examined and found that the deferral of
important decisions occurs regardless of whether the choice set consists of a
single alternative, or of two conflicting alternatives. Decision importance
does not affect the likelihood of deferral when one alternative is clearly
dominant. We propose that people use an importance heuristic when
deciding to defer or not: the time and effort that should be invested in
making a decision is inferred from the importance of the decision, hence
important decisions take longer. We discuss the implications of these
findings for policy in the domain of retirement saving, where deferral is
financially costly.
15.50-16.15
Speaker: Benjamin M. Marx - Dept. of Economics, U. of Illinois at UrbanaChampaign
Additional Author: Lesley J. Turner
Default bias in borrowing: Evidence from a field experiment on federal
student loans
Nearly half of all post-secondary students in the U.S. attend two-year public
institutions (community colleges). All students enrolled in participating
higher education institutions are eligible for federal loans, and the majority
of community college students are also eligible for subsidized loans that do
not accrue interest while the borrower is in school. While eligibility for
subsidized loans is determined at the federal level, colleges are allowed to
offer students different loan amounts, including $0. Approximately half of all
community colleges default students into the maximum subsidized loan
(with unsubsidized loans added in some cases) while the remainder provide
their students with $0 loan offers. In previous work, we show that a $0 loan
default offer creates a fixed cost of borrowing, which leads to substantially
lower loan take-up and potentially reduces educational attainment. For this
study, we conduct a field experiment on student loan default options at two
large community colleges. Within each college, we randomly assign students
to a default loan amount of either zero or a positive amount that includes the
student’s maximum subsidized loan. Using the actual loan amounts chosen
by these students, we examine the nature of default bias. We quantify the
excess share of students who will accept a default, allowing this to depend
on whether the default includes a loan, and we compare the distributions of
non-defaulted amounts to describe borrowing preferences. We examine
heterogeneity by student resources, dependency status, year of school, and
past experience of borrowing or attending the same school. Finally, we relate
our estimates to quasi-experimental estimates of loan effects on educational
attainment to assess the opt-out cost required to explain students’ default
bias.
25
16.15 – 16.40: parallel session 10: Saving & Borrowing
16.15-16.40
Speaker: Anett John - CREST-ENSAE Paris
When commitment fails - Evidence from a regular saver product in the
Philippines (Cancelled)
26
15.25 - 16.15: parallel session 11: Regret
15.25-15.50
Speaker: Andreas Nicklisch - University of Hamburg, Germany
Additional Authors: Sonja Köke and Andreas Lange
Adversity is a school of wisdom: Experimental evidence on cooperative
protection against stochastic losses
We investigate the dynamics of voluntary cooperation to either reduce the
size or the probability of a stochastic shock. For variants of a repeated fourperson prisoner's dilemma game, we show that cooperation is larger and
more stable when it affects the probability rather than the size of damages.
We provide crucial insights on behavioral adaptation following adverse
events: defecting players are more likely to switch to cooperation after
experiencing an adverse event, while existing cooperation is reinforced when
the damage does not occur. This behavior is consistent with simple learning
dynamics based on ex post evaluations of the chosen strategy.
15.50-16.15
Speaker: Philippe P. F. M. van de Calseyde - Eindhoven University of
Technology
Additional Authors: Marcel Zeelenberg and Ellen Evers
How doubt affects the experience of regret
Decisions are often accompanied by considerable levels of doubt and regret
yet little is known about how these experiences are related. In a series of
experiments, using a variety of different methodologies, we consistently find
that (1) doubts arising after a decision (i.e., when people start questioning
whether they made the right choice) intensify the experience of regret via (2)
increased feelings of self-blame for having made a poor choice when
realizing that the outcome could have been better. We further find that this
effect is moderated by the severity of the outcome: Severe outcomes are
regretted irrespective of whether people doubted their decision first. Doubts
arising while reaching a decision seem to be unrelated to the experience of
regret. Implications for how decision-related doubts affect emotional
reactions to outcomes are discussed.
27
16.15 – 16.40: parallel session 11: Regret
16.15-16.40
Speaker: Koen van der Swaluw
Additional Authors: Mattijs Lambooij, Jolanda J.P. Mathijssen, Henriette M.
Prast, and Johan Polder
The Weight Loss Lottery as a commitment device. Who will play, and what
will they feel?
There is evidence that points towards an effect of anticipated regret as a
psychological commitment device on the success of losing weight, when
combined with a lottery. The lottery is designed such that participants can
win but cannot claim their prize when they do not meet their weight-loss
goals. This way, participants know they would have received their prize, if
only they had achieved their goals. The preceding is stated to result in regret.
It remains unclear however, whether regret is in fact a primary anticipated
emotion, motivating procrastinating people in case of a lottery. The current
research aims to determine whether the effectiveness of a lottery-based
weight-loss intervention can be best explained through regret. Using a
questionnaire, twelve groups were presented a different scenario in order to
assess the rewarding-conditions (fixed vs. lottery, prize level, deadline
distance) under which regret is felt. Additionally, personality traits and
background variables were used to predict the willingness to participate in
the intervention. Overweight participants who have plans to lose weight are
more willing to participate in the lottery than overweight participants who do
not have these plans. This effect does not exist in case of a fixed incentive,
indicating that the lottery is especially attractive for commitment rather than
financial gains. Disappointment was the most dominant emotion, regret the
second. The lottery elicited more regret than the fixed reward did, and this
effect reversed for disappointment. We found preliminary evidence that
regret is indeed part of the mechanism that explains why a weight-loss lottery
may help people to overcome procrastination.
28
15.25 - 16.15: parallel session 12: Cheating & Fairness
15.25-15.50
15.50-16.15
Speaker: Celia Blanco - Royal Holloway University of London
Additional Authors: Lara Ezquerra and Ismael Rodriguez
Speaker: Jan T. R. Stoop - Erasmus University Rotterdam
Additional Author: Jan J. M. Potters
License to cheat under loss aversion
We present an experiment designed to measure the effect of three different
incentive schemes on cheating behaviour. Participants receive a dice, which
they roll privately. In the baseline treatment subjects receive a fix amount for
rolling the dice. We can assume that there is no incentive to lie in this
treatment. In the second treatment, they receive a payoff that will depend on
the reported number. In the third treatment, subjects receive an initial
endowment before rolling the dice. The outcome of the die will determine if
subjects lose part of their endowment or keep the whole initial capital that
they we allotted. Since their payoff in the second and third treatment depend
on the reported roll of the die, the subjects have an incentive to be dishonest
and report higher numbers to get a higher payoff. This design has two
advantages. First, cheating cannot be detected on the individual level.
Second, the underlying true distribution of the outcome under full honesty is
known, and hence it is possible to test different theoretical predictions.
Theoretically, Prospect Theory predicts lying behaviour to be higher on a loss
domain; however, our findings show that the individuals do lie less when
they face a loss possibility. These results are consistent with the Reciprocity
Theory, meaning that reciprocity can offset the effect of Prospect Theory's
loss aversion.
Do cheaters in the lab also cheat in the field?
A recent literature in economics looks at various aspects of cheating
behavior in the laboratory. In this paper, we are interested in the correlation
between cheating in the lab and cheating in the field. Do participants who
cheat in the lab also cheat in the field? We conduct a laboratory experiment
using a variant of the Mind Game (Jiang 2013). Subjects have to take a color
in their mind (black or white) before they turn over a card. Depending on the
color they report to have taken in their mind they earn a smaller or a larger
payoff. Subjects play the game 20 times. Payoffs above a certain threshold
are indicative of cheating behavior (significant foresight). One day after the
lab experiment, subjects are paid their earnings by bank transfer. A fraction
of the subjects is deliberately paid more than their earnings. We send
subjects a reminder e-mail stating their earnings and ask them to reply if
they have received their payment. We examine whether subjects who fail to
report that they were paid too much, are also more likely to be (suspected)
cheaters in the laboratory mind game. By exploring this correlation our
results speak to the external validity of cheating behavior in laboratory
experiments.
29
16.15 – 16.40: parallel session 12: Cheating & Fairness
16.15-16.40
Speaker: Sara Rezaei – Utrecht University
Additional Author: Stephanie Rosenkranz
Towards a theory of marginal fairness concerns
This paper develops a new model of inequity aversion by assuming that first,
players are not necessarily fully rational and their decisions do not
necessarily reflects the sub-game perfect equilibrium of a game and second,
they are marginally fairness concerned. Therefore, to have a better
understanding of individual's cooperative behavior, instead of looking at
sub-game perfect equilibrium, all the Nash equilibria of a game will be
studied. In addition, we scrutinize the effect of interaction structure on
individual's fairness concerns on various bargaining and cooperative games.
Our findings show that subjects deviate from sub-game perfect equilibrium
quiet often and even some times SPE fails to explain individual's behavior.
Therefore, Nash equilibria of a game is more effective in explaining
individual's behavior. About the degree of fairness we conclude that people
are marginally fairness concerned. Adding other regarding preferences helps
studying individual's cooperative behavior but these preferences are weak.
30
Poster presentations – Abstracts
1. Carin van der Cruijsen - De Nederlandsche Bank
In love with the debit card but still married to cash
Additional Authors: Lola Hernandez and Nicole Jonker
Using shopping diary survey data we show that changing payment patterns is a
challenging task; even when consumers have fallen in love with the debit card,
they find it hard to divorce from cash. While seven out of ten Dutch consumers
report to prefer using the debit card, only seven out of twenty actually mostly pay
by debit card. The likelihood that reported preferences and actual behaviour do
not match increases with income, education and age. Consumers with payments
in cash-intensive sectors, where the wide acceptance of the debit card is a
relatively recent phenomenon, are more likely to overestimate debit card usage
than other consumers. The likelihood of a gap also increases with the amount of
cash that consumers carry with them and decreases with the average transaction
size. Our findings indicate that persistent habits are an important explanation why
the substitution of cash by debit cards took place at a slower pace than was
expected.
becoming unable to continue in a profession due to an illness or accident. The
goal was to raise young adults’ awareness for such low-probability events. Even if
the prevalence is low (e.g., in Switzerland 3% of disabled persons are aged
between 20 and 30 years), the consequences in case of occurrence are severe, with
financial gaps of up to 40% of income compared to before the negative event took
place. Together with an insurance company, we examined the effectiveness of
“informational nudging” (i.e., providing information which acts as a nudge) in
sensitizing young adults to the potential risk of disability and their perceptions of
disability insurance. In a pre-study, we assessed the main psychological barriers
preventing young Swiss adults from participating in private provision (n = 1003),
to ensure a targeted approach for the intervention. In the main study (n = 240), we
tested in an online experiment the effect of four informational nudges on young
adults’ risk awareness and insurance decisions. We found that presenting
information on a company’s website in such a way, that heuristics such as
availability are exploited, enhanced risk awareness and increased insurance
preferences partially.
4. Petr Houdek - J. E. Purkyne University
2. Takehisa Kumakawa - Osaka University of Economics
An experimental comparison between self- and third-party evaluations
How to pick up the true meaning of messages exchanged in the laboratory is an
important issue for experimental research. The present study investigates, by
experimentally comparing self- and third-party evaluations, to what extent selfevaluations by message receivers can be relied on. After a standard public-good
game, subjects receive from their counterparts a free-form written message
evaluating their decision and self-evaluate its content. Third-party evaluators also
evaluate the content independently. A comparison between both evaluations
showed that a significant proportion of them agreed. Firm evidence of a selfserving bias could not be found.
3. Linda Miesler - Zurich University of Applied Sciences
You only live once? Informational nudges as an effective approach to raise young
adults’ awareness for the future risk of disability
The decision whether or not to obtain private insurance provision is a complex
process, for which many consumers do not have sufficient time or the cognitive
and motivational resources. In the present work, we focused on the risk of
Do people in post-communist countries cheat so much? Laboratory study using a
student sample
Additional Authors: Julie Novakova and Jan Jolic
In our study, we focused on the prevalence of cheating and individual
characteristics, which influence dishonest behavior. We used a dice rolling task
where participants can cheat on their reward to determine whether anonymity
conditions increase the frequency of cheating. Our results show that participants
playing alone did not cheat significantly more than the control group throwing
dice publicly. Although cheating in both groups occurred systematically, the rate
of cheating was low, especially in comparison with a previous study in postcommunist country (Ariely et al. 2015). We did not find that gender, cognitive
ability (measured by the Cognitive Refection Task) or social preferences
(measured by the Dictator and Ultimatum Games) robustly predicted rate of
cheating in our sample.
5. Marjolein Harmsen-van Hout - RWTH Aachen University
Effort and accuracy in social preferences
31
Social preferences, meaning that people take the consequences of their choices
for others into consideration, are often observed in laboratory experiments. For
example, in dictator games participants tend to voluntarily share a considerable
part of an endowment with a passive co-participant. However, this behavior may
not always correctly reflect real-world preferences. If the causes for such
inconsistency are identified, these can be taken into account to improve both the
validity of further experiments and the efficacy of real-world applications.
In this experiment, I investigate a new potential source of inconsistency:
the larger complexity of real-world settings, by which it may not be obvious what
the consequences of individuals’ choices are for themselves and for other
individuals. For instance, in our networked society an individual’s choice may
affect many other individuals in many complex ways. This may lead to a tradeoff
between accuracy in making the right choices and saving of effort in finding out
the exact choice consequences by using simpler decision rules. My hypothesis is
that the willingness to sacrifice accuracy in order to save effort is significantly
larger when it concerns others' payoffs than when it concerns own payoffs. Thus,
in more complex situations, people tend to focus on the consequences of their
choices for themselves.
This hypothesis is formalized and tested by executing an adapted
dictator game in the laboratory with four payoff table variations to manipulate
complexity. I present the connections of this setup with the existing literature and
the results of a small sample pretest.
6. Alko van der Wiel - KU Leuven
The self-defining impact of (dis)advantageous flextime allocations as a function of
scarcity
Additional Authors: Lieven Brebels, Sophie de Winne, and Luc Sels
The group engagement model (Blader & Tyler, 2009) proposes that economic
outcomes help shape social identity and subsequent behaviors that contribute to
effective organizational functioning. However, prior research drawing on the
model have largely neglected macro-economic factors such as an organization’s
economic situation or resource scarcity. Such factors could prove vital moderators
in shaping and maintaining an employee’s identity, and could determine the
success of workforce differentiation as an HRM strategy (‘strategic HR
differentiation’; Huselid & Becker, 2011). Furthermore, differentiation goes
beyond the allocation of just tangible outcomes and also examines the
distribution of more socio-emotional outcomes (e.g. flexible working conditions).
Since time and flexibility are increasingly identified as valued resources for
organizational members (e.g. Valcour, 2007), an interesting question is how the
impact of strategic differentiation in flextime allocation on employees’ social
identity is affected by the amount of flextime available (i.e., scarcity). Based on the
group engagement model and the scarcity principle (with scarcity increasing the
value of those resources), we predict that the identity-relevant implications (e.g.,
pride, respect, and identification) of flextime differentiation are amplified under
high scarcity conditions. In an experimental study, we manipulated scarcity of
valued flexible work hours and the (dis)advantageousness of the differentiation
outcome relative to previously comparable others. This design allows us to
examine the predictions of the group engagement model within a strategic HRM
context, and to explore the moderating role of scarcity as a macro-economic
factor.
7. Rafael Batista - University of Warwick
When saving, who needs targets anyway?
Additional Authors: Joe Gladstone, Evgenii Nikitin, and Nikhil Ravichandar
Setting targets is commonly considered a good strategy when attempting to save.
However, it may not be necessary for all people. In fact, it may be a hindering
practice for those with high levels of self-control.
Using data collected from a representative UK sample (N= 2963) we analyzed the
relationship between savings behavior and subjective measures of self-control,
financial literacy and personality. The results from the linear regression model
showed that setting savings targets is important for those with low to medium
self-control, but is counter productive for those who fall in the top quartile on the
self-control scale (Tangney, et al, 2004). These results remained consistent when
self-control was determined using an intertemporal discount rate.
We also checked whether the effect of self-control is mediated by
financial literacy, finding a significant three-way interaction between self-control,
financial literacy, and effectiveness of target setting.
Setting targets is still an important component for saving for those with low levels
of self-control and financial literacy. Nevertheless, those with high levels of selfcontrol could be saving more by foregoing setting targets.
We also obtained data on types of target used (proportion of income vs. fixed
target) and awareness of personal savings. The relationship to savings behavior
and self-control is still to be analyzed.
Previous work highlights the use of commitment devices and targets;
our results suggest these may be an unnecessary cost to those with high levels of
self-control. These findings add to the literature that looks at the psychology of
financial decision-making.
32
8. Jan Hausfeld - University of Konstanz
The consquences of time pressure on risky decisions
Additional Author: Sven Resnjanskij
We use a novel approach to analyze the trade-off between decision quality and
decision time. Since time is a scarce and valuable resource, a tradeoff arises
between improving decisions and saving time for alternative utility generating
activities. In our experiment, time pressure is not induced by a fixed decision time
but arises endogenously by solving the trade-off of between improved decisions
and time-dependent opportunity costs. As expected, subjects are faster in making
their decisions if time is costly. We also find that time pressure induced by higher
opportunity costs leads to more risk seeking choices and generally increases the
heterogeneity in decision making under risk.
9. Alain Starke - Eindhoven University of Technology
Tailoring energy-saving advice using a one-dimensional Rasch scale of
conservation behaviors
Additional Authors: Martijn Willemsen and Chris Snijders
When analyzing ways in which people save energy, conservation researchers often
conceptually differentiate between curtailment (e.g. unplugging chargers) and
efficiency (e.g. installing PV cells) to understand consumer energy-saving behavior
(Gardner & Stern, 2008). Recently, several authors have questioned such a twodimensional perspective on energy conservation and argued for a onedimensional construct (cf. Urban & Ščasný, 2014).
This study considers whether such a single dimension actually exists, by
applying the Rasch model on an extensive set of energy-saving measures. Based
on self-reports of individuals, Rasch assigns execution difficulty levels (i.e.
behavioral costs) to different behaviors, as well as behavioral propensities (i.e.
attitudes) to individuals (Kaiser et al., 2010). Besides the potential improvement
in terms of measurement, such an analysis allows us to investigate whether such
a one-dimensional approach can be useful to provide tailored energy-saving
advice to consumers. For this purpose, we created a Rasch-based energy
recommender system.
In study 1 (N=263), we successfully constructed a unidimensional scale
of 79 energy-saving measures, containing both curtailment and efficiency
measures. Study 2 (N=198) tested which energy-saving measures, in terms of
relative execution difficulty levels, were perceived as the most appropriate.
The results show that our recommender system is capable of inferring
energy-saving attitudes by using a 13-item questionnaire, as well as providing
tailored conservation advice. Consistent with the Rasch model, we also found that
users perceived energy-saving recommendations with a difficulty below their own
attitudinal level as more appropriate than those with a difficulty above it.
33
Notes
34