sually corruption is thought as a 'personal' matter: an ethical failure of a particular individual that, for his own benefit, does not act according to the law or a moral code. Our standpoint in this paper differs from this common perception of corruption. In our view, corruption results from particular historical processes and how specific human and social relations are configured in these processes. We are, of course, not referring to isolated cases were acts of corruption occur very seldom. Instead, we are interested in systemic forms of corruption. That is, when specific acts of corruption are part of complex systems of social relationships involving institutional agents, organizational traditions and cultural formations. In these contexts, although 'acts' of corruption are, of course, committed by specific social actors, they result from a more complex system of social and cultural forces. Specifically, current research shows that the stability and credibility of a society's institutional system, the perception of a shared fate by most members of society, the levels of inequality and the perception of fair opportunities for personal progress are all elements that either strengthen compliant behaviour or predispose people to anomic acts, championing their own personal interests over the collective good.

In order to show how the social construction of corruption occurs we will concentrate on one particular case, which is how corruption takes place in the relationships between state agents (such as the police, politicians and the judiciary), and members of criminal organizations, but also how members of civil society ('normal citizens') become involved in corrupt social relationships. During the last decades Argentina has faced enormous corruption problems, especially by state agents. The corruption perception index elaborated by Transparency International shows that in 2011 Argentina ranked among the most corrupted countries in the world (a problem common to many other Latin American, African and Eastern Europe countries). 1 II.

This perception is congruent with what has been find by several researches, that state that corruption has become one of the central problems in Latin America in general and Argentina in particular (Manzetti and Blake, 1996;Kurt, 1998).

In this paper we will try to show that these levels of corruption derive from historical and social circumstances. Although we will concentrate in the Argentine case, our aim is to use this as an example to reflect on the historical and cultural processes that usually underlie systemic forms of corruption. Hence, initially we will show how corruption is coproduced by several types of institutional and civil agents and how this responds to specific social conditions. Then we will explore the historical circumstances that derived in the institutional and civil cultures that are associated with the emergence of systemic forms of corruption. Finally, we will draw some conclusions on the historical and social conditions that foster systemic corruption. Waldmann (2003) has shown that, historically, the rule of law in Argentina has been conditioned by the capacity of the state to act according to the law and to generate the conditions that would make the argentine population interact according to basic legal principles. In this line, Nino (1995) has spoken of the anomic character of Argentine society and O'Donnel (2002) of the '(un)rule of law' in Argentina. From different standpoints, these authors reveal the complex causes that explain the particular ways in which norms intervene in the interactions between the state, its agents and the argentine population. Hence, Waldmann (2003) has shown how the Argentine state has historically had a tendency to legislate beyond its means of control. This gives to encoded laws an ambiguous status, since they exist as formal principles but cannot always be enforced as actual ruling norms. This ambiguous character of the law creates a propitious context for 'corruption' to flourish. Since the law is applied contingently, state agents may select when and how to enforce it. Hence, corruption finds a fertile soil, since it results from the suspension of enforcement. This may be easily dissimulated, since it not only happens because of the discretionary capacity of state agents, but also from the mere incapacity they sometimes have to impose the law.

Moreover, the ambiguous status of the law has also propitiated a more active role of the argentine state in the violation of the law. As O'Donnel (1984) has shown, the anomic character of argentine society is related to the recurrent emergence of dictatorships. Where the state not only has been involved in an illegal and brutal repression for political ends, but also in common crimes (we will come back to this later). Thus, O'Donnel's observations reveal that the creation of corruption is not the mere production of the state. Although state agents play the 'leading' role, the multiple ways in which corruption is created in Argentina show that it also arises from a predisposition of members of civil society to take advantage of the possibilities provided by such corrupt practices. In this sense, Nino's contributions should be carefully regarded. He shows how the fact that the instrumentation of norms is conditional in Argentina promotes a particular type of 'civil' conduct. In it, the search for occasional personal benefits prevails over more long term and rewarding collaborative social ties. Hence, corruption develops from the ways in which state agents choose to enforce (or not to enforce) the rule of law, and the predisposition of different actors within civil society to take advantage of the possibilities this opens for them.

The conditions generated by these state of affairs has made corruption rather ubiquitous. As Sain (2008) has shown, when looked at from the top of the institutional ladder corruption appears as produced by structured institutional schemes. Through them state agents control substantial parts of certain illegal markets and generate a more or less constant flux of resources that partly contribute to finance security organizations such as the police and the political system. However, when looked at from the micro social perspective, one sees the proliferation of multiple types of arrangements between state agents and actors with different types and degrees of involvement in organized and semiorganized criminal networks. Hence, partly, corruption comes from powerful institutional actors that in several ways and to several extents control the rule of law. But this has become so naturalized for particular state agents (as the police, judges, etc.), and those who are in conflict with the penal law, that micro forms of corruption are part of the ongoing way in which the argentine state regulates crime. A few examples will allow us to show some of the more habitual forms taken by these types of arrangements. In those interviews the police and judicial agents recurrently appeared as an irreconcilable enemy, but in spite of this the delinquents' narratives were full of anecdotes were several types of negotiations with politicians, the police and members of the judicial power were described as common practice. Hence, although many times the delinquents' tales described politicians, the police or the judges as despicable, they also frequently mentioned for example, how the illegal weapons market was to an important extent in the hands of members of state agencies, and how they often had 'rented' 2 We worked (sic) for this guy, who then became the president of the chamber of deputies. We were not really into politics, but we would go to political rallies as weapons from police officers or members of the armed forces. Also, many young criminals explained how, if they were caught, it was possible to negotiate their freedom with the police or the judiciary in exchange for a part of what they had obtained in their robberies. Or, even more, once they had been caught several times and had previous criminal records they also told how they were extorted by the police who asked for a sort of ransom in exchange for not accusing them of crimes they in fact had not committed (Isla & Valdez Morales, 2003). Thus, although the police and the judiciary were in principle described as an alien power, the narratives showed that there was a constant co-production of crime.

In one of the interviews, Gardelito, a criminal with 35 years of age and an extended criminal record, described very eloquently the more habitual and systematic form in which this type of arrangement functioned.

Volume XII Issue X Version I

a sort of 'security force' (sic) in case members of other parties would go there to make trouble. We did the same when militants of the party would go out to paint the walls [graffitis], in case guys of other political fractions would come to harass them. So, before the rallies this guy would come with a bag full of arms we could use during the campaign, and then we could keep those guns. [?] So in exchange for this sort of help we gave them, we got the guns and they would also arrange with the police and free some zones, so we could work [rob] without interference from the police. We worked as road pirates 3 The story told by Gardelito describes, in a sort of emblematic way, how different forms of exchange and reciprocity affect the relationships between the police, the politicians and the judiciary. However, his tale is far from been an isolated or unique case. For example, in his research on the political organizations in poor Argentine slums, Auyero (2001:95) has also found that young people in conflict with penal law are taken to political rallies to act as 'security forces' and that drugs and a certain 'protection' is given to them in exchange for that participation. In a similar vein, Garriga (2005) in his research on violence among football fans found that politicians are many times involved in football clubs and hold important relations of 'reciprocity' with hooligans and the police. For example, these three parties many times negotiate and allow hooligans to illegally charge for parking spaces next to football stadiums during the most important matches. The earnings from this activity , so he would tell us: "tomorrow from this to that hour you can work on this road that the police won't be there."

This narrative exposes a classical example of how state agents and criminals participate in the coproduction of corruption. In this case, the connivance between politicians, the police and criminals arises from territorialized social ties.

Although the story does not make it explicit, we know from further interviews that the initial contacts between politicians and these members of criminal networks develops out of the connections established by the territorial antennas of the political parties and the collusion with the police that becomes possible through these contacts. Hence, in this case we find that corruption results from a particular constellation of actors in a particular territory. However, the effects produced by this sort of arrangements are not circumscribed to the immediate territory where these ties were initially constituted. As the alliances became more extended involving more powerful state agents the influence of these networks transcended the immediate territory producing their effects well beyond the initial spatial frontiers. are then shared mainly between the hooligans themselves and the involved policemen.

This example reveals a further dimension of corruption. Although corruption results from the collusion between state agents and criminal networks, their effects reach the common citizen. As the previous case illustrates, those who go to football stadiums must pay a quite substantial amount of money in order to park their cars and also to avoid any damage to their properties or their persons if they refuse to pay. In fact, it is an extortive mechanism where resources are extracted from the common citizen by an illegal arrangement between the forces of order and the forces of crime. In this case, the 'victims' of this criminal activity are circumstantial and could easily avoid the problem by not concurring to the stadium. However, in other cases the possibility of common citizens to subtract themselves from these extortive pressures is less simple.

In a research done by Puex (2003) in a slum of the surroundings of Buenos Aires she found that although there was an informal 'prohibition' for delinquents leaving in the neighbourhood to rob other neighbours, in fact robberies were frequent. The inhabitants of the slum were cautious not to leave their homes alone in order to prevent robberies. However, it was common for neighbours to come back from work and find that especially consumer durables like TV sets, microwaves and other domestic goods had been stolen from their homes. Many times these elements were then offered by the thieves to the prior owners at a relative low price. And it was also frequent for those who had suffered the robberies to re-buy the goods from those who had stolen them. This happened because it was cheaper to by them from the thieves that to acquire them in the formal market. In addition, even if in these transactions the victims got to know the identities of those who had robbed them, they were reluctant to denounce them because they knew that the robbers had 'arrangements' with police officers. So, in case the robbed neighbour would denounce the incident in the police station he would be asked for the receipt of the stolen goods (which they generally did not have). And, in case they could not produce them, they would be the ones under suspicion of having stolen goods in the first place.

Hence, denouncing being a victim of crime could backfire turning the victims into accused or suspected for that same crime. Through this mechanism the police precluded the possibility of neighbours to even denounce delinquent activity and with it the responsibility of having to intervene in such affairs. This left the weaker neighbours to their own fate in the hands of the local criminal networks.

The exposed examples suggest at least two things. On the one hand, that corruption does not only affect those state agents and members of criminal networks that partake in spurious arrangements. It also has an effect on the common citizen. In addition, it also shows that the powerless and marginalized members of society are probably the more exposed to suffer the consequences of systemic forms of corruption. Moreover, the previous cases show that while state agents are probably the ones who obtain the greater benefits from systemic corruption, this type of arrangements also result from the fact that, as Block (1974) demonstrated for the Sicilian Mafia, certain members of civil society obtain benefits by exploiting other members of civil society. That is, those who are predisposed to exert force on other members of their own social group can sell this 'good' (violence) to 'corrupt' or 'distant' 4 As I interviewed members of different generations of delinquents, it became apparent that in all age ranges 'exchange' with members of the 'forces of order' was a common practice. But in the younger generation these exchanges were not always part of stable and extended systems of social relationships, as in the case presented by Gardelito or the type of arrangement Garriga describes in relation to hooliganism. What we found in our research in relation to the younger generations is that, in many occasions, the protection offered by state agents was very circumstantial and a matter of 'opportunity'. For example, the younger delinquents many times described how they could buy their 'freedom' from a street police officer just by handing in the purse or the necklace they had just stolen. This kind of arrangement made social bonds more labile. Hence, if always the forms of collaboration between the 'forces of order' and the 'forces of crime' is unstable and may turn into confrontation very easily, what seems to characterize these ties in the case of the later generation is at the same time a: (i) more naturalized practice of this type of state agents, sharing thus the benefits obtained by this extortive kind of arrangement.

Another element that is illustrated by these examples is that corruption is not an isolated phenomenon that occurs exceptionally or only in certain particular moments or occasions. On the contrary, the type of relationships that produce corruption are a naturalized and recurrent form in which illegal activity is regulated by agents of the Argentine state. And, we should also mention that it is not new. As already suggested, the collusion between politics, police and crime may be traced back almost to the origin of the Argentine state. However, what does seem to have developed more in the latest decades is a sort of contingent, short lived and micro social form of corruption.

bond, which (ii) takes place at all the levels of criminal and state organizations (it is not only a systematic arrangement between the 'heads' of the state organizations and 'stable' criminal gangs) and is (iii) more labile, easily turning from collaboration into confrontation -according to Bazzano and Pol (2009) the number of casualties in confrontations between the police and members of civil society has grown steadily since 1996.

What seems to characterize the more recent evolution of the production of corruption in Argentina is a sort of 'democratization' of this practice. That is, it is not only a practice that happens between organized criminal gangs and high ranked members of the controlling state agencies. Although this type of exchange certainly occurs at the top of the ladder, it has become naturalized to the extent that it occurs in a continuum that spans from stable forms of organized crime towards semi-organized and even what could be termed disorganized crime (delinquency that occurs spontaneously, with no previous plan or the intervention of a structured group). Following recent production on the conditions of civic morality and corruption, our premise is that this production of corruption does not happen spontaneously or only because of the particular 'perversion' of a specific set of social actors. When this type of social practice becomes extended and naturalized is because certain social conditions make them a 'possible' or even desirable course of action for certain social agents.

According to our view, the type of social ties that favour corruption are shaped by particular institutional structures, the ways in which public and private actors relate to each other, the opportunities that arise in illegal markets and particular institutional trajectories. Furthermore, complementary studies show that uncivil conduct is also related to sustained economic inequality and lack of fare opportunities for personal progress, that hinder the sentiments of sharing a common fate by most sectors of society (Rothstein and Uslaner, 2005). Also, the quality of political institutions and the levels of trust they inspire in the civil population are important predictors of the levels of 'civic morality'-the predisposition to comply with social norms- (Letki, 2006). Hence, corruption is the result of particular structural and institutional conditions that promote certain forms of (a)moral conduct among the general population.

According to these basic set of premises, the structural and institutional conditions found in Argentina are conducive to the type of social conduct that would produce corruption. For example, equality and the perception of fare opportunities for personal progress promote the perception of a shared fate and the predisposition to act according to social norms instead of personal interests. Notably, several data show that objectively and subjectively these conditions have not been met in recent Argentine history. Estimations show that since the mid 1970s argentine society became increasingly unequal (Smith, 1991;Robinson, 1999;Villalón, 2006). According to Guadagni et al., (2002:27), between 1975 and 2000, the Gini index escalated from 0.39 to 0.54 showing a growing unequal society. Concurrently with this state of affairs, Argentineans do not perceive equal opportunities for everyone in the country. The Latinobarometer 5 However, in order to explain how these conditions may induce public agents to engage in relations of reciprocity with the criminal underworld we need to develop our argument somewhat further. As the argument stands now, what these data would explain is not so much the conduct of institutional agents, but instead how the conduct of these types of agents would impact on members of civil society. Thus, the deviant conduct of public officials would undermine trust in political organizations and thus would induce the population to act following their own personal interests before the collective good or a rule governed conduct. Now, the issue here is that corruption is not generated solely by the actions of 'private actors', but instead mainly by public agents that induce or accept unruly survey estimates for the year 2000 show that after two decades of economic decline and increases in poverty rates, only 14.3 per cent of the population thought that everyone in the country had equal opportunities for personal progress.

Another element that has been mentioned as a condition for civil conduct is trust in political organizations. According to several studies, the quality of public order and the performance of public officials and office holders may induce a moral conduct among citizens. Hence, the lack of trustworthiness and efficiency in institutional agents debases the required foundations for civil conduct (Letki, 2006). Again, the data presented by Latinobarometer shows that the conditions found in Argentina are inconsistent with those required to promote conducts congruent with social norms. Most Argentineans don't find political organizations as trustworthy, for example 66.6 per cent of the population manifested that they have no or scarce trust in the police; 64.3 per cent gave the same answer with regards to the parliament and 78.9 per cent showed these same levels of distrust in relation to political parties. The justice system did not score better, 68 per cent of respondents showed either no trust at all or low level of trust in the judiciary, and the same happened with the military (58.3 per cent). Hence, these estimates make clear that if civil conduct is inspired by trustable political institutions, the perceptions that predominate among the argentine public don't favour rule oriented behaviour.

conduct by members of civil society. The question is then what are the circumstances that explain the illegal behaviour of the members of civil institutions. Our argument is that, in part, these conducts are explained by the traditions that predominate in the institutional settings to which these actors belong and that explain the perceptions the general public have of them. But, in turn, the logics of collective action that are revealed by the aforementioned theories of how moral conduct comes to be also play a part in this game.

The involvement of state agents, like politicians, policemen or the judiciary in crime or in negotiations with the delinquent underworld could probably be traced back to the origins of those institutions. Although there is no space here for a complete historical account of the trajectories of institutions such as the police or the judiciary, available material shows that the collusion between the forces of order and the forces of crime is hardly novel in argentine history. Barreneche's (2007a) account of the efforts overtaken in the 1930s to 'professionalize' the police show how one of the central problems was exactly the long standing collusion between local politicians and the head of local police stations ('comisarios'). Since originally local police chief officers were appointed with the consent of local political powers, there was a strong connivance between the two forces. The police would benefit from the predisposition of politicians to support promotions and nominations for future staff, while politicians could count with the police to prosecute opponents and in the organization of the electoral frauds which were frequent in several moments during argentine history.

In the 1930s an ambiguous effort was made to transform the police into a more centralized and organic professional force. For example, new institutional controls were deployed to try to eradicate traditional focuses of police corruption: the 'administration' of gambling, prostitution and the nomination of new agents which many times implied hiring unqualified personnel who actually did not work within the police force. The efforts failed, basically, because of the contradictory interests within the Conservative Party itself -that remained in power until 1943 thanks to the coup d' Eta of the 1930s and subsequent electoral frauds. Hence, at the same time that from the central powers of the state officials who belonged to the Conservative Party tried to promote these reforms, the local leaders of that same party successfully opposed them essentially because of the limited capacity of the central powers to control distant locations (Barreneche, 2010: 36-41). Furthermore, the traditional practice of using the police to prosecute political opponents was deepened, since the police was explicitly given the task of prosecuting members of the Communist Party by the governor of the state of Buenos Aires (Barreneche, 2010: 40) and was Year also systematically deployed by the political powers in electoral frauds (Béjar, 2004). Hence, even in this brief account it becomes evident that the conditions that accompanied the development of the police force, at least in the province of Buenos Aires, favoured an institutional tradition that naturalized the joint administration of crime by the police and elements of the political system in order to obtain personal and corporate benefits.

During Peron's government (1945)(1946)(1947)(1948)(1949)(1950)(1951)(1952)(1953)(1954)(1955) there were renewed and more successful efforts to reign in the police under a centralized command and to subtract it from the influence of local political leaders that belonged mainly to the Conservative and Radical Parties. Hence, the effort was not only directed to modernize the police forces and also give them a social role, but also to undermine the power of political opponents. Partly, the reforms promoted by peronism resulted in a more centralized and professional force, but under a military model that did not favour a conception of the police as civil servants. Instead, this institutional model fortified the disciplinary, hierarchical and authoritarian traditions that had accrued in the police forces since their creation (Barreneche, 2007b:246).

There is no space here to even try a systematic approach to these institutional histories, but research on the later phases of this process show that the last dictatorial period (1976)(1977)(1978)(1979)(1980)(1981)(1982)(1983) represented an important inflexion point in the way that the police and the military regulated the world of crime. A well established fact in this respect is that during those years the 'military' model became also predominant among the police forces (Kalamanowiecki, 2000;Kaminsky, 2005:47-49). Therefore, the police begun to perceive themselves not as providers of a public service and thus allies of the civil community, but more as 'guards' of a political order they had to 'impose' on that community. Members of civil society were thus perceived as potential enemies of that order, and not as the citizens entitled to receive protection and collaboration from 'civil servants' such as policemen or members of the judiciary.

In addition to this, during the dictatorship the armed forces (police and military) won complete independence from civilian control. They became autonomous organizations that wielded an almost complete discretional power over the rest of society. It is well known that this situation conduced to a rampant violation of human rights that, especially in the case of the police, continued even after the return of democracy in 1983. But the traditions of illegality accrued in the armed forces during the dictatorship surpassed the 'mere' violation of human rights. In addition to political murder and brutal repression, the armed and security forces, frequently joined by groups of civilians acted, at times, directly as common criminal gangs. Taking advantage of the impunity given by their almost absolute discretionary power, they turned the military repressive machinery into a criminal organization and were involved in kidnappings for ransom, theft and other common crimes that had economic instead of political purposes (Dutil & Ragendorfer, 1997;Olivera and Tiscornia, 1997;Pereira and Ungar, 2004;Isla, 2010) In the post dictatorial context (after 1983) the military structures were progressively reformed and subdued to the civil powers to the point that their capacity and traditions of confrontation against the democratic order increasingly waned. 6 III.

Contrastingly, although there were several attempts to introduce reforms in the institutional structures and cultures, the police forces never suffered such radical transformations (Camou and Moreno, 2005:144-145;Brinks, 2003;Kent. 2008). In part this resulted from the fact that the illegal conduct was so ingrained in the institutional culture of the police that any attempt of reform faced enormous challenges (Smulovitz, 2003). But an additional fact was that these same traditions involved the collusion between political parties, the police and the criminal underworld. As shown, the transactions between these three kinds of actors could be traced back to the 1930s when the Conservative Party used elements of the police and the criminal underworld to control electoral polls (Pignatelli, 2005;Cecchi, 2010). Moreover, even the peronist armed leftwing groups of the 1970s had fluent contacts with members of some of the more recognized criminal gangs (Isla, 2007).

Thus, although with democracy the civil powers regained certain levels of control over the police, these controls did not result in a systematic persecution of the illegal practices committed by the security forces. Instead, this capacity to control was many times used by political officials and even members of the judicial power to partake in the resources that circulated in the illegal markets administrated by members of the forces of order. Hence, as several authors (Fuentes, 2005;Sain, 2008;Dewey, 2011) have clearly shown, the impetus of institutional reform of the police forces has many times been limited by the interests that are ingrained in the complex articulations between criminal, police and political networks that promote those same reforms.

When we ask ourselves how and why corruption expanded in Argentina, we need to look at the fact that the traditions ingrained in public organizations aimed at controlling the criminal world were never congruent with 6 Between 1983 and 1991 there were several -failed--attempts of the military forces to overthrow the democratic government. Although with ambiguities and drawbacks the successive democratic governments introduced important institutional reforms that neutralized the capacity of the military to intervene in civil society. the promotion of civic morality. The institutional cultures developed in the police, the political system and the judiciary were forged in processes were those institutions were instrumented according to particular corporate or even personal interests. Compliance with the encoded law was subordinate to those same interests. Hence, the official agents of these organizations have hardly ever understood themselves as members of a national community with a shared destiny (a paramount condition for civic morality). Instead, they think of themselves as part of restricted networks with particular interests to defend. As Tilly (2004) has shown, when the general state is not seen as a trustworthy structure, this type of network tends to preserve their interests over any alien intrusion.

Hence, a possible answer to the question of why even state officials act disregarding public morality is that they, more than anyone else, experience the fact that 'public' organizations act responding to particular interests. They thus 'know' that formal social norms hardly rule concrete social interactions and relationships. Therefore, if as Harding (1993) has postulated, moral conduct results from the fact that social actors comply with normative behaviour as long as they 'experience' that other members of their group share the same values and dispositions, in the argentine case, public officials perceive this is not so. For, they directly observe how even for those who are supposed to be the most compliant-themselves-, conduct is governed by corporate interests.

Considering this, a possible explanation of why and how corruption became naturalized in argentine society is that the traditions and institutional cultures that predominate in public organizations have historically championed fractional or personal interests over normative goals. In these kinds of contexts, all types of social actors, including state officials, tend to act protecting the interests of the networks or systems of interpersonal alliances to which they belong. Since for collective norms to function it is necessary that al l involved parties respond to those norms, and historical and collective experience shows to those same actors that other parties are not trustable in this respect. In addition, another factor is that, particularly since the mid 1970s, argentine society became increasingly unequal. The lack of opportunities for social progress, especially for the less well off sectors, contributes to shatter the perception of a shared fate buy most members of society. And this, in turn, bolsters the tendency to champion personal or corporate interests over the collective good and the social norm. In sum, poor quality of public organizations and growing economic inequality are two paramount factors in explaining the high levels of corruption in Argentine society.

Notably, Argentina does not stand alone. Many underdeveloped nations experience high levels of corruption. Usually, these societies are stigmatized because of this, attributing the levels of corruption to the 'moral fault' of its functionaries and lack of appreciation for the law by the standard citizen. What we have tried to show by exploring the argentine case is that things are not that simple. Moral conduct is not solely an individualistic decision. For most individuals, the possibility of behaving morally or immorally is the result of the system of social relationships in which they participate. And these systems result from complex historical processes and socio-structural conditions. They are not an individual decision.