WHEN CORRUPTION IS CULTURAL: EXPLORING MORAL, INSTITUTIONAL AND RULE-BASED CONCEPTS OF CORRUPTION* CUANDO LA CORRUPCIÓN ES CULTURAL: EXPLORANDO CONCEPTOS DE CORRUPCIÓN MORALES, INSTITUCIONALES

It is often asserted that people are conditioned to act corruptly by their cul ture in a way they cannot help themselves. resumen: Se afirma con frecuencia que la gente está condicionada a actuar de manera corrupta por su propia cultura. Utilizando una aproximación http://www.juridicas.unam.mx Esta revista forma parte del acervo de la Biblioteca Jurídica Virtual del Instituto de Investigaciones Jurídicas de la UNAM https://biblio.juridicas.unam.mx https://revistas.juridicas.unam.mx Universidad Nacional Autónoma de México, IIJ-BJV, 2020 http://revistas.juridicas.unam.mx/index.php/derecho-comparado/issue/archive DOI: http://dx.doi.org/10.22201/iij.24484873e.2019.156.15155

The aim of this paper is to use a multidisciplinary approach, both from political theory and political science, to show that this kind of narrative about corruption is flawed because it is not informative at all about the nature of corruption. This prevents it from leading to any type of meaningful analysis or policy design. We will concentrate on two main flaws: The Triviality Objection, which points out that everything humans do is cultural in some sense or other, and the Circularity Objection, which stresses that attempting to explain why or how corruption becomes part of a specific culture, leads to saying that it is because its members act corruptly. The idea that the cultural causation is flawed becomes persuasive when we contrast that view with our concept of corruption as a special kind of harm to institutional rules: corruption may refer to a parallel set of conventions or rules that undermines the institutional set of morally justified norms.

I. iNtroductioN
"Corruption is a cultural fact" media pundits, public officials and even some academics repeat as a sort of resigned excuse, as if they were talking about the inevitability of the weather. Of course this folk understanding of the cultural origin thesis of corruption -what we call the conventional view-is a strawman when formulated in an absolutely deterministic fash- ion. But determinism is a piecemeal condition. As a result, more plausible formulations of the conventional view may be available. This paper aims to use a multidisciplinary approach with tools from both political philosophy and political science to show that this kind of narrative about corruption is flawed due to it failing to be informative at all about the special harmful nature of corruption, which differentiates it from other harms, such as fraud or theft. We will concentrate on two main flaws: -The Triviality Objection points out that everything humans do is cultural in some sense or other, so it is akin to claiming that poverty or wealth are cultural: it is not informative of the phenomenon. -And the Circularity Objection, which stresses that seeking to explain why or how corruption becomes a part of a specific culture leads to saying that corruption is part of a culture because people act corruptly, which in turn is question begging.
The relevance of such objections may be clearly seen in the realm of political science, where triviality or circularity cause a wide array of concepts of corruption to become inoperant in most policy contexts. Specifically, in liberal democracies, which involve the expectation that government decisions have morally defensible results and require "a high order of responsible behaviour from its citizens" (Steward 1993, 26, 317-330), a trivial or circular concept of corruption renders the process incomplete. Quite simply, it is impossible to achieve any results when the concept to seek them is trivial or circular.
Some important caveats are due, before we proceed to the analysis. First, it is of course tempting to say that corruption is cultural if the members of a cultural groups find it is, so the disagreement about the nature of corruption is in some sense semantic, between folk and normative understandings of the term. But, it is important to stress that this paper moves in the realm of conceptual analysis, which cannot be reduced to semantic disagreements. Instead we will focus on the nature and conditions of institutional corruption in order to disclose its conceptual relationship with culture, as opposed to assigning possible meanings to the usage of the word "corruption". Conceptual analysis involves, for instance, determining what it is about corruption that makes it a special kind of wrong different from similar ones, such as fraud or noncompliance, what follows from the nature of corruption, or the conditions of possibility of corruption. We do not delve into that analysis here. Instead, for the purpose of this paper, we focus on the effects of corruption in order to challenge its cultural cause or its origin. Under this perspective, we assume that corruption is a collective institutional harm. The question, then, is what kind of special harm corruption does that could be culturally determined, culturally caused, or at least culturally originated. We invoke the nature of corruption only insofar as it could be potentially connected to cultural origin. As such, we will not necessarily address explanations that focus on the causes of corruption, such as principal-agent and collective action models (Perssons, Rothsstein y Teorell 2012). While highly useful and in some cases sophisticated, such models have a different purpose from this paper. We are not as concerned with incentive structures that may favour corruption in the future or explain the reasons for its existence, as we are with facing the reality of corruption as an already existing harm to institutions.
Second, and related with the last point: Conceptual analysis is not reduced to semantic disagreements because it does not describe the use of words. This means common-sense notions and folk understanding should not be particularly problematic. For the purposes of this paper, let us distinguish between the traditional folk explanation that claims that corruption is caused by the local culture; from the familiar understanding provided by the social sciences, which depicts many social phenomena as social constructions. We will come back to the approach of social sciences in further sections.
The agenda of this paper runs as follows. In the first two sections we explain the problems with describing corruption as part of culture, and go on to propose a political understanding of culture in order to evaluate if corruption may be politically determined. Nonetheless, corruption could be the result of precisely the opposite: a failure of culture in determining morally justified values, presenting it as a broader ethical issue. In section three, we consider the idea of corruption as a breach of morality, which presents corruption as a special kind of immoral act. At the same time, this concept could be shown to be co-extensional with a conception of morality as culturally embedded. However, the claim that corruption undermines the moral purpose of institutions seems circular: to corrupt an institution is to corrupt its purpose, which corrupts the institution. The fourth section brings up a rule-based conception of corruption that fo- cuses on the validity of rules to point out exactly how corruption harms institutions excluding the cultural origin. In the final section, we use the tools of social constructionism and hermeneutics to attempt an overview of how the rule-based conception of corruption may be seen as a cultural process in Mexico. This analysis attempts to outline how a concept of corruption that actually points out its harm to rules may displace a cultural understanding of corruption by explaining some of the connections between historical processes, institutions, and culture.
II. "corruPtioN is culturAl", theY sAY Very broadly, culture seems to be a kind of basic social technology that allows members of the same community to transfer crucial information from one generation to the next (Pinker 2002). The information is crucial in terms of the struggle for coherence of collective forms of self-understanding expressed in both the symbolic and the functional across various spheres of human activity (see Cassier 2012 y Ypi 2012). Of course, broad culture contains many kinds of irrelevant information, such as whether if one expects to find tortillas, naan or a baguette at a restaurant table, but crucially, it contains the kind of information that children need in order to be socialized. The socialization of children is possible because culture codifies horizons of value and interpretation, and it determines the meaning of ideas, concepts and distinctions (Haslanger 2012;Hacking 1999y Gadamer 1989. As a result, some social phenomena are culturally relative in the sense of being caused by culture itself. This is true for many conventions in everyday life. For example, people from Copenhagen (Denmark) will likely tend to arrive at a party and leave punctually, whereas people from Mexico City are more likely to arrive late and leave when they see fit. This is a good proxy for the behaviour of many people in the specified classes, and the explanation for this behaviour is certainly cultural. Yet, this folk understanding of Copenhageners and Chilangos 1 does not amount to a social understanding or explanation because it does not begin to explain the many reasons why people from said classes behave this way.
The problem of this traditional folk view is that from this commonsense understanding of culture, the traditional view extracts an implausible condition: a tragic sense of inevitability or social determinism. It is not only that Mexicans arrive late to social gatherings and stay for an unspecified amount of time, it is rather that they are causally determined to behave in this fashion by their own culture and we would be wrong to hope they behave differently, at least not until the culture itself changes. 2 Mutatis mutandis if Mexican institutions are corrupt, while Danish institutions are not, it must be because culture causes Mexicans to act corruptly in a way that they cannot help themselves, while Danish culture causes different behaviour; and unless Mexican culture transforms, we cannot expect Mexicans to behave differently. This kind of social determinism is implausible precisely because it eschews any explanation or understanding of the social phenomenon at all. Things are simply what they are.
It seems that, taken solely as social wisdom, the folk understanding ignores several obvious realities of cultural analysis, such as how no culture exists in isolation but is in fact "a dynamic amalgam of indigenous and foreign as well as ancient and modern elements" (Dalton 2005, 237-262). Mary Dalton points this out in her analysis of corruption in the Republic of Korea. According to her, several elements of Korean culture, such as its age-based conception of authority and hierarchy, have, at the same time, created a greater awareness of corruption as a detriment to democracy and provided individuals in power with a set of tools to benefit from it and hide it. 3 Such contradictions are bound to be found in most cultural contexts and would mean the folk view of corruption as a cultural inevitability is, at the very least, an oversimplification. 2 In this paper we remain agnostic about the possibility and dimension of cultural change. Regardless of cultural change in this paper we focus in the relationship between corruption and culture as an already existing problem that needs to be addressed. What is important to note is that while cultures may change, they do so very slowly, with the passing of many generations and usually how they evolve is not under any specific group's control, including the government. Rather it seems a more organic process. We thank Bernardo Bolaños for this observation. 3 In Korean language, people use honorifics to refer to other people based on their age difference, even if it is of a single day, and are expected to treat age difference as a difference in hierarchy. This can be extrapolated to how questioning the decisions of someone who is older or has a higher place in a hierarchy tends to be considered highly inappropriate, especially in formal and professional contexts. This creates a conflict when democratic institutions in the Republic of Korea place the responsibility of reporting corrupt acts and demanding accountability on individual whistle-blowers and society at large, whose members are unlikely to treat people in the government or in positions of authority as equals. Finally, it is important to stress that we are analysing corruption as a dynamic that takes place in institutions, rather than as an incentive structure behind isolated corrupt acts. Our question is more about what is happening to the institution when it becomes corrupt and why that is harmful, as opposed to why the corrupt act itself is taking place. Of course, acts of corruption are a necessary component of corruption, but it is obvious that a single act of corruption is not a sufficient for corruption as an institutional harm. An important corollary of this is that corruption is a special kind of harm conceptually distinct from problems of compliance and moral virtue. Even if noncompliance and unvirtuous behaviour may be conditions taken by several explanations as necessary for an act to qualify as corruption, they are not sufficient to explain corruption as harming institutions. This is because corruption may become prevalent even without breaking any rules or compromise the moral character of persons (Miller S. 2010). 4 If corruption was only a problem of people deviating from the rules of morality or law, we would be able to deal with it by invoking familiar accounts of compliance, virtue or justice ("can't we all just get along?"). However, corruption is a problematic issue precisely because of its distinctive structural, and institutional nature.

III. corruPtioN ANd Public PoliticAl culture
One way to make sense of the cultural causation thesis is the view coming from political philosophy that public political culture determines the preservation of just institutions. Famously, Rawls restated his theory of justice in culturally relative terms by suggesting that a conception of justice may be acceptable when its basic ideas of order, cooperation and fairness are rooted in citizens' public political culture. 5 Perhaps corruption could be the result 4 For example, "before 1977 it was not unlawful for US companies to offer bribes to secure foreign contracts". The same thing was not illegal for Mexican companies until after 2000. A common example of virtuous corruption would be the case of "Schindler's List", where Oskar Schindler corrupts the laws of Nazi Germany by bribing police officers to be allowed to hire persecuted Jewish workers to cut costs in his factories, consequently saving them from being sent to extermination camps. 5 Note that from the point of view of political theory and the state-community relationship, what matters is not broad culture, but only public political culture (Rawls 1999). Political Culture is defined as political culture is not only instrumentally valuable insofar as it helps people who do not know each other to cooperate in the different ways that the of an unreasonable public political culture, which favours injustice and prevents its members from realising the collective consequences of their actions. Babatunde Akanji presents an example of this understanding of corruption when he points to specific elements of Nigerian political culture, such as collective interdependence and hierarchical relationships, 6 as favouring concrete corrupt behaviours, such as bribery, clientelism and nepotism (Akanji 2017). If in Ideal Theory a reasonable political culture will socialize individuals in cooperation and compliance of the community's expectations, in non-Ideal theory an unreasonable political culture will socialize individuals for antisocial behaviour, simply by assuring them that everyone is likely to follow suit. In this section, we hope to show that this option fails to meet the triviality and circularity objections.
This way of restating the cultural causation of corruption seems to deal with the triviality objection, since it identifies one kind of culture that performs one specific task (political culture is only one small subset of a larger pattern of cultural interactions which socializes individuals as members of the community) and also identifies one specific way in which corruption is cultural (corruption is a type of behaviour that is socialised within a specific political culture). However, the problem with this notion is straightforward: it does not deal with the circularity objection. If the community may require. It is also intrinsically valuable because it seems to be part of who people are, of their individual identities (Miller D. 1995). Both objectivists and particularists agree on this point, but each of them arrive to different conclusions from there. 6 Akanji refers to the collectivist element of Nigerian culture, in which people tend to see requests for payments as something that individuals do in the name of groups they belong to, whether if it is their family, ethnicity or political party. Such payments are considered a quid pro quo between the groups represented by each person involved in the transaction, not as an individual request. Questioning the motivation for such requests is seen as highly inappropriate, especially if the person making the request has a higher place in a hierarchy, political or otherwise. While this can lead to innocent interactions, such as a restaurant patron giving a server a big tip in exchange for better service, it can also be seen as bribery when a similar situation happens between a civilian and a civil servant. It would be considered highly disrespectful for the server, for example, to question the motivation of the tip or refuse it in order to be allowed to give mediocre service. In contexts with clear hierarchies, such as in politics, a similar situation would play out when, for example, politicians embezzle government funds to repair infrastructure they see as a priority (generally for clientelistic purposes) or to improve their family's economic position. People who find out about the embezzlement would likely see it as a collective quid pro quo, where the politician is acting in the name of a group and will likely pay it back at some point. question is how people could be culturally determined to act corruptly and the answer is that they have been socialized within a corrupt political culture, nothing is gained with the explanation. We can keep asking what is corrupt in this purported culture and how this culture determines that some individuals bend the rules and not others, never clarifying what exactly favours corruption in a political culture or how it determines corrupt behaviour. Crucially, this refrains from attempting to demonstrate that there is a causal relationship or at least a strong correlation between being socialized in a political culture that favours corruption and actually acting corruptly.
With that in mind we can restate the problem in the following terms: how can we be sure that corruption is caused by socialisation, as opposed to a lack of socialisation or a defective process of socialisation? For if corruption is caused in this other way, it will not be culturally determined but just the opposite: a failure of culture in determining our identities and values. Akanji (2017) makes this point when he suggests long-term cultural change in Nigeria as a solution for corruption. According to him, Nigeria's current political culture shows a defective relationship between Nigerian society's collectively oriented mind-set and the expectations placed on individuals by democratic institutions. While the institutions were designed with the expectation that individuals would make decisions based on moral considerations (for example, to not bribe civil servants so as their services would reach everyone equally), in practice it has allowed individuals to take advantage of institutions for collective purposes. As such, according to Akanji, a long-term strengthening of democratic institutions is a realistic solution for the elements of Nigerian culture that favour corruption.
Consider the hypothetical case of Professor Lazlo, who in this scenario is the director of the Department of Philosophy in a fictional Catholic university in Mexico, a cultural context where corruption is known to be widespread. This fictional university is known for instilling the community values of work ethics in all students, as well as for striving to hire a diverse pool of professors and researchers with the means to explore a wide array of political theories through merit-based open competition. Lazlo is an expert in political critical theory, the kind of political theory that seeks to identify the ideological misuse of political philosophy as a doctrine that regrettably hides social domination and exploitation. Yet, when he faces the task of hiring new researchers for his department by means of open competition, he bends the rules every time to ignore the requirement for diversity in viewpoints and disciplines, and only allowing those whose work he deems compatible with his own research to advance in the contest. By hypothesis, Lazlo is not corrupting the institutional selection process because he is cynically looking for personal gain or because he enjoys torturing applicants. Lazlo sincerely (and wrongly) believes he has a valuable cultural project he is entitled to realize by ignoring institutional requirements, manipulating the contest by bending the rules, and handpicking human resources that accommodate his research interests. Yet, this is a case of corruption, because instead of going through the appropriate channels by convincing the university to define job vacancies according to what professor Lazlo seeks in an applicant and being transparent about such requirements, he instead manipulates the selection process so his preferred option is always selected and his own career benefits. Consequently, applicants are not sure if the selection committee expects them to fulfil the requirements in the contest rules (which advocate diversity in disciplines and viewpoints) or to simply fit in with Lazlo's personal preferences centred on only one kind of critical theory. If Professor Lazlo is not cynical, malicious or overly ambitious, he is well socialized, adjusted and knows that corruption is prima facie wrong, and he ends up acting corruptly, then, in keeping with the explanation of corruption as part of political culture, it must be because he is socially determined to do so, even if he knows better.
The problem with such an explanation is that, even if we reject the idea that corruption is a failure of the appropriate cultural socialization, this does not amount to say that corruption is culturally determined. Rather this case seems to suggest that corruption is a moral issue that becomes cultural when a specific culture's understanding of morality fails to prevent individuals from acting corruptly. Professor Lazlo believes that the intrinsic value of his cultural project trumps the institutional rules for hiring or the moral reasons for the existence of those rules. He is not determined by broad (Mexican) culture or by a bizarre political culture, but by morality, or rather by his own understanding of what morality requires from his performance as a faculty director.
This suggests that corruption may be an ethical issue rather than a cultural one. Note however that this is not precisely an objection against the conventional view insofar as ethics may be part of culture or even culturally relative, particularly within the particularist outlook (see Dancy 2017). The concept of corruption could be presented as a special kind of immoral act, as well as co-extensional with a conception of morality as culturally embedded. Additionally, this way of explaining the cultural specificity of corruption as a breach in morality explains both the nature of corruption and its origin, while also making corrupt practices seem culturally relative. This possibility will be examined in the following section.

IV. corruPtioN As A morAl breAch
Recent philosophical work on the concept of corruption points to it being pre-eminently a moral issue, as opposed to a legal or social problem. 7 Consider Seumas Miller's concept of corruption as a predominantly moral breach: Morality-focused concept of corruption (MC): An act x that is done by an agent A constitutes corruption if and only if: 1. z has the effect Ep of undermining or helping undermine a process or institutional purpose belonging to an institution, I, and/or has the effect Ec of helping to devalue the moral character of agent B while B has an institutional role in I; and 2. At least one of the following is true: There are many important elements in MC. Firstly, it clearly explains that corruption is an institutional phenomenon instead of a personal one. It is true that the moral corruption of individuals is often discussed, for example, when a judge is not impartial or a journalist is not objective (Miller D. 1995). However, for Miller, the type of corruption that constitutes a moral wrong in itself is the institutional kind, which does include the moral detriment of individuals, but most importantly, distorts the character of the institution on behalf of which said individual is acting. The damage to the institution is not just done to the role of its members, but to its processes and institutional purpose. This does not mean that individual acts are not relevant for corruption. Quite the contrary, for Miller, identifying corrupt individuals, either as corrupters or corrupted, is a necessary condition to identify a process of corruption. Using this concept, it would be possible to connect the cultural explanation of corruption with its immoral motivations and consequences. Professor Lazlo would be considered corrupt even though he is not doing anything illegal, as he is subverting the academic purposes of open contest hiring based on merit.
The problem here is that MC seems circular, not just in its application to specific cases (as shown by Dalton for the Korean case), but also in an abstract sense. Institutions, as a concept, exist with a purpose and, and according to Miller, undermining an institution is corrupting the institution. MC would seem to be assuming exactly what it is attempting to explain: that an act undermining an institutional purpose is essentially a moral wrong of a special kind with an institutional dimension, and that all institutions have moral purposes that may be undermined in a way that we can identify with corruption and not with some other reason, such as laziness, insufficient funding, or incompetence. MC does not explain which actions can undermine institutional purposes in a way that resembles what we normally understand as corruption. That is, under MC, maybe Professor Lazlo's undermining of the selection process is corruption because it undermines the department's moral dimension, though the concept lacks an explanation for the nature of the harm in Professor Lazlo's actions. After all, maybe he means well while doing so. With this concept of corruption, we cannot be sure if he is corrupting the department or doing something beneficial or innovative for it.
In order to correct that problem, let us consider a more sophisticated version of the traditional, economic definition of corruption as the abuse of entrusted power for private gain, in order to see if we can find cultural traits of causation. 8 This concept of corruption is widely present in economic and political science literature on the subject 9 and specifies how the act subverts the institutional order, perhaps pointing towards a cultural origin for this subversion: Abuse-of-power-focused concept of corruption (AC): An act X or pattern of behaviour X1, X2, X3… Xn, constitutes an act of corruption when: i) X or X1, X2… Xn is an abuse of power or authority by an government employee, according to what is described in the legal norms or the official rules which structure the institution, and ii) The act or pattern of behaviour in question happens with the motivation to obtain personal benefits by the people involved in the act or pattern of behaviour.
AC is attractive because it explains why Professor Lazlo's act is harmful as corruption. It is an act of corruption because it belongs to a pattern of behaviour which undermines merit within universities. Note however that AC avoids the moral character of corruption by focusing on the structural damage it causes to rules. According to AC, corruption is, in summary, a problem of compliance: a legal offense of economic character, such as bribery or graft (Miller D. 1995). Crucially, as corruption is defined as a pattern of behaviour, it could potentially be restated as a cultural practice: as long as the practice remains pervasive, Mexicans or Nigerians will be socially predisposed to take advantage of institutions without any regard for their instrumental value. Furthermore, AC correctly explains how deviating from institutional rules may be harmful by pointing out its economic consequences in the benefits the actors seek.
According to Miller, the problem with this concept of corruption is not only that it ignores many types of corruption that do not involve economic benefits or institutional roles, but it particularly ignores certain acts of corruption that are not illegal. For example, it is not always illegal for a company to use bribery beyond the borders of its own country. AC is 8 Special thanks to Lucero Fragoso Lugo and Milton Jair Rocha for this point. cultural in a trivial sense that may or may not distinguish a particular kind of harm. Furthermore, AC does not explain how this kind of harm is of a special kind or different from, for example, fraud or theft.
Our analysis then seems to suggest that both MC and AC are not entirely adequate to explain the nature of corruption, as each one captures a different aspect of it. AC makes an excellent point in explaining that it is not necessary to assign a legitimate moral purpose to every institution to identify the damage caused by corruption. It is also true that MC is probably right in pointing out in its first clause that corruption needs to have the effect of undermining or helping undermine an institutional process or purpose, and/or have the effect of helping to devalue the moral character of the actor taking part in corrupt behaviour. The problem is that if corruption needs to be an exclusively moral damage, as MC contends, it would need to explain how "undermining or helping undermine an institutional process or purpose" is a moral damage. The connection between undermining institutional purposes and corruption is easy to spot, but as AC points out, institutions do not need to have a moral justification for its processes to be undermined and corrupted. Therefore, MC would seem to be assuming exactly what it is attempting to explain: that undermining an institutional purpose is essentially a moral wrong, and that all institutions have moral purposes that may be undermined.
One way to supplement AC in order to identify what is special about corruption in a way that could help explain how institutional purposes get corrupted is to invoke the notion of organisational culture. Here we are not suggesting a differentiation between corruption in governments (as emphasised by public policy studies) and the private sector (as is the focus of organisational studies). We use the term 'institution' in a general sense, to refer to "systems of established and embedded social rules that structure social interactions" (Hodgson 2006a, 125), and 'organisation' to refer to "particular kinds of institutions involving rules concerning membership and sovereignty" (Hodgson y Jiang 2007, 1043-1061. Unlike culture in the broad sense, which involves national, ethnic, and religious affiliations; organisational culture constitutes a specific layer of culture that takes place within large organisations in both the public and private sector, and involves the values and behaviours that shape the institutional environment and its dealings with other organisations. Organisational culture may help explain corruption as far as it establishes a link between two conflicting forces that shape all kinds of formal institutions: enforcement and agency. In this context, agency refers to the extent to which an organisation's individual members are able to influence its shape and purpose. Enforcement refers to all sorts of limits to agency, either to curtail it, to shape it in some way, or even to stimulate it. It is necessary to keep in mind that the concept of organisational culture requires both institutions and organisations to be seen as separate from their individual members. Organisational culture may be self-enforced, for instance, when an institution or organisation requires certain behaviours within itself to work towards its own purpose. For example, the organisational culture in an institution that handles emergencies (such as a hospital) would require more strict enforcement of time-management rules than one that does not (such as an art gallery). On the other hand, it could also be externally enforced when requirements are alien to the organisation itself. This would include the requirements coming from the network of other institutions that allows the institution to exist by working towards their own purposes and following their own rules. For example, businesses in the food industry must comply with regulations enforced by the government's health and safety offices, regardless of the scope of their internal rules. It is important to identify the rigidity of the constraints that may prevent individuals from acting rationally, out of mere habit (Hodgson 2006, 16); as an organisational culture becomes corrupt when it awards its members with enough agency to disregard rules or moral consequences. An example of this are doctors in the medical field who prescribe unneeded medications to their patients, because pharmaceutical companies pay them to do so, while their institutions (such as the hospitals where they work) do not discourage the practice.
The problem with conceptualising corruption as organisational culture is that further analysis may reveal that this approach is not different from AC or MC, depending on the degree of moral assessment we use to evaluate the practices emerging from agency and enforcement. isational culture also fails to say what corruption is in a non-circular way; at least in a way that upholds the conventional view and its cultural explanation. If the analysis focuses on institutional enforcement, then AC becomes paramount, as the institution fails to limit agency in the appropriate way; leaving the issue of triviality in AC. If we instead focus on the individual agent, then MC becomes relevant, as the agent harms the institutional structure, leaving the issue of circularity in MC unaddressed.
Once we identify corruption with a problem of rules and compliance, then corruption is cultural only in a trivial sense. For example, consider the organisational culture of the Trump Organisation, where it would seem that being a member of the Trump family might be more relevant for a person's hiring than their professional qualifications. This means nepotism may be a part of its organisational culture, regardless of whether if there is a rule against it. Given Trump family members/employees openness about how they do not see many negatives in allowing nepotism to permeate their companies, 10 the issue of corruption would seem irrelevant precisely because it is part of their organisational culture, even if there may be written rules against it. What matters is how appropriate rules are for certain kinds of enforcement over agency in certain cases. We explore this possibility in the next section.

V. A rule-bAsed coNcePt oF corruPtioN
We hope we have shown that the cultural causation view about corruption is either trivial or circular. It is trivial by only stating the obvious when saying that corruption is cultural: that corruption is something that happens in human societies. But as we have seen, even when we try to flesh out this claim, the conventional view remains circular: to corrupt an institution is to corrupt its purpose, which corrupts the institution. We think, however, that some progress could be made if we explain what it means to corrupt an institution's purposes and how it happens.
In this section we lay the structure of a somewhat different concept of corruption that aims to clarify the specific harm that corruption does to institutions. Under this conception, corruption harms rules by establishing a parallel system of conventions and rules, which compete with institutional rules. For instance, if an institution is structured over the fact that there is a rule requiring you to do R; corruption establishes a convention that requires people to pretend to do R but do C instead; particularly if doing C is part of a pattern of behaviour C1, C2, C3...Cn; where many people may face the choice between doing R or C. Professor Lazlo's tampering with the selection process is corruption not only because it is morally wrong to impose his selection upon the institutional process, tampering with the fairness of the process. It is a special kind of harm we call corruption because his behaviour adds to similar behavioural patterns, which together undermine universities' rules of fairness and merit. The fact that an alternative behavioural pattern exists sets an example that establishes an alternative convention and eventually a rule: now people in a similar institutional position to Lazlo's will wonder whether to act according to fairness and merit, or to act according to the alternative convention Lazlo seems to follow. Therefore, corruption harms institution in three ways: (1) by contributing to the establishment of these parallel conventions and rules that constitute a deviation from the rules that structure the institution; (2) by illegitimately adding personal-gain free-riding purposes to those of the institution, (3) and by consequently reducing certainty; as people need to consider whether to follow R or C instead, given that they cannot be sure if everyone else is following either.
As we will see, saying that corruption is cultural is not much more informative than saying that any set of rules is cultural. A corollary of this obviously is that we should abandon the notion that corruption is cultural.
Before going forward with this concept of corruption, four considerations are in order. First of all, note that, from a philosophical perspective, a rule's validity, requires a critical perspective and careful interpretation of rules (Marmor 2010, 14 and applied in a way that gives it normative authority within the institution. 11 A rule's validity is what allows it to spell out the circumstances under which it is binding within a specific legal system (Shapiro 2009, 4). In general, rules become normative when they provide individuals with relevant reasons to behave according to the rule. In other words, normative rules tell people what to do and give reasons that supersede other possible reasons to deviate from the ruled behaviour.
Second, for the case of public institutions, rules structuring institutions are often rules of law. It is important to stress that although the relationship between validity, normativity, and legality is hugely controversial from a philosophical perspective, the damage that corruption causes to the validity and authority of the law as a system of rules, seems quite straightforward. Corruption establishes parallel rules or conventions to the ones dictated by the law, especially where the law gives shape to institutions and provides reasons for people to respect them. The rules established by corruption end up competing against the law, which undermines the law's authority by making them, at least in practice, optional. In contrast, Miller seems to assume that in pointing out that not all types of corruption are illegal, he is showing that corruption is a moral issue instead of a legal one. However, his assumption is reductive, as legality is only one of the many characteristics of the law.
The relationship between morality and rules is always complicated. Almost nobody believes they are entirely separated or entirely unified. Philosophy of the law analyses this complex relationship, and this complexity is sure to be replicated when discussing the relationship between corruption as a concept, morality, and the law. We do not intend to suggest that the damage caused to the law by corruption does not have a moral dimension, quite the opposite, the thesis is that this damage has the moral consequence of preventing the law from protecting people's rights. The reason for this is explained in a third consideration: as we saw it in the first section, it is only possible to discuss corruption when the moral character of liberal democracy is assumed.
Liberal democracy is a type of political community or legitimate society that adheres to principles such as the protection of basic liberties, a certain conception of equality among citizens, and the protection of human rights. This is important, because the condition of being corrupt may only be conceptualized as relative to moral principles or standards associated with not being corrupt (Miller D.1995). Thus, corruption is not an act in itself, but a process through which liberal democratic institutions lose their moral character. We know this because if the process of corruption is so advanced as to lose all moral character and purpose in an institution, such as in authoritarian regimes where corruption leads to the material collapse of legal institutions, we lose the possibility of describing an institution as corrupt, since the original institution does not exist anymore.
Justice is the preferred standard in political theory to describe the moral character of institutions. However, it seems to be less useful in this context, as distributive justice typically assumes full compliance in its conception of justice that is acceptable to all citizens (Rawls 1999(Rawls y 2000. This assumption of full compliance eliminates the possibility of corruption. Legitimacy, on the other hand, is a political virtue that does not require full compliance and that helps evaluate how well institutions use political power to create a system that protects citizens' rights. Legitimacy does not require full compliance, because it is a matter of degrees. 12 For example, past regimes allowed racial segregation until very recently, but did not stop being legitimate in the same way as current regimes which allow discrimination against the LGBT community.
For the case of public institutions, corruption seems to be more related to the moral obligations generated by the existence of laws and regulations with the objective of creating order and protecting rights, 13 which is less stringent for the private sector. Legitimate purposes for private institutions may be constrained by law, human rights and by the purposes established by the organisation itself. Please consider the following formulation of our concept of corruption: Rule-based concept of corruption (CR): An act X is corrupt when it is part of a pattern of behaviour of the same kind X1, X2, X3… Xn, which, as a set, 12 For more on the distinctions between ideal and non-ideal theory, see Valentini (2012,(654)(655)(656)(657)(658)(659)(660)(661)(662)(663)(664). Special thanks to Lucero Fragoso Lugo for this point. 13 For a more detailed discussion on legitimacy, see Williams (2005), and Sleat, (2015, 230-252 damages the validity of the rules that shape the nature, purpose, and services provided by a morally legitimate institution LI. LI is legitimate if it uses political power or enforcement in a morally justified way and it follows preexisting constitutive rules. The validity of LI is damaged when either of the following conditions are true: (i) X1, X2… Xn establish rules and/or conventions that compete with the authority of the alternative set of rules A1, A2, ...An that in turn shaped the nature, purpose, and services provided by a legitimate institution LI.
(ii) The damage to the validity of LI's constitutive rules undermines the ability of LI to fulfil its defining purpose because after the competing pattern is established, certainty is curtailed. Participants of the institutional practices must now decide whether to do X or A.
(iii) X is an abuse of power or authority by a public servant, according to what the institution's applicable legal rules describe, or (iv) X plays an institutional role even though whoever performs X does not perform an institutional liable role within the institution.
Contrary to MC, AC, and the view of corruption as part of organisational culture, our concept of corruption specifies the harm that corruption inflicts upon institutions with both a social and a moral dimension. CR is attractive because it takes into account the social harm of undermining institutional rules explained in AC and corruption as organisational culture, but it also has a moral dimension like MC. Crucially, our concept of corruption provides an explanation for what it means to undermine or help undermine an institutional process or purpose. It does so by explaining the damage in terms of the subversion of institutional rules by an alternative set of rules or conventions that competes or displaces a legitimate one, thus reducing certainty. Corruption does this as an organisational or legal harm and not an exclusively moral one.
Although the mention of "rules and conventions" may suggest that CR only sees corruption as happening within formal institutions with written rules or that the formalisation of institutions could counteract corruption by itself, both interpretations would be inaccurate. Informal institutions, understood as "patterns of patron-client relations by which power is also exercised" alongside formal institutions (Bratton 2007, 96-110) have their own specific rules and conventions, which would be undermined if a parallel set of rules competed with them. Take, for example, the practice of pooling a community's resources to invest in a morally justified project, such as the construction of clean water facilities. If the person in charge of managing the pooled resources were to use them for a different purpose, such as to buy themselves clothes, the informal institution would be corrupted in much the same way as in any case of embezzlement in a formal institution, such as a government ministry. CR sees corruption as a possible scenario in both formal and informal institutions, as long as there is a moral dimension in the institution that is being undermined. 14 Nonetheless, this moral dimension is puzzling. While the damage caused by corruption to an institution is not directly a moral wrong, such as fraud or murder, it has moral implications in that any action or pattern of behaviour that competes with the authority of a legitimate institution's rules undermines that institution's ability to fulfil its defining purpose. Corruption is morally wrong if the institution itself has a moral dimension. If the institution is legitimate, this defining purpose has a moral justification. Therefore, the connection between the damage to the institution's rules and any moral damage is contingent and not necessary. This moral service performed by institutions is what gives society reasons to cooperate with legitimate institutions, and also may justify moral obligations. 15 Perhaps a more concrete way of thinking about the damage done to a rule's validity is to consider that one of the most basic services provided by rules is to create predictability. When corruption damages the law's validity, society is no longer clear on which rules it is supposed to follow, be it as public servants or not. In this context, people's spheres of action become ambiguous as to which situations will favour corrupt and unwritten political arrangements over written and morally justified laws and regulations. As a consequence, since personal agreements are unwritten and only known to the involved parties, it is no longer clear which institutions are actually working towards their written objective and which institutions only exist to benefit certain individuals. Predictability is di-14 Many thanks to Manuel Balán for suggesting this clarification. 15 Note as well that acts of corruption of rules may be justified by strong reasons as in Schindler's case or when someone needs to act corruptly in order to protect or have access to their own rights. But our concept allows to admit that corruption harms rules and institutions even if it is morally justified in some cases. Thanks to Hazahel Hernández for this suggestion. rectly related to the need for planning and measuring the magnitude of the changes caused by specific public policy interventions, as well as the attractiveness of specific markets for investors in the private sector. While exploring the diverse economic consequences of this loss of predictability would be outside the scope of this paper, it would not be a stretch to link it to several of the manifestations of a loss in productivity outlined by Lambsdorff (2003). According to him, corruption prevents "the allocation of capital goods" from being optimal due to its manipulation of conditions such as efficiency, capacity and quality control in bureaucracies (2003,(457)(458)(459)(460)(461)(462)(463)(464)(465)(466)(467)(468)(469)(470)(471)(472)(473)(474). Lambsdorff (2003) gives the example of how while corrupt contracts may be initially lucrative, they are not legally enforceable (due to being illegal, or at least irregular) and are highly susceptible to arbitrary changes in negotiation terms or even cancellation. As such, they are highly unlikely to allow bureaucracies, both in the private and public sector, to maintain standards of efficiency and quality control.
Our concept of corruption requires a much more extensive defence. However, for the purposes of this paper, it is important to point out the three existing concepts of corruption and that at least one of them, CR, has the validity of institutional rules as its object.

VI. corruPtioN As A sociAl coNstructioN
A defender of cultural explanations may very easily point out that our approach is methodologically flawed. Of course we cannot make sense of the cultural causation of corruption by conceptual analysis of normative political theory precisely because corruption is the kind of object that cannot be analysed in that way, as much as racism, discrimination and other instances of domination cannot either. It's true: people sometimes break rules. When an explanation about corruption leads to trivial and circular accounts it is precisely because we need an understanding of the phenomenon (Little1994). Within human communities, actions and beliefs can be understood by interpretation even if they resist explanation. Interpretation attempts to discover the social meaning that actions and beliefs have for agents themselves, even if these do not make sense for conceptual analysis. Crucially, the cultural understanding here is not trivial or circular; as long as it remains interpretive. One of these projects of social understanding is social constructionism. The constructionist project strives to identify contingent social phenomena that in the present are taken for granted but that can be the object of criticism and change (Haslanger 2012;Hackin 1999). Consider the case of Christiana, a widow in the Republic of the Congo. In communities like hers, widowhood puts women in a category of the most marginalized and invisible members of the society. Note that constructionism avoids the kind of determinism we eschew from the folk interpretation in the introduction. While the folk approach will condemn Congolese society for the inevitability of their behaviour towards widows, the social constructionist view stresses precisely the opposite: the social meanings associated with widowhood are contingent and they may very well be different (Haslanger 2012). From the history of Mexico, the social constructionist view extracts a different conclusion: there is nothing about being Mexican that necessarily entails the kind of corruption that Mexicans ordinarily have to endure in their engagement with public institutions. Bratu, Sotiropoulos and Steyanova put it succinctly when they say that corruption (as well as anti-corruption efforts) is "not a self-evident idea, but the outcome of social actions and political intervention" (Bratu, Sotiropoulos & Stoyanova 2017).
But this is too quick. The constructionist account typically identifies thick social kinds of people. These kinds are thick as opposed to thin in the sense that they come associated with significant social implications. The idea is that, for example, being a Latino Spanish speaker in Mexico City is a very thin position, while being a Latino Spanish speaker in the United States typically carries a significant social weight (Haslanger 2012). So the meaning of being a widow or a Mexican varies across social groups. Social constructionism raises a cluster of issues (Haslanger 2012), but for our purposes here, it is enough to point out that it seems very difficult to construct "corrupt Mexicans" as a thick social kind. It seems problematic, or at least unhelpful, to single out those engaging in or affected by corruption as a disadvantaged group, as it would seem that corruption actually affects everyone. We of course do not mean to deny that corruption is instrumental to the disadvantage of the vast majority of Mexicans, or that corruption may be more detrimental to vulnerable social kinds. Rather the point here is that the harm of corruption is not exclusively associated with one social kind, as the meaning and disadvantages of widows is exclusively as- sociated with women who have lost their husbands. Corruption, as a social collective harm, affects entire societies, not just individuals. 16 An alternative to this may be to propose "middle-to-upper management public officials" as a thick social kind associated with clear advantages and privileges determined by corruption in Mexico's social matrix and practices. It is clear to most Mexicans that senior-level public servants often seek their own interest above the public interest, and that people who aspire to such jobs are often more motivated by rent-seeking than by a desire to help their society. Yet these privileges are not directly available to all the members of that social kind, and said members are clearly not exempt from suffering the harms of corruption. For them, as for everyone else, the predictability of the system is compromised, conditioning their rights and their possibilities for professional advancement with their engagement in corruption. This relationship between Mexicans and corruption is not a prerogative of any specific social kind, but a reality for most members of Mexican society. Now consider a different approach. Michel Foucault famously believed that in order to understand political change, we must understand the practice of political power rather than the conceptual analysis of political theory. According to Foucault, political power is one of those objects that cannot be understood as a concept (as in political theory) but rather as a practice (Foucault y Ewald 2003). The practice of power has several mechanisms, techniques and technologies; so when one seeks to determine the nature of a problem, one must identify the implied technologies of power. Recently, Mexican poststructuralists such as Ariadna Estévez (2007) and Sayak Valencia (2010) have employed the Foucauldian methodology to point out that the extremes in global capitalism have taken a toll in the capacity or even willingness of political institutions to protect the basic rights of people. Much more modestly, we could claim that corruption is rather part of a culture of exploitation or domination. Consider the Mexican constitutional institution (Cruz 2000). While the meaning of social institutions for political theory is described in terms of cooperation sustaining fair and just institutions that uphold freedom 16 This does not preclude the possibility of constructing "corrupt Mexican in the USA" as a thick kind. But even if we do this, the understanding of corrupt Mexicans as a whole remains elusive. Thanks to Bernardo Bolaños for this remark. and equality; for Mexicans, social institutions are rather tolerated as long as they preserve peace and/or distribute resources, in exchange for the vote that preserves the status quo. 17 This may have a cultural swing: The purpose of Mexican institutions is not to establish fair relations of cooperation but to extract votes or rents in exchange for loyalty and allegiance (Diaz 1997).
To be sure, consider Mexico's recent history. Years of civil war related to the Mexican Revolution led to the creation of the Constitution of 1917, which was used as a tool to pacify Mexico by including the many, sometimes contradictory, revolutionary narratives in a document to create an artificial consensus. 18 This pacification was acceptably successful because it took into account parts of every group's claims by avoiding any pretence of a pre-existing consensus and by creating areas of ambiguity and discretionary power for specific stakeholders and local elites to reach their own unwritten agreements. These agreements, regardless of their formality or informality, were not meant to ensure social rights or the rule of law, but to pacify caudillos 19 and mobilize groups by making them feel included in the new constitution. Their objective was to stop the fighting by turning the revolution's various contradictory demands into stable institutions, which would fall under the umbrella of the new constitution's legality (Diaz 1997). These unwritten agreements were eventually controlled by Mexico's presidents, first by Álvaro Obregón, when he won the still-fighting caudillos' loyalty by creating the system where the president chooses which candidates got elected. This gave him control over their decisions and policies, as they aspired to new roles once their period of service was over (Castro 2004). Later Plutarco Elías Calles organized the revolution's many factions under a single political party, the National Revolutionary Party, which continued to enforce an unwritten power structure based on backdoor agreements and discretionality (Beezley 2009 Cárdenas brought this discretionary power directly under the hand of Mexican presidents, cementing the rule of his Institutional Revolutionary Party (PRI) for the rest of the 20th century (Serrano 2007). Historically, ensuring equality was not a priority for Mexican politicians.
The upshot is this: If we follow the poststructuralist approach, facts may support the idea that corruption is culturally caused insofar as it is part of the institutional design of how Mexican institutions work. That is, corruption is embedded in the unwritten customary rules that structure Mexican institutions. In Mexican practices and institutional order, informal institutions do, in fact, compete with formal ones. Corruption is formally forbidden and prosecuted by those in power; but in practice, it is allowed, expected, and even required.
A clear example of this situation is the way Mexico's branches of government deal with the federal budget. During PRI's regime, Congress always passed the federal budget without comments. It was (and still is) the responsibility of the Legislative branch to critically examine and challenge the President's budget proposals, though PRI presidents would have the last say on most of the federal budget's changes. When PRI lost the presidency to the right-wing Partido Acción Nacional (PAN) in 2000, President Vicente Fox sought to follow the constitutional procedure and proposed a federal budget to Congress, expecting them to critically examine it. Instead, factions of Congress returned counter-proposals with "locked" sections of the budget through a process called "labelling", 20 which involved letting the President know which parts of the federal budget were non-negotiable. This labelling did not have any precedent in the law. It is particularly noteworthy that at this point in time, Mexico's constitution did not even specify a course of action if Congress failed to pass a federal budget in time, 21 akin to the federal shutdowns seen when this happens in the US. It was a situation that was clearly not expected to happen under PRI's regime. President Fox's actions and Congress' reaction effectively meant that one of the main unwritten agreements, which allowed the federal budget to pass without complications by means of party loyalty, was forever gone and opened the door for an unregulated and mostly unorthodox practice for Mexican politics: lobbying (Ugalde 2014). From that moment on, groups and organisations with no foundation in laws, such as the National Rural Confederation (CNC) and the National Governors' Confederation (Conago) gained impressive influence in budget negotiations by lobbying Congress for changes to the federal budget that might benefit their members. The extent of the influence exercised by informal organisations like CNC and Conago, as well as unregulated lobbying by interest groups, may all be considered corruption. None of these actors work for the benefit of Mexican society and they are able to influence the government through informal means to benefit their own stakeholders, regardless of the moral reasons for the way the federal budget is meant to be negotiated or the effect their changes to it may have on Mexican society.
The climax of the situation happened in 2004, when PRI and PRD members of Congress labelled around 4% of the federal budget for the following year as non-negotiable. This was ten times the labelled amount from the previous year. The changes to the President's proposal were so substantial as to lead PAN representatives to vote against their own party's budget proposal, which was then passed anyway by the PRI and PRD majority. In response, Fox vetoed the 2005 federal budget, while Congress argued that there was no such thing as presidential veto in Mexico (CNG). Although a constitutional amendment in July 2004 allowed Congress to "modify" the President's budget proposal before voting on it, 22 the ambiguous use of the word "modify" effectively allowed the branches of Mexico's government to work on the budget proposal through lobbyists outside morally justified institutional channels. This amendment, unofficially, welcomed the informal influence of private organisations like CNC and Conago. The situation was so ambiguous, that it reached the Supreme Court through a lawsuit filed by President Fox in 2004 (Art. 74 in México Constitution). Almost a year later, in May 2005, the Supreme Court ruled that presidential veto was validly applied to the federal budget, though the court's ruling did not specify the extent to which Congress was allowed to influence the president's budget proposal through labelling (Proceso 2005). This of course makes sense, since labelling has no legal precedent, but did not resolve the core of the matter: deciding what happens when Congress and the President cannot agree on a federal budget bill. Nonetheless, with the opening of the door to lobbyists and the Supreme Court's tacit approval of labelling, negotiating Mexico's federal budget became an unremarkable process again by the end of the Fox administration in 2006. The difference was federal budget negotiations once again occurred in an informal context, parallel to most written regulations and separate from the normative duty to benefit Mexican society.
This use of poststructuralist approaches to identify and contextualize problems is powerful and attractive. The problem is they do not provide grounded relations of causation or correlation. What critical hermeneutics in general reveal is the meaning of social practices. In this case it reveals what corruption means to Mexicans; it does not predict that Mexicans will act in this way or the other because social practices and meanings change over time, and because it does not attempt a causal explanation for corruption. In particular, the poststructuralist approach powerfully identifies that Mexican institutions have departed from the purpose and structure that institutions with a legitimate purpose ordinarily have. This is hardly news, but it begins the discussion rather than giving it a definitive conclusion. In contrast, as we saw, the normative conceptual analysis proceeds in the opposite direction. In the case of CR, it establishes an ideal account of instrumentally valuable institutions as legitimate practices and rules with a purpose that have a moral character. Then, it proceeds to explain how corruption is a morally impermissible deviation from the ideal account by pointing out the harm it inflicts to intrinsically and instrumentally valuable, morally legitimate institutions. The important thing to note is that these are not competing explanations of reality, but complementary accounts. While hermeneutics in general, and poststructuralism in particular, reveal the nature and dimension of the problem, normative conceptual analysis sets a standard. We need both to deal with the problem.
To see how hermeneutics and conceptual analysis may work together, consider the Mexican case again. Hermeneutics reveals the meaning of corruption within the Mexican order: political institutions establish an arrangement where institutions distribute wealth in exchange for peace. Corruption is simply the social practise of securing a bigger bundle of this wealth without leading to widespread instability, or as Felson puts it, "break the rules yourself but get everybody else to follow them" (Felson 2011, 12-17). While this account helps us understand the meaning of corruption within a cultural outlook, we need to take a step back and explain what corruption in general is and how it harms people if we want a chance to establish effective public policy against it.
Once we know what corruption is and we contrast CR to the poststructuralist account of the meaning of corruption for Mexicans, points of contact emerge. From the poststructuralist interpretation we extract the certainty that corruption is a pervasive social practice embedded in Mexican institutions by the establishment of unwritten rules that allow and even require acts of corruption. But from that, it is not evident how this happened or what we can do to fight this. A reason for this is that the poststructuralist account does not clarify the nature of the harm that corruption does to institutions in general. Therefore, we need a more profound and clear idea about the nature of corruption that the poststructuralist or social constructionist account does not provide, as it only works with the meaning that the people from the case study assign to their corrupt institutions. That is when the conceptual normative account kicks in.
If our conceptual analysis is sound, CR is the most coherent way to understand corruption and the nature of the harm it causes, while showing that the cultural interpretation of corruption is incoherent. CR determines that corruption is harmful because a given pattern of corrupt acts may establish a competing set of rules or conventions that displaces compliance to the set of rules that constitute a legitimate institution. This, then, harms the validity of the rules that conform a given institution, either formal or informal, undermining its institutional purposes and goals. At least for the case of political institutions, the nature of this harm has a legal dimension, but since the institution is legitimate and has intrinsic value, this harm also includes a moral dimension. The upshot here is that we can easily suggest a sort of continuity from the social constructionist and poststructuralist interpretation to the conceptual analysis, as CR explains how corrupt acts make the kind of patterns of behaviour that establish harmful parallel sets of rules that become in turn the kind of social practice that hermeneutics observe. That is, the meaning of corruption for Mexicans accommodates our conceptual analysis.

VII. coNclusioN
Hopefully, this paper has managed to outline a useful conceptual analysis of corruption as it may relate to culture, particularly in the case of Mexico. We began by rejecting the folk traditional view about the cultural origin of corruption as implausibly committed with social determinism. In section two, we proposed a political understanding of the cultural origin of corruption, though this understanding cannot deal with the triviality and circularity objections. We challenged whether corruption is caused by culture or rather by a lack of effective acculturalization. The case of Professor Lazlo showed it is possible to be effectively informed about the wrongs of corruption, fully cultured to be a functioning member of society and yet act corruptly in a way that makes it possible to wonder if culture itself determines people to choose corrupt courses of action. Upon closer inspection, our cases seem to suggest that corruption is not culturally determined in the broad sense or in a political sense, but may be, in a restricted sense, relative to morality.
In section three we examine whether corruption could be understood as a moral issue that becomes cultural when culturally determined morality fails to properly socialize individuals. However, the concept of corruption as a moral fault was circular in the sense that it simply assumed that corruption is harmful to institutions, but failed to specify any characteristics, conditions or reasons for this harm. In contrast, a more descriptive concept that identifies corruption with abuse of office for a personal gain correctly identifies that corruption is harmful partly because it constitutes a pattern of behaviour that undermines institutional purposes. The problem is that this concept is cultural in a merely trivial sense: using public office for personal gain is certainly a cultural phenomenon as much as being an exemplary public servant or as anything else, really. In turn, we invoked the notion of organisational culture in an attempt to specify what exactly is cultural about corruption. Upon closer inspection, this concept is not really different from concepts of corruption focusing on morality (MC) or abuse of power (AC). Nonetheless, the idea of a corrupt organisational culture revealed something useful and interesting. What is harmed is not culture in a broad sense or in a political sense, but specifically the rules that determine organisational culture. In turn, what is harmful is the kind of behaviour that deviates from what is expected, according to the rules that determine the structure and purposes of the institutions. Despite this seeming more explanatory than MC, it is underinclusive in the sense that it ignores many types of corruption that do not involve economic benefits, institutional roles or that are not illegal at all. Furthermore, AC and organisational culture explain how deviations from institutional rules may be harmful, but do not explain how this kind of harm is of a special kind different from other offenses like fraud or theft.
In contrast, in section four we offered a rule-based concept of corruption. This concept explains corruption by pointing out what is distinctively harmful about it. Corruption is the harm that a pattern of behaviour inflicts on a society's institutional life when it establishes a parallel convention or set of conventions to the one that is required by the relevant institution, thus reducing certainty and making people wonder what to do and what others are doing.
In section five, we deal with a potentially powerful objection from the theory of social constructionism and poststructuralism. However, both theories only frame and document the depth of the problem rather than establishing a causal link between a corrupt culture and corruption. To be sure, we examine a modest poststructuralist interpretation of Mexico's recent history, particularly focusing on how the federal budget was negotiated before and after the end of one-party rule in 2000, allowing for the suggestion of a continuity between the legal concept of corruption and the patterns of behaviour that may be identified as a culture of corruption in Mexico.
After this analysis, one significant concern is how to deal with corruption and culture. It is one thing to maybe have an approximate understanding of corruption in Mexico, but it would be a whole other thing to approach a remedy for it. Maybe a good first step in bridging this gap between understanding and solving would be to recognise that corruption may not involve the same situations in every cultural context, but it is also not unique or culturally predetermined in every single place.
Take, for example, corruption scandals involving infrastructure. A quick online search on Mexico's Autopista del Sol scandal will deliver many versions of the situation regarding Mexico's 95D federal highway, connecting Mexico City and Acapulco, and the various problems with the exorbitant cost of its planning, building, and repairs. Similarly, an online search for the "Big Dig" project in Boston, Massachusetts (US), also known as the Central Artery/Tunnel Project, will show similar complaints. The two cases, one taking place in Mexico between 1993 and 2007 and the other in the United States between 1991 and 2007, ended with huge cost overruns, 23 and both have been criticized for their delays, and their use of substandard materials, which cause regular accidents and have hiked up maintenance costs, among other grievances. The two cases, if not entirely equal, are equivalent in their size and implications. However, a clear difference comes up in how the Big Dig has been the object of many audits and investigations, leading to financial restitution by companies that admitted to not performing at their best, and the arrests of at least six individuals accused of hiding the inferior quality of the concrete used in the project (Lindsay 2006). Audits and investigations have also been performed in the case of the Autopista del Sol, but they have led to far fewer legal consequences and are themselves surrounded by accusations of corruption that are regularly ignored by law enforcement institutions (Yamashiro 2013). The causes for this are diverse, but they seem to stem from the involvement of powerful people at the federal level, as well as ambiguous regulations in the Mexican states the highway crosses.
At the risk of entering the realm of corruption perceptions in a conceptual paper, it is worth mentioning that although few residents of Boston would describe their jurisdictions as the opposite of corrupt, few of them would also put Boston (or indeed, the US) on the same level of corruption and government ineffectiveness as Mexico. 24 The upshot is this: Both are prominent cases of corruption in their respective countries, but their consequences were different. In the simplest terms, both infrastructure projects were plagued by corruption and irregularities, but those dealing with the Big Dig had clear rules to apply, and those involved with the Autopista del Sol did not. If the harm done by the guilty in each case is similar, as is the moral outrage surrounding it, what is left as the clear difference between these two cases is the rules present in each context. The difference is US citizens, in the case of the Big Dig, were able to trust that their institutions would investigate the problems with the project and punish the guilty. Mexicans, when it comes to the Autopista del Sol, have no such cer-23 Around 190% in the case of the Big Dig and 275% in the case of the Autopista del Sol (Lindsay 2006;Yamashiro 2013). 24 Many thanks to Bonnie Palifka and Louise Shelley for pointing this out. Perceptions of corruption are rarely the same inside and outside of specific cultural contexts. tainty. This difference in how corruption is handled is not cultural, but rulebased, particularly regarding the clarity and pertinence of the rules applied.
If after all the reader is not convinced, consider one final example: take Torgeir and Emmanuela, hailing from Norway and Italy respectively. 25 They are in Hungary, a place where both are foreigners and corruption is not unheard of, and they have determined after careful deliberation that the easiest way to access a right is to lie to a public official. Let us also suppose that they are actually entitled to that right, but it will be costly and time-consuming for themselves and even for the government to have the public office go through the whole process as it is formally regulated. On good faith, they sincerely believe that it will be better for the government and for themselves to save everyone the trouble of playing along with this very obvious flaw in Hungarian bureaucracy. Once they are finally interviewed by the relevant public official, lying comes naturally to Emmanuela, whereas Torgeir is visibly upset and embarrassed by the situation. From that scene, one is tempted to conclude that corruption (at least in the form of lying to the government to access a right) comes naturally to Emmanuela because she is Italian, and corruption is culturally accepted amongst Italians, whereas Torgeir is upset because he comes from a social context, Norway, where the culture of compliance is pervasive. 26 We can accept this explanation only if the reader accepts what the example suggests: that "culturally determined" maybe means that Emmanuela is more likely to engage in acts of corruption than Torgeir, but this does not answer why she is more likely to do so. In contrast, our rule-based concept of corruption may cast some light by pointing out that maybe Italians find conventional engagement in acts of corruption more effective to access the protection of their rights, whereas Norwegians act under the convention that the compliance of established rules assures this access to that protection better than alternative paths. Of course, this is not all there is to say on the matter, but this is where empirical research, especially in the form of case studies that lead to the evaluation and measurement of corruption, must kick in. Our legal concept of corruption serves the purpose of describing what the cultural explanation is unable to: what corruption actually is. 25 Thank you to Itzel Mayans for insisting on this matter. 26 Many thanks to Andrés Moles for suggesting this example.