216 Abstract Public corruption has received significant attention from scholars across a large array of disciplines. Despite the numerous academic efforts, however, our current understandings of the motivation and the evolution of public cor- ruption still suffer from an agonizing amount of ambiguity, and on many occasions fail to provide practical guidance. One of the reasons for the limited progress in the area is the fact that, given its nature, corruption does not lend itself easily to empirical studies. Especially rare is qualitative research that examines and challenges the find- ings of previous studies within different institu- tional, cultural and social environments. This article provides the results of a study that examined nine theory building propositions suggested by de Graaf and Huberts (2008) within the context of developing democracies, specifically the Republic of Moldova. The study confirmed six of the nine propositions. Among others, it was determined that in developing de- mocracies corruption is highly influenced by the social matrix and public officials often assume their positions with the implicit expectations of reaping rewards from corrupt behaviors. Keywords: corruption, public sector, Moldo- va, social matrix. REFOCUSING PERSPECTIVES ON PUBLIC CORRUPTION AWAY FROM THE INDIVIDUAL: INSIGHTS FROM THE MOLDOVAN SOCIAL MATRIX Alexandru Vasile ROMAN Alexandru Vasile ROMAN Assistant Professor of Public Administration, College of Business and Public Administration, California State University, San Bernardino, United States of America Tel.: 0001-909-537.5762 E-mail: aroman@csusb.edu Transylvanian Review of Administrative Sciences, No. 43 E/2014, pp. 216-231 217 1. Introduction The nature of corruption makes research in the area challenging (Williams, 2000; de Graaf and Huberts, 2008) as what corruption is and how it is defined will depend on the context or one’s theoretical perspective. The latter, however, becomes particu- larly consequential since the proposed solutions, more often than not, will be in large part the byproduct of the original assumptions. Scholars (Rose-Ackerman, 1999; John- ston, 2001; von Alemann, 2004) have argued that our understandings and definitions of corruption are diachronically dependent and are routinely satiated with norma- tive connotations and beliefs. According to Pardo (2004, p. 2), ‘in any given society, corruption is a changing phenomenon, some of its aspects and received morality are culturally specific and its conceptualization is affected by personal interest, cultural values and socio-economic status’. Jain (1998; 2001) has suggested that many of the failures in terms of generating practical solutions, but also the fragmentation in the extant body of literature on corruption, can be attributed to the presence of manifold active definitions and understandings. The study of public corruption is also exposed to debates over causal associations and mechanisms. For example, a corrupt bureaucratic system can be either the cause or the outcome of corrupt behavior; it is essentially impossible to empirically identify which is the first condition to develop. Much of this debate remains to be resolved at the philosophical and theoretical levels and is beyond the scope of this paper. Yet, within this vein it is important to note, that there is a troublesome tendency within the more recent body of literature to attribute perhaps too much decision-making power to the individual. Researchers habitually assume, especially those who embrace ra- tional choice perspectives, that the decision to become corrupt is shaped solely at the individual level (Persson, Rothstein and Teorell, 2013). The principal-agent model, typically invoked in such studies, has also been challenged on the grounds that it misses multiple critical dimensions of corruption (Kahana and Qijun, 2010). Further- more, an important stream of academic literature subtly enforces overly-simplistic dichotomies of ‘us versus the other’ (Haller and Shore, 2005) and accepts material or power interests as the primary motivators of corrupt behavior. While, assuming that the individual has full control over the decision of whether or not to undertake a corrupt act, does indeed reduce some of the theoretical complex- ity associated with the study of corruption – such supposition, however, also imposes strict constraints on the broader dynamics that the researcher can ultimately uncover. Attempts to enforce stereotypical dichotomies are vulnerable to similar criticisms; la- beling behavior within rigid categories is a way of observing some things, but also a way of overlooking others. Notwithstanding the difficulties associated with research on public corruption, scholars appear to agree that developing practical interpreta- tions about the nature of corruption and unethical behavior is critical for achieving improved democratic governance (Menzel, 2009). In a recent effort to address the inability of the extant body of literature to adequate- ly explain public corruption and the lack of a convincing theoretical framework, de 218 Graaf and Huberts (2008) undertook a qualitative investigation of the nature of corrup- tion in Western democracies. In their study of corruption in Netherlands, the authors studied in detail 10 important public corruption cases and offered nine propositions. The Dutch researchers concluded that corruption develops within enduring relation- ships and the process of becoming corrupt can be described as a slippery slope. While the usual suspect, material gain, was part of the motivational context, their research also recognized that corruption could be motivated by love, friendship, status, and the desire to impress others. The conclusions reached in the Dutch study were intriguing, but their claim to generalizability was limited since the suggestions were framed with- in the context of economically developed countries with established democracies. Given the potential of the nine propositions in terms of generating a powerful the- oretical framework, this study examined them within a different environment, treat- ing them as hypotheses. The research presented here attempted to determine whether the nine propositions derived from the Dutch study could be transferred and upheld within the context of developing countries of Eastern Europe, which are inexperi- enced democracies. The methodological approach was reproduced in studying 28 Moldovan corruption cases. Different from the original study, however, 33 additional interviews were conducted. The interviews were found to be necessary in order to gauge the nuances of the social and cultural structures within the new setting. The findings confirmed six of the nine original propositions. This provides hope that a practical and encompassing understanding of the motivational constructs be- hind public corruption can indeed be achieved. The three propositions that were not confirmed, nevertheless, suggest that the de Graaf and Huberts’ (2008) framework somewhat underestimates the role of the social context, at least when applied to de- veloping democracies. Consequently, by underestimating the effect of social associa- tions and kinship pressures, the framework empowers the individual decision-maker beyond the pragmatic reality of his or her social environment. This paper, then, makes a two-fold contribution to current public administration literature. First, it is one of a small number of studies on public corruption that exam- ines, challenges and attempts to confirm propositions suggested as a result of a pre- vious qualitative study. Second, the research is sufficiently generalizable in character in order to be applicable and of academic, and practical interest for established and fledgling democracies alike. Taken together, given the methodological variation from the original study, the research presented here should be treated both as confirmatory and exploratory in nature. Following this introduction the discussion is broken down into three main sec- tions. In the next section I introduce the de Graaf and Huberts (2008) study and their methodology, and explain why replicating their research design was found to be ap- propriate and theoretically justifiable. In the section that follows, I present the results of the study on the background of the original nine propositions. The discussion fo- cuses mainly on the propositions that were not confirmed. In the final section, the broader implications of the findings are further discussed. 219 2. From Dutch to Moldovan: an overview of the methodological approach The difficulty that is associated with the attempts to frame the nature of public corruption and the limited knowledge about causal relationships, and their directions – make studying corruption highly complex. The normative overtones and the implic- it assumptions behind what corruption is dictate what is being examined and what is eventually being offered as a policy solution (Anechiarico and Jacobs, 1996; Philp, 1997; Johnston, 2001). Presently, the literature does not offer one single or reliable indicator that would provide a substantive anticipation of corrupt behavior. Whilst quantitative examinations have their place in this field of study, in instances when the researcher would like to uncover dynamics unfettered by theoretical assumptions, it is probably appropriate to undertake a qualitatively-heuristic approach to research (Menzel, 2003). Qualitative methodologies are often found more useful than quanti- tative approaches in terms of their capacity to develop understandings of corruption (Doig, 2011) and in terms of providing improved policy suggestions (Anechiarico and Jacobs, 1996). In their research, de Graaf and Huberts (2008) employed a multiple case study de- sign within an explorative and inductive research strategy (Höffling, 2002), in order to focus on a specific setting (Yin, 2009) and generate theory through propositions (Har- ris and Sutton, 1986). The researchers selected and studied in great detail 10 important corruption cases from the Netherlands. The complex nature of corruption and the lack of academic agreement on the topic made an explorative inductive research approach fitting for deconstructing the character of the public officials and the dimensions of the motivators behind their corrupt behavior. The cases were analyzed within their con- texts and following the frameworks offered by Anechiarico and Jacobs (1996), Della Porta and Vannucci (1997), and Höffling (2002). For the purposes of the original study, to be corrupt, as per Huberts and Nelen (2005), meant to act (or fail to act) as a result of receiving personal rewards from inter- ested third parties. The original exploration focused specifically on the Netherlands, a high-income country, where it is commonly believed that corrupt behavior in the pub- lic sector is the exception rather than the norm. In this sense, the nine propositions that were suggested by de Graaf and Huberts (2008), as a result of their analysis, should be interpreted within the situational context of Western democracies, with developed economies and stable institutionalized political systems. While powerful in their own right, examining their conclusions within a different setting would provide a valid test for the robustness and transferability potential of the nine propositions. The Republic of Moldova was identified as an appropriate environment for repli- cating the study. The choice was driven by two main considerations. First, Moldova is sufficiently different1 in terms of economic development, historical traditions and 1 Moldova is commonly considered a country with high levels of corruption (Spinei et. al, 2008) – for instance in 2011 it was ranked as the 112th least corrupt country in the world, 220 cultural structures to provide a challenging test for the propositions generated in the Dutch study. Yet, at the same time, the institutional structures were believed to be sufficiently similar to warrant employing an analogous research design. Second, this represented a convenience sample; it was the only research setting, where it was pos- sible to assure the needed richness of information and level of access. A total of 28 complete corruption cases, from 2003 to 2009, were made available by the Supreme Court of Justice of Republic of Moldova (SCJRM, 2010) and by the Center for Combating Economic Crimes and Corruption (CCECC, 2010). Each file provided comprehensive information on the investigations and court proceedings, while on a number of occasions the files also contained audio or video recordings. Additional data, whenever possible, were collected from other publicly available sources, which most of the time were Internet-based. The cases were examined following the meth- odological format employed by de Graaf and Huberts (2008) as closely as reasonably possible. Multiple data sources were iteratively reviewed and triangulated in order to reconstruct an individual profile for each of the corrupt public officials and to under- stand the motivational framework for engaging in the specific corrupt act. Differently from the original study, however, an additional 33 interviews with business owners (five), current (four) and former (three) anti-corruption agencies’ em- ployees, fiscal/judicial/custom employees (five), journalists (eight) and other public officials (eight) – were conducted. The interviews with the informants were conducted in person in December of 2010. The interviews were motivated by the intention to provide the data necessary to evaluate the cultural dimensions of corruption, but also to assist with interpretations of uncovered patterns. The interviews were semi-structured in nature and were se- lected through a snowball approach. Although the informants were allowed to dis- cuss any particular matter that they found important to be shared, they were direct- ed to focus on describing existing social perceptions of corruption. They were also encouraged to frame their insights within stories. The latter were found to provide a great number of thick and useful descriptions and understandings regarding the inner fundamentals of the Moldovan social matrix. Interviews were stopped when it was believed that a theoretical saturation was achieved, for which even sets less than 50 interviews can frequently suffice (Strauss, 1987; Ritchie, Lewis and Elam, 2003). A number of the informants were also intimately familiar with some of the cases being examined; as such, they provided an additional useful source that complemented the information derived from the investigation files. sharing the same ranking with Algeria, Egypt, Kosovo, Senegal and Vietnam (Transparency International, 2012). In addition, the nation has been experimenting with democracy for less than three decades. 221 3. Findings 3.1. Confirmations By focusing on predisposing causes of corruption, de Graaf and Huberts (2008) determine that next to material gain, friendship, love, status2 and attempts to impress others, are the most significant motives for corrupt behavior. Their results also sug- gest that the process of becoming corrupt is a slippery slope and as a rule the individ- ual becomes entrapped by developing dynamics, which lead to limited self-awareness or control, hence increased levels of corrupt behavior. These findings are intriguing in several important ways. First, the results lend support to the argument that in order to properly delineate behavior the researchers should rely on both individual and societal lenses (Vardi and Weitz, 2004; de Graaf, 2007). Interpreting corruption from one single perspective can easily lead to the re-en- forcement of simplistic and banal beliefs about corruption motivating mechanisms. Second, de Graaf and Huberts (2008) convincingly argue that anticorruption policies have to be flexible in order to be effective as context is found to be critically important in terms of evaluating and managing corruption. Specific circumstances will demand a different mix of anticorruption controls and ethical constructs (de Graaf and Hu- berts, 2008). Finally, and most importantly, given the methodological approach, the Dutch study provides the basis for developing theoretical insights about the nature of corruption unconstrained by over-simplistic assumptions about the individual, orga- nizational or societal constructs. The patterns that were uncovered as a result of the in-depth analysis of 28 Moldo- van cases supported six of the nine propositions put forward as a result of the original study. Table 1 provides a summary description of the nine leading propositions by de Graaf and Huberts (2008) and their confirmation. Appendix 13 provides a summary description of the coding for each Moldovan corruption case. The aggregate analysis of the Moldovan cases juxtaposed with content analysis of the interviews confirmed that the relationships between the interested third parties and the corrupt public officials are typically lasting and constructed based on mutual trust (proposition eight). One of the suspected public officials was quoted in one of 2 When it is argued that corruption was motivated by status it is meant that one engaged in corruption due to one’s belief that one’s social position excused or allowed such behavior. In bland terms, one’s position made one feel untouchable and supposedly gave one prefer- ential treatment under the law. 3 In the analysis, one’s personality was coded based on the deconstruction of one’s answers and dialogues during integration, one’s court statements and evaluation provided in one’s fi le, as well, whenever present, the characterizations off ered by one’s co-workers. Although, as in de Graaf and Huberts (2008), personality labeling was unstandardized, the data was suffi ciently thick and rich to provide the appropriate basis to identify one’s att itude and gen- eral personality traits. On multiple occasions the evaluation by investigators also provided solid insights into one’s personality. 222 Table 1: de Graaf and Huberts propositions (2008) The Individual Motive 1. Next to material gain the most important motives for offi cials to become corrupt are friendship or love, status, and making an impression on colleagues and friends Process of Becoming Corrupt 2. Offi cials ‘slide down’ toward corruption; most processes of becoming corrupt can be considered a slippery slope. Character 3. Often corrupt offi cials have dominant and strong personalities, know how to ‘get things done’, take or get the freedom to do things independently, and oversee formal boundaries of authority. [confi rmed4] Space to Maneuver 4. More ‘business type’ public offi cials, brings (sic) the risk of more corruption. [confi rmed] The Organization Supervision 5. In most corruption cases, supervision of the corrupt offi cial is not strong. [con-fi rmed] Organizational Structure 6. In most corruption cases, management has not promoted a clear integrity policy. [confi rmed] Organizational Culture 7. Because of loyalty and solidarity, colleagues are hesitant to report suspicions of another’s corrupt activities. The Relationship between the Briber and the Corrupt Offi cial The Exchange of Favors; the Length, Initiative, and Nature of the Relation- ship; and the Interest of Bribers 8. The relationship between briber and offi cial is most often enduring. [confi rmed] Corrupt Networks 9. Corrupt offi cials, including those who operate outside-called corrupt networks, do not limit their corruption to one incident. [confi rmed] the case files stating: ‘I didn’t do favors just for anyone. I usually helped or dealt only with people whom I trusted. Most of the time they were guys I knew from school. For me it was very important that I could depend on them. When it came to big staff, I normally had two guys that I went through…no one else. We were on the same page in terms of what needed to be done’.4 Corrupt behavior is habitually repetitive/not limited to one incident (proposition nine). In one of the files, for instance, the investigation team identified the following corruption pattern: ‘there have been at least four other known instances in which the suspect has engaged in unlawful activities. In the first two months of 2008, the suspect is believed to have subtracted 3000 LEI [approximately $240] from four different wholesalers in exchange for favorable inspections. A clear and lasting pattern of unlawful acts can be established. The suspect had a well-known system for receiving fees for inspections from wholesalers’. The latter dynamic is regularly aided by weak organizational supervision and fail- ures to institute policies that would otherwise uphold and enforce ethical behavior (propositions five and six). During one of the interviews, a long time anti-corruption agent stated: ‘one huge issue in Moldova is the fact that there is total disinterest in fighting corruption. Moral behavior is not something high on the priority list for public agencies. High ranking state employees are not interested in developing or enforcing strict moral codes. It is against their interest to promote such policies. It makes doing things on the side more difficult. But why would anyone want to do that? It makes no sense for them to do that’. 4 Here ‘confi rmed’ means that the data from the Moldovan cases supported the proposition made by de Graaf and Huberts (2008). In other terms, Moldovan data warranted reaching similar conclusions about the nature and triggers of corrupt behavior. 223 The Moldovan cases confirmed that public officials, who eventually undertake cor- rupt behavior, can be normally described as having strong personalities (proposition three). They were able to claim ample discretionary spaces to maneuver within their organizations, which with time increased the risk of corruption (proposition four). One of the informants, an anti-corruption investigator, concluded: ‘if I would have to choose one specific thing that characterized most of the corrupt state workers whom I have en- countered over the years, it would be arrogance. They are usually domineering and feel self-im- portant. They truly believe, most of the time for good reason, that nobody can do anything to them. People are either afraid of them or just don’t care enough to do anything. As time goes by, they assume more and more liberties. They don’t like to answer to anyone and can get nasty when challenged. The longer they do what they are doing and the higher they climb, the more powerful and entitled they become. It is actually not easy [to be so dishonest] … you have to be a different kind of man to be able to do some of the things that I have seen’. While the confirmation of six of the nine propositions is a commanding outcome that warrants significant attention; the dynamics behind the three propositions that were not confirmed are perhaps even more theoretically intriguing. The next sub-sec- tion introduces and discusses the three propositions that were not confirmed. 3.2. Differences The analysis of the 28 cases did not provide the grounds for a complete and un- challenged acceptance of propositions one, two and seven. Within the Moldovan context, neither friendship nor love was confirmed as important motives that lead public officials to become corrupt. In fact, no case revealed love or friendship as the primary reason for engaging in corrupt acts (proposition one). Next to material gain, kinship associations join status as the two most important motives for corrupt behav- ior. It was determined that in ten cases kinship pressures were the main trigger of the specific corrupt act while in another five cases, public officials acted corruptly due primarily to status considerations. One of the informants, a high ranking government official, elegantly captured the condition during one of the interviews: ‘our people have an obsession with power and status. They always chase it. They want more, and more and more… Moldovans have this unexplainable dumb definition of democracy. For Moldovans democracy means doing anything that you want. There is no respect for the law. Everybody is in it for themselves and for their clans. Communism has killed any sense of community and now all everyone wants is to feel important. All people want is to take care of their families and their relatives. People say that it is always about money. That’s not true. Yes, money is very important, but I can tell you for a fact that it is not always true. Many times our officials want to show off that they can do anything that they want. In such cases corruption has noth- ing to do with money. It is just about showing off. It’s Moldovenism through and through…a disease’. Most critically, however, unlike Dutch public officials, Moldovan public officials often come into their positions already corruptly-predisposed; they expect and de- mand status and benefits that would, on paper, not be sanctioned for holders of public 224 office (proposition two). An interrogation transcript from one of the files reveals the following exchange: ‘Investigator: You don’t think there is anything wrong with what you did? Suspected official (S.O.): No, of course not. I only did what everyone else does and I only took what is rightfully mine. Investigator: Rightfully yours? S.O.: Yes. It comes with the job. Sometimes you do things on the side. You should know this better than me. It is not like anyone in this country actually lives on salary alone. Investigator: I do. S.O.: No. You don’t. I guarantee you don’t. Investigator: Why are you so sure? S.O.: Because you work for the state and nobody can live on a state salary. People only work for the state because they can make money off the books or they want to get to power. That’s the only reason anyone wants a state job’. Explaining misbehavior as a normal expectation of one’s position represented a common occurrence within the examined cases. Indeed, the possibility of command- ing additional, job unsanctioned benefits, was a pervasive theme with the motivation- al structures for corrupt public officials. Finally, different from the Dutch cases, neither individual nor organizational loy- alty could be confirmed as an inhibitor of the willingness of other public employees to report suspicious behavior (proposition seven). On the contrary, minimal levels of organizational identification and dominance of kinship associations made reporting of suspicious behavior – which one would expect to be an organizational priority – unimportant. Public officials rarely, if ever, would identify with their coworkers or organizations. The latter, for instance, is well captured in the following dialogue from one of the interrogation transcripts: ‘Investigator: You have worked here for five years, right? Suspected official (S.O.): Yes, almost six. Investigator: You don’t seem too upset that you have lost your job. S.O.: Why should I be? The mayoralty didn’t do much for me. It was a job. I got what I could while I could’. Table 2 delineates the three unconfirmed propositions and their proposed alterna- tive interpretations. There are several consequential theoretical implications that are bound to come with these findings. First, within the context of post-communist countries kinship and other social associations might rival and at times might even dominate material in- terest (Karklins, 2005; Barsukova, 2009; Roman and Miller, 2014). In more than half of the cases in the data set, other motives than material interest acted as predisposing causes for corrupt behavior (Appendix 1). Along these lines, the nature of the material interest-corruption nexus within low income, developing democracies might not be 225 Table 2: Proposed alternatives to unconfirmed propositions Original Propositions Alternative Propositions The Individual Motive Next to material gain the most important motives for of- fi cials to become corrupt are friendship or love, status, and making an impression on colleagues and friends. Next to material gain the most important motives for offi cials to become corrupt are kinship and status. Process of Becoming Corrupt Offi cials slide down toward corruption; most process- es of becoming corrupt can be considered a slippery slope. Offi cials come into a position expecting the status and benefi ts outside their salary, the offi cials are corrupt oriented even before taking a position. The Organization Organizational Culture Because of loyalty and solidarity, colleagues are hesi- tant to report suspicions of another’s corrupt activities. Colleagues are not interested in reporting suspicions of anoth- er’s corrupt activities for a variety of intertwined reasons having to do with culture, lack of organizational loyalty, unclear reporting procedures, unwillingness to get involved and predominance of corruption in the organization. significantly different from those in high income, established democracies. Under this scenario, the expectations that better economic conditions would lead to less corrup- tion (e.g., Treisman, 2000; Paldam, 2001; Fishman and Gatti, 2002; Rauch and Evans, 2000; Gundlach and Paldam, 2009; Blackburn and Forgues-Puccio, 2010) become the- oretically strenuous. Corruption might indeed be less of an economic problem than it is commonly assumed. Second, within studies of corruption in fledgling democracies with developing economies, it is probably ill-advised to make any assumptions about the normative expectations, values and behaviors of individuals upon entry into public service. The process of becoming corrupt can certainly be described as a slippery slope. The slope, in developing countries, however, is probably much steeper and has a different start- ing point than it is the case of established democracies, with developed economies and time-honored institutional structures. As such, it is inappropriate to assume that most public officials become corrupt on the job. This latter realization induces a great deal of challenges since few organizational controls are powerful enough to account for the behavioral predispositions of the individual upon entry into public service. There is a whole new set of lessons and dynamics, outside decision-making at the individual level, which researchers have to learn in terms of managing public corrup- tion. If the individual enters public service already corruptly-orientated, we could ex- pect that integrity control structures will be perceived as nothing more than a minute inconvenience. In developing democracies, the critical question of managing public corruption, then, will not be that of guiding individual decision-making but of that of reconstructing the broader telos behind public service employment. Thirdly, while the findings suggest substantial levels of commonality among the two settings, hence possible theoretical transferability, in terms of policy design, pub- lic corruption in low income countries might demand approaches that differ in ways that matter from anticorruption policies in established democracies. I have previously 226 argued that boiler plate anticorruption approaches, that assume that public corruption in developing countries is driven by poverty and material gain, have been rather inef- fective and might in fact backfired by motivating more corruption (Roman, 2012). In fledgling democracies, institutional and social constructs are not well behaved nor do they lend themselves to rational predictability (Persson, Rothstein and Teorell, 2013; Ledeneva, 2011). McCarthy (2010), for instance, has suggested that the failures and in- adequacies of institutional structures are much more consequential in terms of explain- ing public corruption than it is led to believe. In Post-Soviet democracies it is common for anticorruption and law-enforcing agencies, which are critical for enforcing anticor- ruption policies, to be some of the most corrupt organizations (Barsukova, 2009; Lede- neva, 2011). As a result, public corruption in low income countries often remains un- responsive and unaffected by policies imported from established democracies, which take for granted the integrity and effectiveness of local anticorruption agencies. Finally, there are valid reasons to believe that corruption is a self-generating and self-fulfilling condition. Meaning that corrupt behavior in the public sector might si- multaneously be the outcome and the cause of specific circumstantial developments within the individual-social environment nexus. The latter does not mean that any directional assumption about causality becomes worthless or theoretically inappro- priate; it merely means that if simplifying suppositions should be made, they should be based on well-constructed contextual understandings of the condition. 3.3. Limitations At this point, it is central to note that there are four, possibly critical, weaknesses in term of the research approach that could endanger the validity of the confirmatory efforts. First, the Dutch cases were considered important. The Moldovan cases were not necessarily important in the same manner that the Dutch cases were; furthermore, their selection was based on availability/convenience and not on discretionary aca- demic judgment. In fact, only two of the 28 studied cases could have been considered as important. Thus, in general lines, the original study dealt with salient or elite cases whereas in the second setting the data was a mix of cases primarily reflecting petty or routine corruption. There was an obvious bias in the type of corruption cases being prosecuted in Moldova, which in itself represented a very powerful realization and should be investigated by further research. Second, while in terms of institutional structure and stated legal processes, in the- ory, the two countries might appear rather similar, in reality the two settings might respond dramatically different to an identical set of conditions. Third, as it is the case for all research efforts of this type, the final interpretations are as much the reflection of the researcher’s training as of the methodological ap- proaches being employed. Even the earnest of efforts and the most exacting research designs cannot properly remove all traces of bias of the unstated and unrealized as- sumptions that the researcher possesses. This analysis is certainly not immune to that. I cannot nor do I wish to claim such exactness. 227 Lastly, the methodological approach was reproduced to the extent made possible by the detail provided in the original article. While, all sources used by the Dutch researchers were consulted closely, it is not excluded that the methodology used in studying the Moldovan cases might not have been an exact replication. Nevertheless, despite these possibly significant vulnerabilities, I find sufficient grounds to believe that the comparison of the findings under the two contexts is war- ranted and provides meaningful insights into expanding our understandings about triggers and motivators of public corruption. Although the differences between the two environments and the nature of the data might raise some concerns, passing such a demanding test would provide increased validity to the original suggestions. 4. Conclusions de Graaf and Huberts note that ‘in the literature on corruption, there is much spec- ulation on its nature, but there are hardly any empirical qualitative studies on the nature of corruption’ (2008, p. 650). If it accomplished anything, this study at a min- imum provides a rare examination and confirmation effort of previous qualitative empirical research. Under the best case scenario the findings presented here laid the grounds for a practical and generalizable theoretical framework that is driven by heu- ristic logic and makes few simplifying assumptions. The successful confirmation of six of the nine propositions gives hope that it might indeed be possible to generate a satisfactory understanding of the nature of corruption. Yet, the three unconfirmed dimensions induce a necessary level of realism to such enthusiasm by suggesting that much remains to be done. Due to the methodological difficulties of confirming qualitative research and the inherent weaknesses of explorative efforts, the conclusions reached in this article should be treated with an adequate level of academic skepticism. Having in mind the potential imitations of the study, nevertheless, the richness of the data and the methodological rigor provide acceptable levels of confidence in order to believe that the suggestions are indeed robust. Paradoxically, the two main possible weaknesses of this study, that are associated with reproducing the research within such a dramati- cally different environment, could, in some sense, be considered the primary strengths of this scholarly effort. The original propositions were confirmed not only within the context of a different environment, but also within a data set that was primarily com- posed of petty corruption cases. The main contribution of this research, then, rests in its efforts to confirm and ex- pand our knowledge about the nature of public corruption. While the profile and nature of the individuals are confirmed as important in constructing the predisposing causes for corrupt behavior, I argue that in low income developing democracies the social matrix is more powerful in shaping individual behavioral dynamics than the other way around. Curiously, when de Tocqueville (1835/1840/2001) first introduced the term individualism, he used it to describe the selfishness with which one pursues the interests of self and of a limited circle of family and friends. Behavior is influenced 228 by social ties well before the individual becomes a public servant. It is theoretically convenient to assume that public servants leave their social identity at ‘the organi- zational door’, but this is clearly rarely the case. While the role played by the social pressures might be more evident in inexperienced democracies, it is still undoubtedly present in established democracies as well. The individual shapes his or her organi- zational environment as well as his or her social associations; the social matrix, none- theless, might ultimately dictate and establish the criteria under which those causal interactions take place. References: 1. Anechiarico, F. and Jacobs, J.B., The Pursuit of Absolute Integrity: How Corruption Control Make Government Ineffective, Chicago: University of Chicago Press, 1996. 2. Barsukova, S., ‘Corruption’, 2009, Russian Politics and Law, vol. 47, no. 4, pp. 8-27. 3. Blackburn, K. and Forgues-Puccio, G.F., ‘Financial Liberalization, Bureaucratic Cor- ruption and Economic Development’, 2010, Journal of International Money and Finance, vol. 29, no 7, pp. 1321-1339. 4. Center for Combating Economic Crime and Corruption, 2010, [Online] available at http://en.cccec.md/, accessed on October 5, 2010. 5. de Graaf, G., ‘Causes of Corruption: Towards a Contextual Theory of Corruption’, 2007, Public Administration Quarterly, vol. 31, no. 1, pp. 39-86. 6. de Graaf, G. and Huberts, L.W.J.C., ‘Portraying the Nature of Corruption Using an Explorative Case Study Design’, 2008, Public Administration Review, vol. 68, no. 4, pp. 640-651. 7. de Tocqueville, A., Democracy in America, New York, NY: Penguin, 1835/1840/2001. 8. Della Porta, D. and Vannucci, A., ‘The Resources of Corruption: Some Reflections from the Italian Case’, 1997, Crime, Law and Social Change, vol. 27, no. 3-4, pp. 231-254. 9. Doig, A., ‘Numbers, Words and KYC: Knowing Your Country and Knowing Your Cor- ruption’, 2011, Critical Perspectives on International Business, vol. 7, no. 2, pp. 142-158. 10. Fishman, R. and Gatti, R., ‘Decentralization and Corruption: Evidence across Coun- tries’, 2002, Journal of Public Economics, vol. 83, no. 3, pp. 325-345. 11. Gundlach, E. and Paldam, M., ‘The Transition of Corruption: From Poverty to Hones- ty’, 2009, Economics Letters, vol. 103, no. 3, pp. 146-148. 12. Haller, D. and Shore, C., Corruption: Anthropological Perspectives, London: Pluto Press, 2005. 13. Harris, S.G. and Sutton, I.R., ‘Functions of Parting Ceremonies in Dying Organiza- tions’, 1986, Academy of Management Journal, vol. 29, no. 1, pp. 5-30. 14. Höffling, C., Korruption als Soziale Beziehung (Corruption as Social Ties) Opladen: Lseke Budrich Verlag, 2002. 15. Huberts, L.W.J.C. and Nelen, J.M., Corruptie in het Nederlandse Openbaar Bestuur: Om- vang, aard en Afdoening (Corruption in Government in the Netherlands: Extent, Nature and Prosecution), Utrecht: Lemma, 2005. 16. Jain, A.K. (ed.), Economics of Corruption, Boston: Kluwer Academic Publishers, 1998. 17. Jain, A.K., ‘Corruption: A Review’, Journal of Economic Surveys, 2001, vol. 15, no. 1, pp. 71-121. 18. Johnston, M., ‘The Definitions Debate: Good Politics, Governance, and the Concept of Corruption’, in Jain, A.K. (ed.), Political Economy of Corruption, London: Routledge, 2001, pp. 11-32. 229 19. Kahana, N. and Qijun, L., ‘Endemic Corruption’, 2010, European Journal of Political Econ- omy, vol. 26, no. 1, pp. 82-88. 20. Karklins, R., The System Made Me Do It: Corruption in Post-Communist Societies, Armonk, NY: M.E. Sharpe, 2005. 21. Ledeneva, A., ‘Open Secrets and Knowing Smiles’, 2011, East European Politics and So- cieties, vol. 25, no. 4, pp. 720-736. 22. McCarthy, L.A., ‘Beyond Corruption: An Assessment of Russian Law Enforcement’s Fight against Human Trafficking’, 2010, Demokratizatsiya: The Journal of Post-Soviet De- mocratization, vol. 18, no. 1, pp. 5-27. 23. Menzel, D.C., ‘State of the Art of Empirical Research on Ethics and Integrity in Gov- ernance’, in Frederickson, G.H. and Ghere, R.K. (eds.), Ethics in Public Management, Armonk, NY: M.E. Sharpe, 2003, pp. 16-48. 24. Menzel, D.C., ‘In Pursuit of Ethical Governance’, 2009, The Public Manager, vol. 38, no. 2, pp. 30-35. 25. Paldam, M., ‘Corruption and Religion Adding to the Economic Model’, 2001, Kyklos, vol. 54, no. 2-3, pp. 383-413. 26. Pardo, I. (ed.), Between Morality and the Law: Corruption, Anthropology and Comparative Society, Aldershot: Ashgate, 2004. 27. Persson, A., Rothstein, B. and Teorell, J., ‘Why Anticorruption Reforms Fail – Systemic Corruption as a Collective Action Problem’, 2013, Governance: An International Journal of Policy, Administration, and Institutions, vol. 26, no. 3, pp. 449-471. 28. Philp, M., ‘Defining Political Corruption’, 1997, Political Studies, vol. 45, no. 3, pp. 436- 462. 29. Rauch, J.E. and Evans, P.B., ‘Bureaucratic Structure and Bureaucratic Performance in Less Developed Countries’, 2000, Journal of Public Economics, vol. 75, no. 1, pp. 49-71. 30. Ritchie, J., Lewis, J. and Elam, G., ‘Designing and Selecting Samples’, in Ritchie, J. and Lewis, J. (eds.), Qualitative Research Practice: A Guide for Social Science Students and Re- searchers, London: Sage Publications, 2003, pp. 77-108. 31. Roman, A.V., ‘The Myths within Anticorruption Policies’, 2012, Administrative Theory & Praxis, vol. 34, no. 2, pp. 237-254. 32. Roman, A.V. and Miller, H.T., ‘Building Social Cohesion: Family, Friends, and Corrup- tion’, 2014, Administration & Society, vol. 46, no. 7, pp. 775-795. 33. Rose-Ackerman, S., Corruption and Government: Causes, Consequences, and Reform, New York: Cambridge University Press, 1999. 34. Spinei, I., Carașciuc, L., Obreia, E. and Lazari, C., Măsurăm Corupția: De la Sondaj la Sondaj (Measuring Corruption: From Survey to Survey), 2008, Chisinau, Republic of Moldova: Transparency International. [Online] available at http://www.transparency. md/content/blogcategory/16/48/4/8/lang,en/, accessed on October 1, 2010. 35. Strauss, A.L., Qualitative Analysis for Social Scientists, Cambridge, UK: Cambridge Uni- versity Press, 1987. 36. Supreme Court of Justice of Republic of Moldova, 2010, [Online] available at http:// www.csj.md/index.php?lang=4, accessed on October 5, 2010. 37. Transparency International, Corruption Perception Index 2011, 2012, [Online] available at http://cpi.transparency.org/cpi2011/, accessed on March 15, 2012. 38. Treisman, D., ‘The Causes of Corruption: A Cross-National Study’, 2000, Journal of Pub- lic Economics, vol. 76, no. 3, pp. 399-457. 230 39. Vardi, Y. and Weitz, E., Misbehavior in Organizations: Theory, Research and Management, Mahwah, NJ: Lawrence Erlbaum, 2004. 40. von Alemann, U., ‘The Unknown Depths of Political Theory: The Case for a Multidi- mensional Concept of Corruption’, 2004, Crime, Law and Social Change, vol. 42, no. 1, pp. 25-34. 41. Williams, R. (ed.), Explaining Corruption, Cheltenham, UK: Edward Elgar, 2000. 42. Yin, B.R., Case Study Research: Design and Methods, Thousand Oaks, CA: Sage, 2009. 231 Appendix 1 Key Variables for the 28 Corruption Cases5 Ye ar G en de r Ag e Po sit io n/ De pa rtm en t/ Ar ea Fi rs t a cc us at io n/ cr im in al o ffe ns e Br ib e Pr es en ce Fi na nc ia l/ Fa vo r Im pl ica tio ns M ot iva tio n5 As se rti ve / St ro ng Pe rs on al ity 1 2008 male 38 police/judicial yes yes low fi nancial Yes 2 2008 female 56 education yes yes low fi nancial Yes 3 2003 male 39 police/judicial yes yes high fi nancial Yes 4 2005 female 25 customs/fi scal yes yes low fi nancial Yes 5 2005 male 25 police/judicial no yes medium fi nancial Yes 6 2008 male 60 administration/services yes yes high fi nancial Yes 7 2008 male 30 administration/services yes yes high fi nancial Yes 8 2008 male 28 police/judicial yes yes medium fi nancial Yes 9 2002 male 27 administration/service yes no high fi nancial Yes 10 2008 male 53 education no yes high fi nancial Yes 11 2006 male 24 police/judicial no yes low fi nancial Yes 12 2007 male 26 police/judicial no yes low fi nancial Yes 13 2002 female 26 administration/service yes no high fi nancial Yes 14 2003 female 51 police/judicial yes no high kinship No 15 2008 female 46 education no yes low kinship No 16 2007 male 34 police/judicial yes no medium kinship Yes 17 2006 male 48 administration/service no no low kinship Yes 18 2007 male 54 education no no low kinship Yes 19 2007 male 53 education no no low kinship No 20 2006 male 35 customs/fi scal yes yes low kinship Yes 21 2008 male 54 police/judicial yes no low kinship Yes 22 2008 male 26 police/judicial yes no high kinship Yes 23 2005 male 28 police/judicial no no low kinship No 24 2009 male 29 customs/fi scal no yes low status Yes 25 2008 male 26 police/judicial no no high status Yes 26 2004 male 55 police/judicial yes yes low status Yes 27 2004 male 43 police/judicial yes yes low status Yes 28 2007 male 52 administration/service yes no low status Yes 5 The motive of the corrupt act was determined based on the triangulation of several data sources: the statements made by suspected public offi cials, the information obtained from interrogation and court transcripts. Although, on most occasions there might have been sev- eral motives for engaging in corruption, in each case there was a clear dominant motive for engaging in the specifi c corrupt act for which one was being investigated. It is important to note here that the presence of a bribe did not necessary mean that the act was motivated by fi nancial considerations. For instance, in 12 of the studied cases there was no bribe involved. In eight of the twelve instances, the suspected public offi cials motivated their actions by the need to satisfy the interest of their networks of association. In two cases although no bribe was involved it was still determined that the main reason behind the corrupt act was fi nan- cial. Even if the public offi cials did not receive a bribe, they engaged in corrupt behavior with the expectation of future fi nancial benefi ts. Furthermore, in two of the instances when bribe was demanded, it was done so in order to pay off someone else in the system; the actu- al corrupt act was still undertaken for the benefi t of someone in one’s appropriate network.