Corruption as a Self-Fulfilling Prophecy: Evidence from a Survey Experiment in Costa Rica

Gingerich, Daniel W.; Oliveros, Virgina; Corbacho, Ana; Ruiz-Vega, Mauricio Working Paper Corruption as a Self-Fulfilling Prophecy: Evidence from a Survey Experiment in Costa Rica IDB Working Paper Series, No. IDB-WP-546 Provided in Cooperation with: Inter-American Development Bank (IDB), Washington, DC Suggested Citation: Gingerich, Daniel W.; Oliveros, Virgina; Corbacho, Ana; Ruiz-Vega, Mauricio (2015) : Corruption as a Self-Fulfilling Prophecy: Evidence from a Survey Experiment in Costa Rica, IDB Working Paper Series, No. IDB-WP-546, Inter-American Development Bank (IDB), Washington, DC, https://hdl.handle.net/11319/6826 This Version is available at: https://hdl.handle.net/10419/115508 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. http://creativecommons.org/licenses/by-nc-nd/3.0/igo/legalcode Corruption as a Sel f - Fulfilling Prophecy: Evidence from a Survey Experiment in Costa Rica Daniel W. Gingerich Virginia Oliveros Ana Corbacho Mauricio Ruiz-Vega Department of Research and Chief Economist / Institutions for Development IDB-WP-546 IDB WORKING PAPER SERIES No. Inter-American Development Bank February 2015 Corruption as a Sel f -Ful f illing Prophecy: Evidence from a Survey Experiment in Costa Rica Daniel W. Gingerich* Virginia Oliveros** Ana Corbacho*** Mauricio Ruiz-Vega*** * University of Virginia ** Tulane University *** International Monetary Fund 2015 Inter-American Development Bank Cataloging-in-Publication data provided by the Inter-American Development Bank Felipe Herrera Library Corruption as a self-fulfilling prophecy: evidence from a survey experiment in Costa Rica / Daniel W. Gingerich, Virginia Oliveros, Ana Corbacho, Mauricio Ruiz-Vega. p. cm. — (IDB Working Paper Series ; 546) Includes bibliographic references. 1. Corruption—Public opinion—Costa Rica. 2. Corruption—Surveys—Costa Rica. 3. Political corruption —Public opinion—Costa Rica. 4. Political corruption—Surveys—Costa Rica. I. Gingerich, Daniel W., 1977-. II. Oliveros, Virginia. III. Corbacho, Ana. IV. Ruiz-Vega, Mauricio. V. Inter-American Development Bank. Department of Research and Chief Economist. VI. Series. IDB-WP-546 http://www.iadb.org Any dispute related to the use of the works of the IDB that cannot be settled amicably shall be submitted to arbitration pursuant to the UNCITRAL rules. The use of the IDB’s name for any purpose other than for attribution, and the use of IDB’s logo shall be subject to a separate written license agreement between the IDB and the user and is not authorized as part of this CC-IGO license. Following a peer review process, and with previous written consent by the Inter-American Development 2015 Copyright © Inter-American Development Bank. This work is licensed under a Creative Commons IGO 3.0 Attribution-NonCommercial-NoDerivatives (CC-IGO BY-NC-ND 3.0 IGO) license ( http://creativecommons.org/licenses/by-nc-nd/3.0/igo/legalcode) and may be reproduced with attribution to the IDB and for any non-commercial purpose. No derivative work is allowed. Bank (IDB), a revised version of this work may also be reproduced in any academic journal, including those indexed by the American Economic Association’s EconLit, provided that the IDB is credited and that the author(s) receive no income from the publication. Therefore, the restriction to receive income from such publication shall only extend to the publication’s author(s). With regard to such restriction, in case of any inconsistency between the Creative Commons IGO 3.0 Attribution-NonCommercial-NoDerivatives license and these statements, the latter shall prevail. Note that link provided above includes additional terms and conditions of the license. The opinions expressed in this publication are those of the authors and do not necessarily reflect the views of the Inter-American Development Bank, its Board of Directors, or the countries they represent. Abstract1 An influential body of scholarship argues that corruption behaves as a selffulfilling prophecy. The idea of this work is that levels of corruption emerge endogenously as a result of a society-wide coordination game in which the individual returns to corrupt behavior are a function of how disposed towards corruption the other members of society are perceived as being. An empirical implication of this logic is that if one were to exogenously perturb beliefs about societal levels of corruption upward, willingness to engage in corruption should increase as a consequence. The current paper evaluates this claim by utilizing an information experiment embedded in a large-scale household survey conducted in the Gran Área Metropolitana (GAM) of Costa Rica from October 2013 to April 2014 (n=4200). Changes in beliefs about corruption were induced via the random assignment of an informational display depicting the increasing percentage of Costa Ricans who have experienced or directly observed an act of corruption (from 2006 to 2011). The paper finds that, on average, assignment to this display (relative to the control condition) increased the probability that a respondent would be willing to pay a bribe to a police officer by approximately 0.04 to 0.08, thereby providing supporting evidence for the self-fulfilling prophecy hypothesis. JEL Classification: D73, D84, C83, C90 Keywords: Corruption, Spillover effects, Survey experiments, Sensitive question techniques 1 This paper was prepared when Ana Corbacho was Sector Economic Advisor in the Institutions for Development Sector at the Inter-American Development Bank. Correspondence should be directed to Daniel W. Gingerich (dwg4c@virginia. 1 Introduction Scholarship on corruption has long emphasized the notion that corruption may corrupt. In particular, a number of scholars have developed the argument that levels of corruption may emerge endogenously as a result of a society-wide coordination game in which the individual returns to corrupt behavior are a function of the inclination towards corruption of the other members of society (Andvig and Moene, 1990; Mishra, 2006; Persson, Rothstein and Teorell, 2013). This view of corruption as a coordination game has some relatively straightforward empirical implications. One of these is the expectation that the higher the level of perceived corruption in a society, the more likely it is that individual members of that society would be willing to engage in corrupt behavior. In other words, corruption behaves as a self-fulfilling prophecy. Much is at stake in empirically assessing whether or not corruption corrupts. Evidence in favor of this hypothesis would support theoretical claims that corruption is an endeavor characterized by strategic complementarities among economic agents. As is well known, the presence of such complementarities lends itself to the existence of multiple equilibria (e.g., Milgrom and Roberts, 1990). In the context studied here, this means that a given polity might find itself locked into a relatively high or relatively low equilibrium level of corruption depending on the precise manner in which citizens resolve their coordination dilemma. Thus, accepting the view that corruption corrupts implies that long-run variation in corruption encountered across polities must be understood as a consequence of two factors: first, the set of institutions that determine the costs and benefits that accrue to corrupt behavior (thereby affecting the incidence of corruption in any feasible equilibrium) and, second, the mix of cultural norms and beliefs that determine which of the potential corruption equilibria will prevail at a given time. For social scientists, this means that a comprehensive account of long-run differences across space and time in the quality of governance would necessarily consider both the organization of the formal rules of the game, so to speak, as well as the informal beliefs that determine how nation-states and other collectivities come to think about themselves. Does corruption corrupt? Does the level of perceived corruption in a society affect an individual’s willingness to engage in corrupt behavior? In this paper, we provide evidence that an individual’s willingness to engage in corrupt behavior is indeed affected by her perception of the level of corruption in society. Using data from an original survey experiment in Costa Rica, we show that learning about increasing levels of corruption increases the likelihood that citizens would be willing to bribe a police officer in order to avoid paying a traffic ticket. Costa Rica provides a propitious environment for studying the questions that concern us here. Although corruption in Costa Rica is relatively low by regional standards, the country has experienced a substantial increase in perceived corruption in recent years. According to the 2013 Corruption Perception Index by Transparency International, Costa Rica is ranked 49 among 177 2 countries, considerably above Honduras (140), Nicaragua (127), Guatemala (123), Mexico (106), and Panama (102). In fact, the only two Latin American countries that perform better than Costa Rica on this metric are Uruguay (19) and Chile (22). However, according to a nationally representative survey conducted by Latinobarómetro, the number of people who have witnessed an act of corruption has increased from 16 percent in 2006 to 24 percent in 2011.2At the elite level, the country has seen repeated political corruption scandals over the last decade, including the indictment of three different past presidents for bribery (and the conviction of two of them) as well as a number of forced resignations among cabinet officials. Corruption scandals of that magnitude have no precedent in earlier periods of the modern history of the country. Moreover, the growth of corruption in Costa Rica also appears to have been accompanied by an increase in other forms of societal disorder, in particular, a rise in violent crime and the use of the country as a transshipment point in the international narcotics trade.3Given these abrupt changes occurring in a country once characterized as the “Switzerland of Central America,” beliefs about how deeply embedded corruption is in Costa Rica society are likely to be in flux for many individuals. For this reason, the country is a natural setting for exploring how information about the scope of corruption may drive corrupt behavior. 2 Corruption as a Self-Fulfilling Prophecy: Theory and Evidence Theoretical approaches to the study of corruption have long emphasized that the phenomenon is subject to coordination dilemmas and herd behavior, and, as a consequence, that high (or low) levels of corruption tend to feed upon themselves and persist over time. The specific mechanisms adduced to explain why corruption tends to operate as a self-fulfilling prophecy are many and varied. Some accounts concentrate on how the existence of corruption undermines sanctioning mechanisms, thereby furthering the incidence of corruption in the first place. Such a dynamic may be operative if the probability of detection is inversely related to the frequency of corrupt behavior (Lui, 1986) or if the costs of detection are mitigated by the fact that supervisors responsible for monitoring corruption are also subject to bribery (Cadot, 1987; Andvig and Moene, 1990). Moreover, in a regulatory setting, corruption can undermine sanctioning (thereby fomenting greater corruption) if the frequency of illicit behavior by firms reduces the returns to skill formation by regulators (Mishra, 2006). Another perspective holds that corruption can corrupt by affecting search costs. The idea here is that the more inclined towards corruption are members of the private sector and government, 2This perception of increasing corruption was corroborated by the focus groups we conducted in Costa Rica prior to the survey. See the Appendix for further details on the focus groups. 3Although relatively safe by regional standards, the country’s homicide rate nearly doubled from 2001 to 2010, growing from 6.4 per 100,000 to 11.5 per 100,000 (Departamento de Planificación, Anuario Policial 2010). 3 or, equivalently, the more certain each set of actors is about the corruptibility of the other, the less costly it is to find a partner willing to consummate a corrupt bargain and the more attractive initiating a corrupt transaction will be (Ryvkin and Serra, 2012; see also Andvig and Moene, 1990). Recent work from a psychological perspective emphasizes the role of guilt aversion in generating corruption spillovers. In this view, the moral cost public officials associate with engaging in corruption is a function of how corrupt they think the public perceives them (the officials) to be. Greater corruption by officials pushes citizens’ beliefs about public probity downwards, thereby reducing moral costs among officials and instigating a corruption spiral (Huang and Wu, 1994; Balafoutas, 2011). Accounts focusing explicitly on long-run dynamics tend to emphasize the consequences of reputational lock-in for collectivities or the intergenerational transmission of cultural values. In an influential paper, Tirole, 1996 noted that due to imperfect observability of past individual behavior, economic agents may condition trading decisions on the reputation of the collectivity to which their partner belongs. As such, agents in organizations whose founding members have acted corruptly have little incentive to behave honestly, since potential trading partners will anticipate corrupt dealing on their part. In a related contribution, Hauk and Sáez-Marti, 2002 argue that in the long run corruption corrupts due to educational decisions made in the home. If parents anticipate that future public policy will favor the dishonest, they will invest in transmitting their cultural values to their children only if they themselves are dishonest, and abstain from doing so otherwise. Consistent with these ideas is work that advances the notion of corruption spirals being driven by imitative behavior in which agents act corruptly if doing so has been a successful strategy in the past for their neighbors (e.g., Accinelli and Carrera, 2012).4 Political economy treatments of corruption note that externalities generated by political exchange can also give birth to corruption spirals. Two contributions argue that corruption can corrupt by affecting the selection mechanism determining who holds public office. One such account claims that, since the value of holding office (i.e., ego rents) is a function of the ability cum honesty of existing officeholders, once the current political elite has been corrupted, it will typically be only dishonest citizens who will be willing to stand for public office in the future (Caselli and Morelli, 2004). The other claims that, since the capacity of interest groups to smear or threaten politicians is greater the lower the overall ability/honesty of the political elite, the less attractive public office will be for honest would-be politicians when the current political elite is corrupt (Dal Bó and Di Tella, 2006). A third political economy treatment of how corruption corrupts focuses on the organization of party patronage networks in the bureaucracy (Gingerich, 2009). It claims that since the value of the rewards offered by a party leader to rank-and-file militants in the bureaucracy 4For other examples of imitation (or “learning by observing”) see Bikhchandani, Hirshleifer and Welch, 1998. 4 is greater the stronger the party in electoral terms, said militants will be more inclined to engage in state plunder at the behest of their party if they anticipate their copartisans are doing so as well. Relative to the extensive theoretical work on the topic, the empirical literature examining the degree to which corruption corrupts is still at a very early stage. Several papers have used aggregate data to show that corruption might be contagious. For example, Dong and Torgler, 2012 use province-level data for China from 1998 to 2007 to show that social interaction has a significant positive effect on corruption. Using state-level U.S. data from 1995 to 2004, Goel and Nelson, 2007 also find evidence of neighboring corruption: a 10 percent increase in corruption in neighboring states increases corruption in a state by about 4 to 11 percent. López-Valcarcel, Jiménez and Perdiguero, 2014 similarly find evidence that corruption is contagious using a dataset of local Spanish municipalities from 2001-2010. Studies by Becker, Egger and Seidel, 2009 and Goel and Saunoris, 2014 utilize cross-national data and spatial econometric techniques to estimate the degree to which corruption in one country affects its neighbors. In both cases, the authors find evidence of spillover effects. However, contrary to the findings of these studies, Márquez, Salinas-Jiménez and Salinas-Jiménez, 2011 also use a spatial econometric approach applied to a cross-section of countries and find no evidence of corruption spillovers. Articles that explore the contagiousness or self-fulfilling hypothesis using individual-level data, as we do in this paper, are few and far between. To the best of our knowledge, the only article that directly studies the relationship between perceived corruption in a society and individual behavior using individual-level data is an article by Dong, Dulleck and Torgler, 2012. Using data from the European Values Survey, they show that the more respondents perceive others as being corrupt, the more tolerant they are towards someone accepting a bribe in the course of their duties. Recent work by Simpser, 2013 provides individual-level evidence on one mechanism by which corruption can corrupt, namely, the intergenerational transmission of attitudes towards corruption. Using data from the U.S. and European General Social Surveys, he finds that, holding current institutional context constant, attitudes towards corruption in a respondent’s ancestral country of origin strongly predict current attitudes towards the same. In laboratory experiments conducted with undergraduates in the United States and India, Innes and Mitra, 2013 find that information treatments indicating a peer propensity towards dishonesty in a game setting increased the likelihood of dishonest play, a finding that they interpret as evidence for the contagiousness of corruption hypothesis. In a broader reading, our paper can be seen as a contribution to the large empirical literature on how social context affects individuals’ willingness to engage in crime. In a seminal paper, Kahan, 1997 outlined the logic behind these findings. He argues that individuals are much more likely to commit crimes when crime is widespread in their communities because they are likely to assume that both the probability of being caught and the potential cost of stigmatization are 5 severe problems of confounding, we induce exogenous variation in beliefs about corruption via the random assignment of respodents to distinct informational treatments. Three informational treatments were employed in our experiment: a corruption treatment, an inefficiency treatment, and a control condition. In the corruption treatment, respondents were presented with a flyer depicting the increasing percentage of Costa Ricans who have directly observed an act of corruption (covering the period from 2006 to 2011). A second treatment, the inefficiency treatment, was introduced as a “placebo” treatment in order to test whether respondents were affected by the information included in the corruption treatment or just the fact that they were given a flyer with negative information about the capacity of the Costa Rican state to deal with illicit behavior. In the inefficiency treatment, respondents were presented with a flyer presenting the (lack of) productivity of the legal system in dealing with a particular crime: assault with a deadly weapon. In the control condition, respondents were not presented with a flyer. Randomization of treatment assignment was programmed directly into the portable digital assistants (PDAs) the enumerators used to conduct the survey. Random assignment to different types of information ensured that, on average, groups of respondents were indistinguishable on both observable and unobservable characteristics. Appendix Table 4 provides evidence on balance in observable respondent characteristics across treatment groups. Excluding the treatment assignments, respondents were asked questions from otherwise identical questionnaires. The two flyers are presented in Figure 2. The flyer on the left-hand side is the corruption treatment. In bold text, it states, “Did you know that corruption in Costa Rica has increased?” Below the statement appears a bar graph showing the increase in the percentage of Costa Ricans who had personally witnessed an act of corruption from 2006 (16 percent) to 2011 (24 percent). At the bottom right-hand side of the flyer, the source of the information, a nationally representative survey conducted by Latinobarómetro, is displayed. The flyer on the right hand side is the inefficiency treatment. It states, “In 2011, 6,812 cases of assault with a deadly weapon entered into the judicial system. However, only 47 individuals were sent to jail for this crime.” Below the statement appears a bar graph showing the relative magnitudes of the number of cases filed for this crime (6812), the number of judicial decisions made on cases of the crime (333), and the number of individuals actually sent to jail (47). The source of the information, National Judicial Statistics, is displayed on the bottom right corner. After exposure to oneof thethree experimental conditions, respondents werelater prompted to respond to the aforementioned question about willingness to bribe, first in crosswise format and then, at the end of the survey, in direct questioning format. 12 Figure 2. The Two Treatments, Corruption and Judicial Inefficiency 5 Corruption Does Corrupt We begin our analysis by examining simple differences in the estimated proportion of respondents willing to bribe according to treatment condition. Results are presented utilizing our joint response approach, the direct survey responses by themselves, as well as the crosswise responses by themselves. The findings are presented in a series of barplots displayed in Figure 3. All told, the results provide support for the self-fulfilling hypothesis that an individual’s willingness to engage in corrupt behavior is affected by her perception of the level of corruption in society. According to the estimates based on the joint response model, the proportion of respondents assigned to the corruption information treatment who would be willing to bribe a police officer was 0.35, whereas the proportion of respondents assigned to the control condition willing to do the same was only 0.27. Thus, the average treatment effect of information about the increasing scope of corruption was equal to 0.08. This is a substantively large effect: exposure to the corruption treatment was estimated to increase the proportion of respondents willing to bribe by 28 percent. In addition to being large in magnitude, the average treatment effect was statistically significant by conventional standards. Our placebo treatment–information about judicial inefficiency–did not have a statistically significant impact on willingness to bribe. 13 Figure 3. Estimated Proportion of Respondents Willing to Bribe by Treatment Condition and Questioning Method control inefficiency corruption Joint Response Model ATE(corr. vs cont.) = 0.08 [0.01,0.15] 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.35 (0.03) 0.24 (0.03) 0.27 (0.03) control inefficiency corruption Direct Responses Only ATE(corr. vs cont.) = 0.03 [−0.00,0.06] 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.20 (0.01) 0.18 (0.01) 0.17 (0.01) control inefficiency corruption Crosswise Responses Only ATE(corr. vs cont.) = 0.06 [−0.02,0.13] 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.27 (0.03) 0.17 (0.03) 0.21 (0.03) Note: Standard errors in parentheses. Ninety-five percent confidence intervals for the average treatment effect in square brackets. 14 In addition to providing evidence that corruption corrupts, the figure also clearly illustrates the virtues of using the joint response approach for asking questions about bribery. In every treatment condition, estimates of willingness to bribe based on direct responses were below those based on the joint response model and crosswise responses only. Moreover, it would appear that desirability bias substantially attenuated downward the estimate of the average treatment effect of exposure to information about corruption: the average treatment effect based only on direct responses was about half that based only on the crosswise responses and just about one third as large as the average treatment effect based on the joint responses. The use of only the crosswise responses resulted in an estimated average treatment effect equal to 0.06, below but roughly similar to that based on the joint responses. However, this estimate was not statistically significant at conventional standards, due in part to the fact that it is based on an inefficient statistical approach that failed to incorporate the potentially useful information available from the pattern of direct responses. Table 2. Relationship between Information Treatments and Bribery of Police (joint response model) Parameters Model 1 (n= 4193)Model 2 (n= 4192) estimate s.e. 95% int. estimate s.e. 95% int. Diagnostic parameters b T 10.61 0.03 [0.55,0.67] 0.63 0.03 [0.58,0.68] b L 10.36 0.03 [0.30,0.41] 0.34 0.03 [0.28,0.39] b T 00.97 0.00 [0.96,0.98] 0.97 0.00 [0.96,0.98] Explanatory parameters Constant -0.98 0.09 [-1.17,-0.80] 2.81 0.46 [1.97,3.77] Corruption treatment 0.24 0.10 [0.04,0.45] 0.23 0.11 [0.01,0.43] Inefficiency treatment 0.01 0.11 [-0.18,0.22] 0.03 0.12 [-0.19,0.24] Male _ _ _ 0.89 0.10 [0.70,1.10] Log(age) _ _ _ -1.29 0.12 [-1.54,-1.06] Education (base=some college) _ _ _ Primary or less _ _ _ 0.12 0.16 [-0.17,0.40] Secondary incomplete _ _ _ 0.40 0.15 [0.13,0.68] Secondary complete _ _ _ 0.29 0.15 [-0.01,0.59] Some technical _ _ _ 0.05 0.28 [-0.51,0.62] ATE (corruption vs. control) 0.05 0.02 [0.01,0.09] 0.04 0.02 [0.00,0.08] The next step of our analysis consisted of the use of the joint response-modified logistic regression framework described earlier. (Unlike the joint response estimates presented in Figure 3, this approach pools the estimation of the diagnostic parameters across all three treatment conditions.) Table 2 depicts coefficient estimates showing the impact of the two information treatments 15 on a respondent’s willingness to offer a bribe to a police officer to avoid paying a traffic ticket. Two regression models were estimated: one in which the informational treatments entered as the sole explanatory variables and one in which we included the age, gender, and education of the respondent. Previous findings in the literature suggest that these covariates might relate to willingness to bribe in a number of ways. There is some empirical evidence that women and older individuals might be less involved in corruption and/or less likely to condone corruption than men (Swamy, Knack, Y Lee et al., 2003; Torgler and Valev, 2006; Torgler and Valev, 2010). With respect to education„ some scholars have found that the more educated (or richer, variables that are often used as proxies) are more tolerant of corruption (Winters and Weitz-Shapiro, 2013), while others have presented results that imply the opposite (cf. Manzetti and Wilson, 2007). Moreover, there is evidence that more informed citizens (presumably, the more educated) are less likely to be receptive to new information (Lupu, 2013). If that is the case, the more educated might oppose corruption equally or more than the less educated, but the effect of our information experiment would show a smaller effect for the more educated. Both estimated models told a very similar story: exposing respondents to information about the growing scope of corruption in society made them significantly more likely to indicate a willingness to bribe a police officer to avoid paying a traffic ticket. Again, it thus appears that, as predicted by theory, corruption really does corrupt. As shown in the table, the average treatment effect of exposure to information about the growing scope of corruption was estimated to be 0.05 in the baseline model and 0.04 in the model with additional covariates. In both cases, the average treatment effects were statistically significant by conventional standards. As above, in neither of the estimations did the judicial inefficiency treatment have any discernible impact on willingness to bribe, suggesting that the effect of the corruption treatment was caused by the specific information contained in it and not by just the fact that respondents were exposed to some negative information about the capacity of the Costa Rican state to deal with illicit activity. In terms of background characteristics, men appeared to be substantially more inclined to bribe than women, younger respondents more inclined to bribe than older respondents, and individuals with incomplete secondary school educations more inclined to bribe than individuals with some exposure to college (the baseline education category). The estimated values of the diagnostic parameters point to the importance of using sensitive survey techniques for asking questions about bribery. According to the estimated models, the probability that an individual inclined to bribe would admit as much under direct questioning was 0.61-0.63, whereas the probability that such an individual would lie under direct questioning was 0.34-0.36. Thus, our results indicated that there was a fairly substantial degree of respondent evasiveness under direct questioning. 16 6 Conclusion This paper provides new evidence in favor of the proposition that corruption corrupts. Utilizing data from a large-scale social survey conducted in the Gran Área Metropolitana of Costa Rica, one that combined both a survey experiment and a novel format for asking sensitive questions, we find that exposing citizens to information about the growing scope of corruption in their society made them individually more disposed to engage in corruption. In this way, our paper offers empirical support to a large theoretical literature on corruption that has long claimed that the phenomenon is characterized by strategic complementarities between economic agents, a state of affairs that implies that multiple, stable corruption regimes may be feasible for a given polity at a given time. For students of governance, the import of our findings is that if one accepts that corruption operates as a self-fulfilling prophecy, then one also must accept that a comprehensive understanding of the determinants of the quality of government cannot be attained without properlyintegrating all three of the traditional paradigms of comparative politics: rationality, culture and identity, and structure (read: institutions), per Lichbach and Zuckerman, 1997. Rational decision-making under the constraints imposed by institutions defines the realm of the possible, i.e., just how corrupt any of the feasible corruption regimes in a polity would end up being. On the other hand, culture and identity help determine whether a vicious or virtuous circle of corruption/probity takes hold in a polity, thereby determining which of the potential corruption regimes is actualized. Future research on corruption would do well to begin the hard work of systematically integrating these often competing analytical traditions into a coherent general theory of governance. 17 References Accinelli, E., and E.J.S. Carrera. 2012. “Corruption Driven by Imitative Behavior.” Economics Letters 117(1): 84-87. Andvig, J. C., and K. O. Moene. 1990. “How Corruption May Corrupt.” Journal of Economic Behavior & Organization 13(1): 63-76. Balafoutas, L. 2011. “Public Beliefs and Corruption in a Repeated Psychological Game.” Journal of Economic Behavior & Organization 78(1): 51-59. Becker, S., P.H. Egger and T. Seidel. 2009. “Common Political Culture: Evidence on Regional Corruption Contagion.” European Journal of Political Economy 25(3): 300-310. Bikhchandani, S., D. Hirshleifer and I. Welch. 1998. “Learning from the Behavior of Others: Conformity, Fads, and Informational Cascades.” Journal of Economic Perspectives 12(3): 151-170. Blair, G., and K. Imai. 2012. “Statistical Analysis of List Experiments.” Political Analysis 20(1): 47-77. Cadot, O. 1987. “Corruption as a Gamble.” Journal of Public Economics 33(2): 223-244. Carrell, S.E., F.V. Malmstrom and J.E. West. 2008. “Peer Effects in Academic Cheating.” Journal of Human Resources 43(1): 173-207. Caselli, F., and M Morelli. 2004. “Bad Politicians.” Journal of Public Economics 88(3): 759-782. Chong, A., A.L. De La O, D. Karlan et al. 2011. “Looking Beyond the Incumbent: The Effects of Exposing Corruption on Electoral Outcomes.” NBER Working Paper 17679. Cambridge, United States: National Bureau of Economic Research. Dal Bó, E., P. Dal Bó, and R. Di Tella. 2006. “‘Plata o Plomo?’: Bribe and Punishment in a Theory of Political Influence.” American Political Science Review 100(1): 41-53. Dong, B., and B. Torgler. 2012. “Corruption and Social Interaction: Evidence from China.” Journal of Policy Modeling 34(6): 932-947. Dong, B., U. Dulleck and B. Torgler. 2012. “Conditional Corruption.” Journal of Economic Psychology 33(3): 609-627. Ferraz, C., and F. Finan. 2008. “Exposing Corrupt Politicians: The Effects of Brazil’s Publicly Released Audits on Electoral Outcomes.” Quarterly Journal of Economics 123(2): 703745. Gingerich, D. W. 2009. “Ballot structure, political corruption, and the performance of proportional representation.” Journal of Theoretical Politics. 21(4): 509-541. Gingerich, D.W. 2010. “Understanding Off-the-Books Politics: Conducting Inference on the Determinants of Sensitive Behavior with Randomized Response Surveys.” Political Analysis 18: 349-380. 18 . 2013. Political Institutions and Party-Directed Corruption in South America: Stealing for the Team. Cambridge, United Kingdom: Cambridge University Press. . 2014. “When to Protect? Using the Crosswise Model to Integrate Protected and Direct Responses in Surveys of Sensitive Behavior.” Technical Note TN-702. Washington, DC, United States: Inter-American Development Bank. Glaeser, E., B. Sacerdote and J. Scheinkman. 1996. “Crime and Social Interactions.” Quarterly Journal of Economics 111(2): 507-548. Goel, R.K., and M.A. Nelson. 2007. “Are Corrupt Acts Contagious?: Evidence from the United States.” Journal of Policy Modeling 29(6): 839-850. Goel, R.K., and J.W. Saunoris. 2014. “Global Corruption and the Shadow Economy: Spatial Aspects.” Public Choice 161(1-2): 119-139. Hauk, E., and M. Sáez-Martí. 2002. “On the Cultural Transmission of Corruption.” Journal of Economic Theory 107(2): 311-335. Huang, P.H., and H.M. Wu. 1994. “More Order without More Law: A Theory of Social Norms and Organizational Cultures.” Journal of Law, Economics, and Organization 10(2): 390-406. Ichino, A., and G. Maggi. 2000. “Work Environment and Individual Background: Explaining Regional Shirking Differentials in a Large Italian Firm.” Quarterly Journal of Economics 115(3): 1057-1090. Innes, R., and A. Mitra. 2013. “Is Dishonesty Contagious?” Economic Inquiry 51(1): 722-734. Jann, B., J. Jerke and I. Krumpal. 2012. “Asking Sensitive Questions Using the Crosswise Model: An Experimental Survey Measuring Plagiarism.” Public Opinion Quarterly 76(1): 32-49. Kahan, D. M. 1997. “Social Influence, Social Meaning, and Deterrence.” Virginia Law Review, 349-395. Lichbach, M.I., and A.S. Zuckerman, editors. 1997. Comparative Politics: Rationality, Culture, and Structure. Cambridge: Cambridge University Press. López-Valcarcel, B.G., J.L. Jiménez and J. Perdiguero, J. 2014. “DANGER: LOCAL CORRUPTION IS CONTAGIOUS!”. Accessed February 11, 2015. http://www.alde.es/encuentros/anteriores/xviieea/trabajos/g/pdf/179.pdf/. Lui, F.T., 1986. “A Dynamic Model of Corruption Deterrence.” Journal of Public Economics 31(2): 215-236. Lupu, N. 2013. “Party Brands and Partisanship: Theory with Evidence from a Survey Experiment in Argentina.” American Journal of Political Science 57(1): 49-64. Malesky, E.J., D. Gueorguiev and N. M. Jensen. 2013. “Monopoly Money: Foreign Investment and Bribery in Vietnam, a Survey Experiment.” American Journal of Political Science. doi: 10.1111/ajps.12126. 19 Manzetti, L., and C.J. Wilson. 2007. “Why Do Corrupt Governments Maintain Public Support?” Comparative Political Studies 40(8): 949-970. Márquez, M.A., J. Salinas-Jiménez and M.D.M. Salinas-Jiménez. 2011. “Exploring Differences in Corruption: The Role of Neighboring Countries.” Journal of Economic Policy Reform 14(1): 11-19. Milgrom, P., and J. Roberts. 1990. “Rationalizability, Learning, and Equilibrium in Games with Strategic Complementarities.” Econometrica: Journal of the Econometric Society 58(6): 1255-1277. Mishra, A. (2006). “Persistence of Corruption: Some Theoretical Perspectives.” World Development 34(2): 349-358. Persson, A., B. Rothstein and J. Teorell. 2013. “Why Anticorruption Reforms Fail–Systemic Corruption as a Collective Action Problem.” Governance 26(3): 449-471. Pierce, L., and J. Snyder. 2008. “Ethical Spillovers in Firms: Evidence from Vehicle Emissions Testing.” Management Science 54(11): 1891-1903. Ryvkin, D., and D. Serra. 2012. “How Corruptible Are You? Bribery under Uncertainty.” Journal of Economic Behavior & Organization 81(2): 466-477. Sheffrin, S.M., and R.K. Triest. 1992. “Can Brute Force Deterrence Backfire? Perceptions and Attitudes in Taxpayer Compliance.” In: J. Slemrod, editor. Why People Pay Taxes: Tax Compliance and Enforcement. Ann Arbor, United States: University of Michigan Press: 193–218. Simpser, A. 2013. “The Intergenerational Persistence of Attitudes Toward Corruption.” Available at SSRN 2241295. Swamy, A., S Knack, Y Lee et al. 2003. “Gender and Corruption.” In: S. Knack, editor. Democracy, Governance and Growth. Ann Arbor, United States: University of Michigan Press. Tan, M.T., G.L. Tian and M.L. Tang. 2009. “Sample Surveys with Sensitive Questions: A Nonrandomized Response Approach.” American Statistician 63(1): 9-16. Tirole, J. 1996. “A Theory of Collective Reputations (with Applications to the Persistence of Corruption and to Firm Quality).” Review of Economic Studies 63(1): 1-22. Torgler, B., and N.T. Valev. 2006. “Corruption and Age.” Journal of Bioeconomics 8(2): 133-145. . 2010. “Gender and Public Attitudes Toward Corruption and Tax Evasion.” Contemporary Economic Policy 28(4): 554-568. Treisman, D. 2007. “What Have We Learned about the Causes of Corruption from Ten Years of Cross-National Empirical Research?” Annual Review of Political Science 10: 211-244. Winters, M.S., and R. Weitz-Shapiro. 2013. “Lacking Information or Condoning Corruption: When Do Voters Support Corrupt Politicians?” Comparative Politics 45(4): 418-436. 20 Yu, J.W., G.L. Tian and M.L. Tang. 2008. “Two New Models for Survey Sampling with Sensitive Characteristic: Design and Analysis.” Metrika 67(3): 251-263. 21 Appendix Table 4. Balance in Respondent Characteristics across Treatment Assignments Covariates Treatment groups corruption inefficiency control mean s.e. mean s.e. mean s.e. p.value Demographics Male 0.52 0.01 0.48 0.01 0.50 0.01 0.19 Age 38.1 0.42 38.6 0.42 38.3 0.42 0.72 Education 0.23 without study 0.01 0.00 0.01 0.00 0.01 0.00 primary incomplete 0.07 0.01 0.06 0.01 0.07 0.01 primary complete 0.19 0.01 0.20 0.01 0.22 0.01 secondary incomplete 0.27 0.01 0.28 0.01 0.28 0.01 secondary complete 0.24 0.01 0.24 0.01 0.23 0.01 technical studies incomplete 0.02 0.00 0.01 0.00 0.01 0.00 technical studies complete 0.03 0.00 0.02 0.00 0.03 0.00 university incomplete 0.09 0.01 0.09 0.01 0.07 0.01 university complete 0.08 0.01 0.09 0.01 0.07 0.01 post-graduate 0.01 0.00 0.01 0.00 0.01 0.00 Head of household 0.49 0.01 0.46 0.01 0.48 0.01 0.32 Cellphone 0.94 0.01 0.93 0.01 0.94 0.01 0.38 Laptop 0.43 0.01 0.41 0.01 0.42 0.01 0.60 Tablet 0.26 0.01 0.26 0.01 0.24 0.01 0.31 Car 0.39 0.01 0.37 0.01 0.39 0.01 0.34 Motorcycle 0.14 0.01 0.14 0.01 0.13 0.01 0.41 Plasma, LCD, or LED TV 0.56 0.01 0.53 0.01 0.55 0.01 0.26 Cable or Satellite TV 0.70 0.01 0.71 0.01 0.72 0.01 0.73 Internet 0.56 0.01 0.54 0.01 0.54 0.01 0.59 Costa Rican national 0.91 0.01 0.90 0.01 0.90 0.01 0.68 Prior beliefs In recent years, insecurity in the GAM has: 0.96 increased 0.72 0.01 0.71 0.01 0.73 0.01 decreased 0.05 0.01 0.05 0.01 0.05 0.01 stayed the same 0.23 0.01 0.24 0.01 0.22 0.01 In recent years, corruption in Costa Rica has: 0.45 increased 0.83 0.01 0.84 0.01 0.83 0.01 decreased 0.01 0.00 0.02 0.00 0.02 0.00 stayed the same 0.16 0.01 0.14 0.01 0.16 0.01 Of all the cases that enter the legal system, how many do You think are resolved? 0.38 the majority 0.08 0.01 0.08 0.01 0.07 0.01 many 0.09 0.01 0.08 0.01 0.08 0.01 few 0.37 0.01 0.41 0.01 0.38 0.01 very few 0.41 0.01 0.38 0.01 0.41 0.01 none 0.05 0.01 0.05 0.01 0.05 0.01 Prior experiences Direct contact with police or transit officer in previous year 0.25 0.01 0.26 0.01 0.26 0.01 0.54 Knows personally a police officer 0.42 0.01 0.41 0.01 0.41 0.01 0.87 Bribe solicited by police officer in previous year 0.03 0.00 0.03 0.00 0.03 0.00 0.81 Knows personally someone accused, prosecuted, or Sentenced by the criminal justice system 0.42 0.01 0.40 0.01 0.41 0.01 0.74 Note: P-values based on chi-square test of homogeneity for categorical covariates and an ANOVA test for equality of group means for continuous covariates (e.g., age). 28