ArticlePDF Available

Abstract and Figures

Corruption is a welfare issue worldwide, but it is difficult to study because of its secret nature. We here did a lab economic experiment on bribery to study different compliance mechanisms through which people might be deterred from corruption. We focused on two elements of norms which people might respond to: information about the function of the norm (to avoid harm to third parties) and information about the prescriptive content of the norm (rights and duties). We show that information about both negative externalities of bribery and prescriptive norms are effective deterrents, and that bribe offers and acceptances are most discouraged with their synergic interaction. We find that participants followed prescriptive information, even when it was inefficient to do so, and implied choosing against their material self-interest (by rejecting a bribe) and not reciprocating bribe offers. Such compliance regardless of costs to the self and to others suggests a rule-based “mindless” process, like a normative heuristic. We conclude by highlighting the relevance of these findings as behavioral insights in the elaboration of strategies to combat corruption and norm transgressions.
Content may be subject to copyright.
The effect of prescriptive norms and negative externalities on bribery decisions
Hipólito Hasrun, Carlos Maximiliano Senci, Rodrigo Moro, Esteban Freidin
Corruption is a welfare issue worldwide, but it is difficult to study because of its secret
nature. We here did a lab economic experiment on bribery to study different compliance
mechanisms through which people might be deterred from corruption. We focused on two
elements of norms which people might respond to: information about the function of the
norm (to avoid harm to third parties) and information about the prescriptive content of the
norm (rights and duties). We show that information about both negative externalities of
bribery and prescriptive norms are effective deterrents, and that bribe offers and
acceptances are most discouraged with their synergic interaction. We find that participants
followed prescriptive information, even when it was inefficient to do so, and implied
choosing against their material self-interest (by rejecting a bribe) and not reciprocating
bribe offers. Such compliance regardless of costs to the self and to others suggests a rule-
based “mindless” process, like a normative heuristic. We conclude by highlighting the
relevance of these findings as behavioral insights in the elaboration of strategies to combat
corruption and norm transgressions.
We ran a bribery laboratory experiment in which participants made decisions involving real
money with consequences for them and others. The experiment aimed at testing whether
people may refrain from corruption because it simply goes against explicit norms and/or
because it damages others’ welfare (has negative externalities). We varied whether
participants faced a fully-fledged normative environment (which included clearly defined
rights and duties) or a bribery game devoid of explicit normative information. In addition,
counter-normative behavior could infringe externalities upon others or be innocuous.
Whereas the separate presence of norms and externalities each caused a lessening of
bribery, the strongest deterrence occurred when they acted together. This synergy may
serve as a behavioral insight to inform policy.
Corruption is an institutional issue that imposes negative externalities on the welfare
of societies worldwide and distorts the efficient allocation of resources (1-2). Monitoring
and sanctions are obvious deterrents (3-4); however, their efficacy is limited by several
country-wise or regional factors, such as the existence of corrupt monitors, politics with
discretionary power, and a weak judicial system (5-6). Additionally, increases in
transparency and, therefore, accountability may be countervailed by increased efforts to
hide the corrupt deeds (7). Furthermore, options based on informal enforcement present
other issues, such as the second-order free-rider problem –who bares the costs of
punishment- (8), and the fact that diffuse punishment may lead to more (than less) anti-
social behavior (e.g., 9).
Beyond external incentives, converging lines of evidence from cultural
anthropology (10-12) and sociology (13) to social psychology (14-15), behavioral
economics (16-18), and neuroscience (19-21) suggest that people follow norms at levels
that could not be predicted from selfish instrumental reasons alone. Indeed, according to
Norm Focus Theory, directing people´s attention to prescriptive norms (socially approved
behaviors) may increase compliance (14). This prediction has received empirical support in
different domains, such as littering (22), distribution of monetary resources (23-24), and
theft (25), to mention some examples. In line with this, a complementary, easily
implementable, and cheaper strategy to combat corruption may involve the use of
information that makes moral and normative appeals.
In the present study, we focused on two elements of norms which people might
respond to: information about the function of the norm (to avoid negative externalities; 13,
26-27) and information about the content of the norm (an agent´s normative status –e.g.,
whether she has a particular right- and actions prescribed/proscribed as a function of such
status –e.g., provision of a rightful benefit). From a theoretical perspective, the distinction
may entail different compliance mechanisms. On the one hand, people may act as moral
consequentialists, and thus, be motivated to weight the harm on others associated with each
decision option (see e.g., 28-29). According to this, people may avoid norm transgressions
because of being sensitive to the associated negative externalities. On the other hand, there
is plenty of evidence showing that, in contrast to reasoned decisions, people may make
choices based on heuristics or simple rules of thumb (30). Heuristics involve fast and frugal
decisions, and could be contrasted against more effortful reasoned choices (31). Indeed, the
heuristic approach has been extended to social decision making, including domains such as
cooperation (32) and morality (33). From this perspective, people may comply with norms
as if simply following “normative” heuristics or rules, without necessarily taking into
account the expected consequences (34). We believe that practical consequences of
discriminating between these compliance (function- vs. content-based) mechanisms may
stimulate the elaboration of effective behavioral insights that can inform anti-corruption
policies (see 35).
Despite its theoretical and practical relevance, the distinction between sensitivity to
externalities and to detailed prescriptive information remains mostly unexplored in bribery
experiments (see 36 for a review; see also 37-40; but see 41) as well as in experimental
studies of dishonest behavior (42-44). Furthermore, bribery experiments focused on
studying the effect of either externalities or norms on their own have not managed to
succeed in clarifying these matters either. The conditions under which participants show
sensitivity to negative externalities associated with bribery has not been clearly established
in the experimental literature (41, 45). It is important to approach this issue because
normative appeals may be less effective in certain contexts, such as those involving
economic transactions (46). In turn, the use of loaded wording to create a normative frame
in bribery experiments has produced mixed results which did not allow systematic
conclusions either (41, 47-49).
Our Study
The goal of the present study was to assess the effect of detailed prescriptive
information and negative externalities as causes of norm compliance in a collusive bribery
game. In this context, roles, rights, and duties were clearly stated, and negative externalities
were suffered by a passive third party. With this goal in mind, we invited university
students (N=202; 54% women; mean age ±1SD: 22.6 ±4 years old) from a wide range of
academic disciplines to participate in a one-shot collusive bribery game involving monetary
stakes in the PCs lab at IIESS CONICET Bahía Blanca, Argentina. The game was
implemented using Ztree (50).
Participants were randomly assigned to one of two roles (citizen and public official
–hereafter, the latter simply called “official”), and a local conservationist NGO experienced
the inefficient negative externalities of corruption (i.e., a drastic decrease in the
experimenters´ donation; see SI). Across between-subject treatments, we systematically
varied whether there were negative externalities associated with the official´s decision of
favoring the citizen, and whether participants were presented with explicit normative
information. These instructions comprised information about the citizen´s normative status,
that is, whether she acquired the right to receive a monetary benefit to be provided by the
official; and information about the official’s duty, that is, whether she should provide the
benefit given the citizen’s normative status (see SI Appendix I for an English version of the
written instructions). Overall, we had five independent treatments: 1) condition
Right_Externality (n=42); 2) condition NoRight_Externality (n=44); 3) condition
Right_NoExternality (n=38); 4) condition NoRight_NoExternality (n=38); and 5) condition
NoNorm_Externality (n=40).
Figure 1 shows a schematic representation of the three stages involved in the one-
shot bribery game used. In all conditions, the game began with the citizen having 120 sec to
perform a real-effort task which consisted of counting the number of letters “a” in a two-
paragraph text (stage 1). In conditions with normative information (Right and NoRight), we
systematically varied the successful-performance threshold through which the citizen
acquired her normative status (the right to a monetary benefit). In condition Right, the
successful-performance threshold was very low so that almost all citizens acquired the
right, in which case the officials´ explicit duty was to provide the benefit. In turn, in
condition NoRight, the successful-performance threshold was very high so that almost all
citizens did not acquire the right, in which case the officials´ explicit duty was not to
provide the benefit (see SI for more details). In the NoNorm condition, the initial effort task
had no consequences upon subsequent stages of the game (participants knew this) and,
therefore, citizens did not have any normative status and officials no explicit duty. In stage
2, after being informed about their performance and corresponding normative status,
citizens had to make a monetary transfer to their corresponding official (a randomly
associated participant in the room). In the NoNorm condition, there was no information
about any normative status and participants knew that performance information would only
be provided at the end of the game. In the instructions and relevant screens for all
conditions, we specified that the transfer in stage 2 represented an administrative fee of $2
which was the minimal transfer admitted, but that they could transfer more if they wished
(i.e., up to $40 which was the amount of the benefit the citizen could obtain). The transfer
amount came from citizens´ initial endowment of $52. In stage 3, each official was
informed about the normative status of the corresponding citizen (only in conditions with
normative information) and the transfer amount received from her (in all conditions). The
amount transferred was presented as comprising the $2-administrative fee and a“surplus”
(i.e., a bribe, if there was one). If there was no bribe, the official had a simple choice
between providing or not providing the $40-benefit to the citizen (the benefit amount did
not come from the official´s endowment -which was initially $68- as if it came from public
funds). If there was a bribe (i.e., a transfer >2), the official could accept it, which
automatically implied providing the $40-benefit to the citizen (an accepted transfer of $12
was the only case in which citizen and official ended with the same final payoff, i.e., $80),
or reject it and decide whether to provide the benefit or not (the official could not accept the
bribe and not provide the benefit; see Figure 1).
Figure 1. Representation of the three stages involved in the present one-shot collusive
bribery game for conditions Right and NoRight.
Approximately a random half of participants in Right and NoRight conditions were
assigned to the Externality condition (n=86), whereas the other half were assigned to the
NoExternality condition (n=76). In turn, the NoNorm condition (n=40) did involve
externalities with the aim of having a treatment similar to other bribery games in the
literature (41, 48). In all conditions, externalities were implemented as a drastic and
inefficient decrease (from $50 to $5, i.e., a $45-loss compared to the citizen´s $40-gain for
receiving the benefit) in the experimenters´ donation to a local conservationist NGO. For
each pair of participants, a negative externality was caused by the official´s decision in the
game: in conditions with normative information, externalities were caused by the official´s
transgression of her duty, which was to provide the $40-benefit if and only if the citizen
had acquired the right to it (this was informed both in written instructions and relevant
screens for both roles). In the NoNorm condition, externalities were simply caused by the
official´s provision of the benefit to the citizen (this implementation of externalities in the
NoNorm condition was similar to other bribery games in the literature; 41, 45, 48). In
addition, all conditions had the same loaded wording in reference to roles as “citizen” and
“official”, and to the bribe as a “surplus”, which was also similar to loaded frames
implemented in other bribery games (41, 47-48). However, the NoNorm condition had no
information about the citizen´s normative status or the official´s duty, hence served as
baseline control (representative of other loaded bribery games in the literature) against
which to assess whether more specific and detailed normative information could affect
When the game ended, participants completed several post-decision questionnaires,
which also included providing socio-demographic information (see SI Appendix II). Finally,
participants were sequentially called by their PC-terminal number, and each was privately
handed a closed envelop with his or her earnings in cash inside.
Behavioral predictions
In the present study, we attempted to disentangle whether people respond to a norm
because of its function and consequences (i.e., avoiding negative externalities), or because
of its content, namely the mere presence of prescriptive information (rights and duties).
Some predictions derived from these two hypotheses contrast with each other, and are also
in contrast with predictions derived from selfish rationality as we explain below.
Selfish rationality: since the provision of the benefit was costless to the official, and
accepting any bribe automatically implied its provision, the citizen had an incentive to send
the minimal bribe possible (i.e, $1). If the official was selfishly rational, then she would
accept any bribe>0, and thus provide the benefit. This hypothesis did not predict any
difference among conditions.
Moral consequentialism: this hypothesis could take different flavors depending on whether
the citizen and the official were sensitive to causing negative externalities, and/or the
citizen believed the official was sensitive to causing externalities. Through preferences or
beliefs, this hypothesis predicts that the proportion of bribe offers and bribe acceptances (or
more generally, benefit provisions) should be lower in conditions where the provision of
the benefit led to negative externalities (i.e., less bribing in conditions NoRight_Externality
and NoNorm_Externality than in the remaining conditions).
Normative heuristic: Contrary to consequentialist models which emphasize the attunement
of decisions to the consequences of each choice, the heuristic approach assumes that
participants rely on ‘fast and frugal’ rules that guide their behavior in a more or less
unreflective way (30, 32). If this is the case and, therefore, citizens and officials respond to
norms as if they were rules, or citizens expect officials to respond in such a way, then the
proportion of bribe offers and bribe acceptances (or more generally, benefit provisions)
should be lower in conditions where providing the benefit was counter-normative (i.e., less
bribing in conditions NoRight_Externality and NoRight_NoExternality than in the
remaining conditions).
Double criterion: Finally, it is possible that norms and consequences interact in their effect
on people´s decisions (e.g., 41, 51). This could be so because the presence of negative
externalities could be more or less tolerated depending on the norms in place. Indeed,
negative externalities are part of the accepted consequences in certain contexts, such as it is
the case, for example, with sports competitions or other sum-zero situations (51). In the
context of corruption, the probability of refraining from it may increase when the
possibility of causing negative externalities is associated with an explicit relevant norm
against it (41, 52). On the contrary, the absence of explicit norms may allow self-serving
interpretations of what is appropriate in a given context (43). If we apply this double
criterion hypothesis to the present game, we predict that the proportion of bribe offers and
bribe acceptances (or more generally, benefit provisions) should be the lowest in condition
Figure 2 shows the proportion of bribes offered by citizens as a function of
condition. The percentage of bribe offers was lower in condition NoRight_Externality
(45%) than in condition NoRight_NoExternality (80%) (Fisher´s exact test, P < 0.05),
showing an effect of negative externalities on citizens´ decisions. In addition, 85% of
citizens in treatment NoNorm_Externality offered bribes which, against the 45% of bribes
in treatment NoRight_Externality (Fisher´s exact test, P < 0.05), shows citizens´ sensitivity
to normative information. In short, citizens behaved as predicted by the double criterion
hypothesis in the sense that they most strongly refrained from offering bribes when citizens
had not acquired the right to the benefit and obtaining the benefit was associated with
negative externalities.
The reported effects of prescriptive information and externalities on citizens´
decisions could mean that citizens themselves were sensitive to those normative elements
and/or that they expected officials to be sensitivity to them. The information provided by
citizens in the post-decision questionnaires helps disentangling these possibilities. A probit
regression of bribe offers as dependent variable showed a significant predictive effect of
citizens´ rating of the appropriateness of offering a bribe in the game (b=0.25; P=0.09). In
turn, citizens´ estimation of the percentage of officials who would accept a bribe turned out
not to be a significant predictor of bribe offers (b=0.002; P=0.64). This suggests that
variation in citizens´ own moral sensitivity to bribing may have underlied variation in bribe
Figure 2. Proportion of bribe offers as a function of experimental condition.
Also worth noting is that citizens in Right conditions (white bars in Figure 2)
offered bribes in 75% of cases on average, which is not significantly different from the
highest bribery frequency in the experiment (80%; see Figure 2) (Fisher´s exact test, P >
0.1). Bribes in Right conditions could have been motivated by citizens´ concern for
externalities (caused by the official not providing the benefitin these cases), which is
consistent with the slightly, though not significant, higher proportion of bribes in condition
Right_Externality (80%) than in condition Right_NoExternality (69%; Fisher´s exact test, P
> 0.1). Non-exclusive alternatives involve the notion of reciprocity. On one hand, bribery
could be an instance of citizens´ intention to secure their rightful benefit by appealing to the
official´s positive reciprocity after receiving the bribe. On the other hand, if the citizen
thought that the official was expecting a bribe, bribing could be intended not to disappoint
the official´s expectations and avoid triggering his negative reciprocity. The citizen´s
expectations of reciprocity from the official could parallel beliefs triggered in Ultimatum
Games (e.g., 49, 53), or even harassment bribe contexts (49, 54).
In terms of transfer amounts sent by citizens to their corresponding officials, bribes
amounts (transfers>$2) did not systematically vary among conditions (Kruskal Wallis
ANOVA by ranks, 2=2.73, df=4, P= 0.60). Notwithstanding, citizens´ scores in the
Machiavellianism scale (Cronbach´s=0.76) significantly and positively predicted the
amount transferred (=3.84, P=0.016), meaning that more manipulative individuals were
prone to offer larger bribes. Moreover, alike in other bribery experiments (55-56), women
were less likely to offer bribes (b=-1.38, P<0.001), whereas bribe amounts were unrelated
to gender (b=-0.05, P=0.97).
Figure 3 shows the proportion of officials who provided the $40 benefit to the
citizen as a function of condition. Similarly to citizens, officials were affected by both
normative information and the possibility of externalities. The overall majority of officials
granted the benefit when it was deserved (95% in conditions Right_Externality and
Right_NoExternality, together), and denied it when it was not (21% granted benefits in
conditions NoRight_Externality and NoRight_NoExternality, together; Right vs. NoRight,
Fisher´s exact test, P < 0.001). This normative pattern occurred even when officials were
offered bribes: 96% vs. 23% of benefits granted when officials were bribed in Right vs.
NoRight conditions, respectively (Fisher´s exact test, P < 0.001). This meant that officials
rejected bribes in 77% of occasions in NoRight conditions (89% in condition
NoRight_Externality and 71% in condition NoRight_NoExternality; Fisher´s exact test, P
>0.1). Also in support of the effectiveness of normative information, only 11% of officials
provided the benefit after a bribe attempt in condition NoRight_Externality (1 out of 9),
whereas bribed officials provided benefits in 37.5% of occasions (6 out of 16) when there
was no explicit norm but still benefit provisions caused inefficient externalities.
Figure 3. Proportion of officials who provided the $40-benefit to the citizen as a function
of condition.
Despite the overall normative behavior of officials, bribe offers did increase the
probability of benefit provisions (b=0.85, P=0.05; see Figure S1 in SI), even after
controlling for the presence of normative information, externalities, and the acquisition of
the right to the benefit. In addition to their sensitivity to the explicit duty, officials were less
likely to grant the benefit when doing so caused an externality than when it did not (5% vs.
29% of granted benefits in conditions NoRight_Externality vs. NoRight_NoExternality,
respectively; Fisher´s exact test, P=0.05; see the black bars in Figure 3; this result is
similar, though turns not significant, when one considers only benefits provided after bribe
offers: 11% vs. 29% of granted benefits in conditions NoRight_Externality vs.
NoRight_NoExternality, respectively; Fisher´s exact test, P=0.29). All in all, officials
generally behaved according to the normative heuristic hypothesis, providing the benefit
when deserved and refusing its provision when undeserved. Nevertheless, there is also
evidence that officials were also sensitive to negative externalities. As it occurred with
citizens, the strongest anti-corruption result was found when both a negative externality and
an explicit prescriptive norm were in place (i.e., in condition NoRight_Externality; see
Figure 3).
Bribery experiments can have a clear practical goal which is finding ways to fight
corruption (36). In this sense, it seems important to experimentally study the underlying
mechanisms of norm compliance by which both citizens and public officials are deterred
from corrupt exchanges, such as bribery. In the present study, we found evidence
supporting two different norm-compliance mechanisms, which, in turn, had synergies.
On the one hand, we evaluated whether participants´ choices revealed a moral
consequentialist motivation to avoid harm to third parties (29).When there were negative
externalities associated with the provision of the benefit, citizens were less prone to offer
bribes and officials were less likely to provide the benefit (i.e., lower bribery in condition
NoRight_Externality than in condition NoRight_NoExternality; see Figures 2 and 3, and
also see SI for further discussion of citizens´ bribe offers in condition NoNorm).This
overall pattern of benefit provisions across conditions was similar, even if we only consider
officials´ behavior after bribe attempts. In short, the possibility of causing material harm
inhibited bribery in the present protocol. This is consistent with results from Barr and Serra
(41) who found that high (relative to low) negative externalities reduced bribe offers in a
one-shot bribery game with a loaded frame similar to the current condition
NoNorm_Externality. This effect of negative externalities was not present when their game
was framed in abstract terms, which coincides with results fromAbbink et al.(47) who did
not find any effect of negative externalities on participants´ decisions in a collusive bribery
game with repeated-rounds framed in abstract terms. Together, results from these studies
suggest increased sensitivity to externalities in the presence of consistent prescriptive
norms. We further discuss the effect of normative information below.
Alternatively, participants´ consequentialism could be directed at minimizing payoff
differences between citizen and official (i.e., an Inequity Averse motivation, IA hereafter,
e.g., 28). With present payoffs, equality between partners could only result from the
official´s acceptance of a $12-transfer ($10-bribe) and the consequently $40-benefit
provision to the citizen (which led citizen and official, each, to end with $80). According to
this reasoning, IA does not predict differences among conditions. However, if citizens
expected a higher proportion of benefit provisions in Right than in NoRight conditions (i.e.,
an expectation of normative compliance), then IA would predict a higher percentage of
bribe offers in the former than the latter conditions (indeed, citizens showed higher
expectations of bribe acceptances in Right than in NoRight conditions; see SI). This
prediction derived from IA was, nonetheless, only corroborated in conditions with
externalities, whereas there was no statistical difference in the percentage of bribe offers
between conditions Right_NoExternality and NoRight_NoExternality. Moreover, officials´
provision of benefits was drastically higher in condition Right_NoExternality than in
condition NoRight_NoExternality, which would not be anticipated by IA at all. Even if
participants not only strived to reduce inequality between partners but also in relationship
with the NGO (i.e., in conditions with negative externalities), IA still predicts bribery under
sensible values for the “guilt” parameter (see SI). Also important for present purposes, IA
does not predict the inhibitory effect of normative information on bribery as observed when
comparing conditions NoRight_Externality and NoNorm_Externality (which we discuss
On the other hand, we found supporting evidence for a compliance mechanism
based on the mere following of what the prescriptive norm commands. More specifically,
participants in both roles were sensitive to the presence of information about the normative
status of the citizen and the corresponding official´s duty. For citizens, this was evident
only in conditions with externalities, where the percentage of bribes offered was much
lower in condition NoRight relative to conditions Right and NoNorm (see the three bars in
the left-hand side of Figure 2). The effect of normative information was even more
impressive for officials, who tended to follow the norm of not providing the benefit to
citizens who did not acquire the right even when there were no externalities and they were
offered bribes. This result shows officials´ conformity to norms even when it was
inefficient to do so (providing the benefit introduced $40 to the game), and implied
choosing against their material self-interest (by rejecting a bribe),and not reciprocating
bribe offers. Such compliance regardless of costs to the self and to others suggests a rule-
based “mindless” process, like a normative heuristic (32-34; also see SI for arguments
against an interpretation based on an experimenter demand effect).
Normative-heuristic-like responses seem consistent with Norm Focus Theory,
which posits that cognitively activating normative information increases the probability of
norm consistent behaviors (22). Instead, the effect of explicit norms could be related to
conditions of greater accountability. Arguably, introducing an explicit normative demand
may limit the expression of ethical blindness (57) and self-serving biases (43). Whether it is
a matter of mere availability of normative information, as Normative Focus Theory
suggests, or norm compliance requires a more cynical explanation could be tested by
varying whether the normative information provided is public or private, or by changing
people´s feelings of being seen (58).
These two compliance mechanisms, namely weighing consequences and following
normative rules, may indeed present synergies. This can be seen in the fact that the lowest
level of bribe offers and benefit provisions occurred in the condition NoRight_Externality.
In this condition, information about norms and externalities converged to deter bribery. The
fact that sensitivity to externalities was increased with the presence of normative
information shows some parallel with results from Barr and Serra (41). These authors tested
an abstract versus a loaded frame in a one-shot collusive bribery game and found that
bribery was less frequent in the latter than the former condition, though only for
participants in the role of citizens (not for officials) and when externalities were high.
Taking Barr and Serra´s (41) study and ours together, one could see the effect of increasing
the normative information provided, from an abstract game, through a game framed with
words related to bribery, to a game with explicit information about normative status and
duties: more normative information leads to less bribery. Indeed, this normative
interpretation is consistent with the asymmetric effect of normative information as a
function of participants´ roles found in the present experiment. That is, citizens were only
deterred from offering bribes when normative information converged with the existence of
externalities (there was no difference in bribe offers between Right and NoRight conditions
when there were no externalities). For officials, instead, normative information deterred the
provision of underserved benefits even in the absence of externalities. This asymmetry
between roles could be linked to the fact that normative information directly prescribed
actions for officials (“your duty is to…”), whereas the content of the norm was relatively
mute in terms of whether citizens should or should not offer bribes. This suggests that
norms directly prescribing behavior may have a higher chance of affecting decisions. All
together, present results point toward a behavioral insight to apply in the context of
corruption deterrence, namely elaborating strategies that highlight both prescriptive actions
and negative externalities caused by the transgression of norms. Normative information,
through a heuristic-like process, may make some decision options more salient while
precluding counter-normative options to come to mind (59). In turn, as mentioned in a
previous paragraph, explicit norms may increase a sense of accountability (58).
Regarding the influence of normative information on corruption, we believe that the
present protocol and results introduce a new effective dimension to the experimental
literature on bribery. In most bribery experiments, researchers do not give participants
explicit normative information. At most, authors have relied on loaded frames, sanctions,
and/or the presence of externalities to implicitly signal norms, for example (41, 45, 48, 60;
but see 49 for an exception). We actually had a control condition that resembled some
protocols in the literature (i.e., the NoNorm_Externality condition) in which, as in other
bribery experiments, we used a loaded frame and the provision of the benefit by the official
caused a negative externality. The aim of the externality is to represent a situation in which
the provision of the benefit is not rightfully granted. Notice, however, that participants in
the NoNorm_Externality condition behaved differently from participants in condition
NoRight_Externality which explicitly represented the case where citizens did not acquire
the right to the benefit. More specifically, citizens tended to offer more bribes and officials
tended to grant more benefits when there was no explicit normative information (i.e., in the
NoNorm_Externality condition relative to the NoRight_Externality condition). This result
suggests that the mere presence of externalities associated with a decision option does not
unequivocally represent an unrightful option to participants, even in a framed scenario.
Last but not least, to our knowledge, the present protocol is the first to show an
effect of prescriptive norms on bribery in an experimental game. Another study reported
failures to find an effect of estimation of prescriptive norms on decisions in collusive
bribery games (39). Despite this study using loaded wording to describe the game, it is
possible that different participants interpreted differently whether the citizen´s benefit was
rightful, given that the citizen´s normative status was not explicitly provided. Getting rid of
this ambiguity in the present protocol could explain why prescriptive norms did affect
decisions in the experiment reported here.
All in all, the present contribution to the experimental bribery literature can be
summarized in the following three points: 1. the mere presence of negative externalities
may not be enough to trigger full normative expectations; 2. nonetheless, reminders of the
negative externalities associated with corruption may deter people from it; and 3. explicit
normative information may present synergies with information about negative externalities
in inhibiting bribery. These conclusions could be taken as preliminary behavioral insights
to nudge people away from corruption.
We would like to thank Urs Fischbacher, Irenaeus Wolff, Katrin Schmelz, and
David Hugh-Jones for useful comments on oral presentations of the present study. This
work was supported by The Argentine Council of Science and Technology (Spanish
acronym: CONICET; grant number PIP 2014-2016 N° 112-201301-00600-CO) and The
Argentine Agency of Science and Technology Promotion (Grant code: PICT 2013-1582).
1. Mauro P (1995) Corruption and Growth. The Quarterly Journal of Economics 110
2. Bertrand M, Djankov S, Hanna R, Mullainathan, S (2007)Obtaining a Driver's License in
India: An Experimental Approach to Studying Corruption. TheQuarterly Journal of
Economics 122(4):1639-1676.
3.Becker GS (1968) Crime and Punishment: An Economic Approach. The Journal of
Political Economy, 76 (2): 169-217.
4. Becker GS, Stigler GS (1974)Law Enforcement, Malfeasance, and Compensation of
Enforcers, The Journal of Legal Studies 3(1): 1-18.
5.La Porta R, Lopez-de-Silanes F, Shleifer, A (1999) Corporate ownership around the
world. The Journal of Finance 54 (2): 471-517.
6.Jain AK (2001) Corruption: A review. Journal of Economic Surveys 15 (1): 71-121.
7.Olken, BA (2007) Monitoring Corruption: Evidence from a Field Experiment in
Indonesia. Journal of Political Economy 115(2): 200-249.
8.Yamagishi T (1986) The provision of a sanctioning system as a public good. Journal of
Personality and Social Psychology 51(1): 110-116.
9. Hermann, B., Thöni, C., &Gachter, S. (2008). Antisocial punishment across societies.
Science 319, 1362-1367.
10.Richerson PJ, Boyd R (2005) Not by genes alone: How culture transformed human
evolution (Univ of Chicago Press, Chicago).
11.Sober E, Wilson DS (1998) Unto others: The evolution and psychology of unselfish
behavior (Harvard Univ Press, Cambridge).
12.Chudek M, Henrich J (2011) Culture-gene coevolution, norm-psychology and the
emergence of human prosociality. Trends CognSci 15(5):218–226.
13.Coleman JS (1990) Foundations of Social Theory (Harvard Univ Press, Cambridge).
14.Cialdini, R. B., Kallgren, C. A., & Reno, R. R. (1991). A focus theory of normative
conduct. Advances in Experimental Social Psychology 24: 201–234.
15.Sripada C, Stich S (2007) A Framework for the Psychology of Norms. In: P Carruthers,
S Laurence and S Stich, ed., The Innate Mind: Volume 2: Culture and Cognition, 1st ed.
Oxford: Oxford University Press, pp.281-301.
16.Fehr E, Fischbacher U, Gächter S (2002) Strong reciprocity, human cooperation and the
enforcement of social norms. Human Nature 13: 1–25.
17.Fehr E, Fischbacher U (2004) Social norms and human cooperation. Trends inCognitive
Sciences 8 (4), 185-190.
18. Bowles S, Gintis G (2011) A Cooperative Species: Human Reciprocity and its
Evolution (Princeton Univ Press, Princeton). Quervain, DJF, Fischbacher U, Treyer V, Schellhammer M, Schnyder U, Buck A,
Fehr E (2004) The neural basis of altruistic punishment. Science 305: 1254–1258.
20.Buckholtz JW, Asplund CL, Dux PE, Zald, DH, Gore JC, Jones OD, Marois R (2008)
The Neural Correlates of Third-Party Punishment. Neuron60 (5): 930–940.
21.Ruff CC, Ugazio G, Fehr E (2013) Changing social norm compliance with noninvasive
brain stimulation. Science 342: 482–484.
22.Cialdini RB, Reno RR, Kallgren CA (1990) A focus theory of normative conduct:
Recycling the concept of norms to reduce littering in public places. Journal of Personality
and Social Psychology 58: 1015–1026. doi:10.1037/0022-514.58.6.1015.
23.Krupka E, Weber, RA (2009) The focusing and informational effects of norms on pro-
social behavior. Journal of Economic Psychology 30 (3): 307–320.
24.Raihani NJ, McAuliffe K(2014)Dictator Game giving: The importance of descriptive
versus injunctive norms. PLoS ONE9(12): e113826. doi:10.1371/journal.pone.0113826
25.Cialdini RB, Demaine LJ, Sagarin BJ, Barrett DW, Rhoads K, Winter PL (2006)
Managing social norms for persuasive impact. Social Influence 1(1): 3-15.
26.Elster J (1989) Nuts and bolts for the social sciences (Cambridge Univ Press,
27.Bicchieri C (2006) The grammar of society: The nature and dynamics of social norms
(Cambridge Univ Press, Cambridge).
28.Fehr, E, Schmidt K (1999) A Theory of Fairness, Competition, and Cooperation.
Quarterly Journal of Economics 114: 817-868.
29.Haidt, J (2007) The New Synthesis in Moral Psychology.Science 316:9981002. DOI:
30. Gigerenzer G, Todd PM, ABC Reseacrh Group (1999) Simple Heuristics that make us
smart (Oxford Univ Press, Oxford New York).
31.Kahneman D (2011) Thinking, Fast and Slow (Straus and Giroux, New York).
32.Rand DG, Peysakhovich A, Kraft-Todd GT, Newman GE, Wurzbacher O, Nowak MA,
Greene JD (2014) Social heuristics shape intuitive cooperation. Nat Commun 5:3677.doi:
33.Sunstein C (2005) Moral Heuristics. Behavioral and Brain Sciences 28 (4):531-542.
34.Baron, J (1994) Nonconsequentialist decisions. Behavioral and Brain Sciences 17(1): 1-
35.Sousa Lourenço J, Ciriolo E, Rafael Almeida S, Troussard X (2016) Behavioural
insights applied to policy: European Report 2016.
f. doi: 10.2760/903938
36.Abbink K, Serra D (2012) Anticorruption Policies: Lessons from the Lab. New
Advances in Experimental Research on Corruption (Research in Experimental Economics,
Volume 15), eds Serra D, Wantchekon L (Emerald Group Publishing Ltd.), pp. 77–115.
37. Kobis NC, van Prooijen JW, Righetti F, Van Lange PAM (2015) “Who doesn´t?” – The
oimpact of descriptive norms on corruption. PLoS ONE 10(6): e0131830. doi:
38. Chaudhuri A, Paichayontvijit T,Sbai E(2016) The Role of Framing, Inequity and
History in a Corruption Game: Some Experimental Evidence.Games, 7, 13, 1-24;
39. D´Adda G, Drouvelis M, Nosenzo D (2016) Norm elicitation in within-subject designs:
Testing for order effects. Journal of Behavioral and Experimental Economics 62, 1-7.
40.Engel C, Georg SJ (2016) Symmetric vs. Asymmetric PunishmentRegimes for Collusive
Bribery.American Law and Economics Review. doi:10.1093/aler/ahw005
41.Barr A, Serra D (2009) The effects of externalities and framing on bribery in a petty
corruption experiment. Experimental Economics 12(4): 488-503.
42.Mazar N, Amir O, Ariely D (2008).The dishonesty of honest people: a theory of self-
concept maintenance. J.Market.Res. 45: 633–644.doi:10.1509/jmkr.45.6.633
43.Ariely, D (2012) The (Honest) Truth about Dishonesty (Harper Collins, New York).
44.Fischbacher U, Föllmi-Heusi F (2013) Lies In Disguise—An Experimental Study on
Cheating. Journal of the European Economic Association 11: 525–547.
45.Abbink K, Irlenbusch B, Renner E (2002) An Experimental Bribery Game. Journal of
Law, Economics and Organization 18(2): 428-454.
46.Falk A, Czech N (2013) Morals and Markets. Science 340: 707-711.DOI:
47.Abbink A, &Hennig-Schmidt H (2006) Neutral vs loaded instructions in bribery
experiment. Experimental Economics 9(2):103-121.
48.Cameron L, Chaudhuri A, Erkal N, Gangadharan L (2009). Propensities to engage in
and punish corrupt behavior: Experimental evidence from Australia, India, Indonesia and
Singapore. Journal of Public Economics 93: 843–851.
49.Banerjee, R (2016) On the interpretation of bribery in a laboratory corruption game:
moral frames and social norms. Experimental Economics 19(1): 240-267.
50. Fischbacher, U (2007) z-Tree. Zurich Toolbox for Ready-made Economic Experiments.
Experimental Economics, 10: 171-178.
51.Fershtman C, Gneezy U, List JA (2012). Equity Aversion: Social Norms and the Desire
to be Ahead. American Economic Journal: Microeconomics 4(4): 131–144. doi:
52.Alekseev A, Charness, G, Gneezy U (2017) Experimental methods: When and why
contextual instructions are important. Journal of Economic Behavior & Organization134,
53.Guth W, Schmittberger R, Schwarze B (1982) An experimental analysis of ultimatum
bargaining. Journal of Economic Behavior & Organization 3(4): 367-388.
54.Abbink K, Dasgupta U, Gangadharan L, Jain T. (2014) Letting the briber go free: an
experiment on mitigating harassment bribes. Journal of Public Economics 111: 17-28.
55.Lambsdorff JG, Frank B (2010) Bribing versus Gift-Giving: An Experiment. Journal of
Economic Psychology 31(3): 347-357.
56.Rivas MF (2013) An Experiment On Corruption And Gender. Bulletin of Economic
Research 65: 10–42. doi:10.1111/j.1467-8586.2012.00450.x
57.Sezer O, Gino F, Bazerman M (2015) Ethical blind spots: explaining unintentional
unethical behavior, Current Opinion in Psychology 6:77–81.
58.Pfattheicher S, Keller J(2015) The watching eyes phenomenon: The role of a sense of
being seen and public self-awareness. European Journal of Social Psychology 45 (5): 560-
59.Hlobil U (2015) Social norms and unthinkable options. Synthese 193: 2519-2537
60. Barr A, Lindelow M, Serneels P (2009) Corruption in public service delivery: an
experimental analysis. Journal of Economic Behaviour and Organization 72 (1): 225-239.
... trillion per year 2 . In light of these tremendous costs, a broad array of research has identified multiple factors shaping engagement in bribery and corruption, including social norms [3][4][5] , negative externalities 6,7 , as well as various types of punishment [8][9][10] . ...
Full-text available
People differ fundamentally in their propensity to engage in bribery. Herein, we tested the effects of inequality of (monetary) resources between bribers (individuals who can offer bribes) and bribees (individuals who might receive bribes) on bribery. We implemented a bribery game where bribers and bribees could engage in bribery to benefit themselves at the cost of others. To manipulate resource inequality, we allocated each briber with a high or a low endowment (resulting in “rich” and “poor” bribers) and matched them with a bribee who received either a high or a low endowment (resulting in “rich” and “poor” bribees). In a large-scale pre-registered study (N = 2,401), we found that rich (vs. poor) bribees exhibited a stronger preference for (1) accepting bribes from rich, rather than poor bribers, and (2) accepting higher, rather than lower bribes. Furthermore, we found that both poor and rich bribers indicated they would only accept relatively high bribes from rich (rather than poor) bribees. Finally, rich and poor bribers benefitted in different ways when engaging in bribery: to get their bribes accepted, rich bribers had to spend lower proportions of their endowments than poor bribees, however, rich bribers had to spend more in absolute terms.
Full-text available
El enfoque de Behavioral Insights (BI) implica utilizar conocimientos de las ciencias del comportamiento en el diseño, implementación y evaluación de políticas públicas. En general, los proyectos asociados a BI buscan afectar la conducta de los ciudadanos en sentidos socialmente deseables. En este artículo planteamos que el mismo enfoque puede ser utilizado como marco para "pensar" cómo moldear la conducta de los gobernantes con fines prosociales. Con este objetivo, discutimos dos conceptos: 1) la alineación de los incentivos de gobernantes y ciudadanos, y 2) el enfoque Nudge, que implica estructurar los ambientes de decisión para guiar a los agentes hacia fines prosociales. Concluimos mencionando algunas promesas y limitaciones del enfoque BI en general y de nuestras propuestas específicas en particular
Full-text available
With respect to questions of fact, people use heuristics – mental short-cuts, or rules of thumb, that generally work well, but that also lead to systematic errors. People use moral heuristics too – moral short-cuts, or rules of thumb, that lead to mistaken and even ab- surd moral judgments. These judgments are highly relevant not only to morality, but to law and politics as well. Examples are given from a number of domains, including risk regulation, punishment, reproduction and sexuality, and the act/omission distinction. In all of these contexts, rapid, intuitive judgments make a great deal of sense, but sometimes produce moral mistakes that are replicated in law and pol- icy. One implication is that moral assessments ought not to be made by appealing to intuitions about exotic cases and problems; those intuitions are particularly unlikely to be reliable. Another implication is that some deeply held moral judgments are unsound if they are products of moral heuristics. The idea of error-prone heuristics is especially controversial in the moral domain, where agreement on the correct answer may be hard to elicit; but in many contexts, heuristics are at work and they do real damage. Moral framing effects, in- cluding those in the context of obligations to future generations, are also discussed.
Full-text available
We investigate the role of framing, inequity in initial endowments and history in shaping behavior in a corrupt transaction by extending the one-shot bribery game introduced by Cameron et al. (2009) to a repeated game setting. We find that the use of loaded language significantly reduces the incidence of bribery and increases the level of punishment. Punishment of bribery leads to reduced bribery in future. The evidence suggests that this game captures essential features of a corrupt transaction, over and above any sentiments of inequity aversion or negative reciprocity However, showing subjects the history of past play has little effect on the level of corruption.
Full-text available
We investigate norms of corruption using the norm-elicitation procedure introduced by Krupka and Weber (2013). We use a within-subject design whereby the norms are elicited from the same subjects who are observed making choices in a bribery game. We test whether the order in which the norm-elicitation task and the bribery game are conducted affects elicited norms and behavior. We find little evidence of order effects in our experiment. We discuss how these results compare with those reported in the existing literature.
An important methodological issue in experimental research is the extent to which one should use context-rich or abstract language in the instructions for an experiment. The traditional use of abstract context in experimental economics is commonly viewed as a way to achieve experimental control. However, there are some advantages to using context-framed instructions, such as “employer and worker” instead of “player 1 and player 2.” Meaningful context can enhance understanding of an environment and reduce confusion among participants, particularly when a task requires sophisticated reasoning, and hence may yield responses of better quality. In emotionally-charged research questions, such as pollution or bribes, contextual instructions may affect behavior in the experiment, but this effect may be appropriate as it relates to the research question. Our review of the evidence from the literature indicates that in the great majority of cases meaningful language is either useful or produces no change in behavior. Nevertheless, a few important considerations are worth keeping in mind when using rich context. Finally, we see the choice of context as being an expansion of the experimenter’s toolkit and a factor to consider in experimental design.
In major legal orders such as the United Kingdom, the United States, and France, bribers and recipients face equally severe criminal sanctions. In contrast, countries like China, Russia, and Japan treat the briber more mildly. Asymmetric punishment has been shown to help deter harassment bribery. However, we conjecture that asymmetry is ineffective when applied to collusive bribes. Instead of deterring bribes, asymmetry might enable the briber to enforce the corrupt deal. To test this hypothesis, we design and run a lab experiment in Bonn (Germany) and Shanghai (China) with exactly the same design. The results show that, in both countries, with symmetric punishment bribers are less likely to report to the authorities. Officials are less likely to grant the favor. In Shanghai, corrupt offers are then also less likely. If we frame the experiment as collusive corruption, effects are less pronounced, but we can replicate all of them.
We sometimes violate social norms in order to express our views and to trigger public debates. Many extant accounts of social norms don’t give us any insight into this phenomenon. Drawing on Cristina Bicchieri’s work, I am putting forward an empirical hypothesis that helps us to understand such norm violations. The hypothesis says, roughly, that we often adhere to norms because we are systematically blind to norm-violating options. I argue that this hypothesis is independently plausible and has interesting consequences. It implies, e.g., that some experimental paradigms for investigating social norms have hitherto unnoticed shortcomings.