Recent studies on Latin American politics have convincingly demonstrated that the democratization processes that swept across the region during the last three decades did not bring an end to clientelist politics. On the contrary, the regional democratization processes, which went hand in hand with the neoliberalization of Latin American societies, even enhanced the historically pervasive presence of political clientelism throughout the region (Auyero Reference Auyero2000; Brusco et al. Reference Brusco, Nazareno and Stokes2004; Escobar Reference Escobar2002; Garcia-Guadilla and Perez Reference Garcia-Guadilla and Perez2002; Levitsky Reference Levitsky2003; Penfold-Becerra Reference Penfold-Becerra2007; Scherlis Reference Scherlis2008). It is through clientelist practices, conceived as ‘the giving of material resources as a quid pro quo for political support’ (Stokes 2009: 667), that a growing number of Latin American citizens, confronted with their relegation to the status of ‘second-class citizenship’ (Koonings and Kruijt Reference Koonings and Kruijt2007; Kruijt Reference Kruijt2011) – a position marked by persistent socioeconomic and political exclusion – are struggling on a daily basis to ensure their socioeconomic reproduction in the world region with the most unequal income distribution (Frankema Reference Frankema2009). But ‘second-class citizenship’ is not only a type of citizenship characterized by socioeconomic and political exclusion and marginalization; it is also a decisively violent form of citizenship. In this regard, the expansion, if not predominance, of this form of citizenship reflects that throughout the last two decades or so Latin America has become one of the most violent regions in the world. This ‘new violence’ (Koonings and Kruijt Reference Koonings and Kruijt1999) is most of all related to a dramatic upsurge in violent criminal activities and a general unrule of law, transforming most Latin American states into ‘violent democracies’ (Arias and Goldstein Reference Arias and Goldstein2010). This rise in crime and violence has also left its mark on the character of Latin American clientelism. It has contributed to the emergence of ‘armed’ or ‘violent’ patron–client exchanges in which ‘violence brokers’ (Koonings and Kruijt Reference Koonings and Kruijt2007: 14) and formal political actors negotiate deals regarding the delivery of votes, policy implementation and more general provision of political support (Arias Reference Arias2006; Clarke Reference Clarke2006; Gay Reference Gay1999).
Whereas much attention has been paid to the violent and coercive dimension of clientelism in contemporary Latin America, the closely related role of public security provision and policing within the deepening penetration of Latin American societies by clientelist practices has received strikingly little attention. This lack of interest might be in part explained by the conceptual exclusion of public security from the realm of clientelist exchanges in the related literature. For instance, Kitschelt and Wilkinson (Reference Kitschelt and Wilkinson2007: 11) argue that public goods ‘that are desired by everyone in society and from whose enjoyment no one can be excluded’, including ‘the provision of external and internal security’, ‘can by definition not be traded through clientelist exchanges’. Such arguments, however, ignore that criminologists have long pointed out that internal or public security is hardly a public good in the economic understanding of the term as defined by non-rival consumption and non-excludability. Even in consolidated democracies, ‘public’ security is politicized, ‘targeted’ and unevenly distributed.Footnote 1 In this regard, ‘the purity of security's credentials as a public good in the economic sense do not withstand close enquiry’ (Loader and Walker Reference Loader and Walker2007: 147). This politicized and uneven nature of ‘public’ security, to paraphrase Scott (Reference Scott1990: 4), a notion that better describes the state's own ‘public transcript’ – that is, an image of how the state would wish to see things happening, rather than actually existing security policies – is well reflected in the historical experiences of patron–client exchanges in Latin America. For example, Fischer's detailed analysis (Reference Fischer2008) of citizenship and inequality in twentieth-century Rio de Janeiro highlighted how selective policing and strategic tolerance of illegality became a powerful engine of local patronage politics, linking politicians, local political bosses and favela residents into complex networks through which selective law enforcement (or non-enforcement) and policing were traded for political support. Another telling account of the inscription of public security and policing into the realm of patron–client exchanges in Latin America can be found in Roldán's study (2002) of the violence that haunted Colombia between 1946 and 1966, and which probably killed more than 200,000 people. Focusing on the department of Antioquia, Roldán (Reference Roldán2002: 106) demonstrated how this episode, known as la violencia, is inseparable from partisan clientelism and patron–client exchanges that linked competing clientelist networks with the locally appointed police forces, converting the latter into ‘nothing more than one among a competing array of armed groups, all of whom ultimately answered to private and particular interests and not to the interests of the public’. In a very similar fashion, Kalmanowiecki (Reference Kalmanowiecki2000) stressed the deeply entrenched history of patron–client exchanges throughout the history of modern Argentine policing, which due to the resulting politicization and extra-legal autonomization of police practices subverted, and in the end undermined, the conditions for democratic development in the country throughout much of the twentieth century. In a telling passage reflecting on the role of the police forces during and after the overthrow of the populist government of Juan Domingo Perón in 1955, she writes:
Changes in regimes brought changes in police personnel through clientelist appointments and purges . . . Shifts in the balance of power influenced the police, and police survival depended on loyalty to the regime. Thus when Perón was ousted, the police were unlikely to stake everything on the losing side in order to defend their own interests. By now an autonomous agency not accountable to the rule of law, the police could be put to work for any government. (Kalmanowiecki Reference Kalmanowiecki2000: 210)
In light of such historical evidence, the current lack of interest in assessing the role of policing within the more recent transformations of clientelism in Latin America's violent democracies is surprising. By addressing this void from the vantage point of Mexico City,Footnote 2 this article contributes to the debates on the relationship between clientelism and democracy in contemporary Latin America by illustrating and explaining how and why clientelist practices are increasingly expanding into the realm of policing and public security provision in a democratic political context. It is argued that in a political environment marked by a precarious security situation that designates security as a top priority for the local electorate, the delivery of security and protection services through clientelist channels has become an important resource for politicians interested in building up political support. This ‘societal’ dimension of security-related clientelist practices is embedded in an ‘institutional’ or ‘bureaucratic’ dimension. In the bureaucratic realm, the strengthening and remaking of clientelist structures served as an important means for regaining political control over a highly autonomous police apparatus whose autonomy even increased after the fragmentation of national patronage structures that under the 71 years of one-party rule by the Revolutionary Institutional Party (Partido Revolucionario Institucional, PRI) linked local policing to national politics. By putting these arguments in perspective with related experiences in other Latin American countries, the article demonstrates that, despite undeniably existing path dependencies, the observed patterns are not unique to Mexico, thereby indicating the ongoing and expanding role of clientelist exchanges within the realm of policing and urban governance in contemporary Latin America.
Policing Mexico City: From Authoritarianism To Democracy
A central pillar of the political system that emerged out of the Mexican Revolution of 1917 has been the thorough penetration of social and political relations by clientelist practices.Footnote 3 The type of clientelism that stood at the heart of the Mexican political system during the seven decades of the Revolutionary Institutional Party's one-party rule has been described as ‘authoritarian clientelism’ – that is, a form of ‘political subordination in exchange for material rewards’ in which ‘imbalanced bargaining relations require the enduring political subordination of clients and are reinforced by the threat of coercion’ (Fox Reference Fox1994: 153). These authoritarian patron–client exchanges overdetermined the entire political architecture of the post-revolutionary state, including capital city politics in Mexico City (Cornelius Reference Cornelius1975; Eckstein Reference Eckstein1988; Ward Reference Ward1990: 63–91).
From an analytical point of view, authoritarian clientelism in Mexico City, as throughout Mexico, had both a societal and an institutional, or bureaucratic, dimension. In Mexico City's urban politics under Revolutionary Institutional Party rule, the societal dimension of authoritarian clientelism was most visible in the way the Revolutionary Institutional Party machine organized political support. The basic exchange relationship was rather simple: by offering highly personalized access to urban infrastructure, public services or jobs, the Revolutionary Institutional Party expected electoral support and other demonstrations of political loyalty from the beneficiaries of these services. Local residents participated in these relations because it was impossible to gain access to public resources and state bureaucracies on an impersonal, formal legal basis (Grindle Reference Grindle1977: 39). Under such conditions, the Revolutionary Institutional Party's ‘punishment regime’ that threatened to exclude oppositional actors – politicians as well as voters – from the clientelist spoil system was largely efficient in securing high levels of electoral turnout (Magaloni Reference Magaloni2007: 20). However, as the reference to coercion in the above-mentioned definition of authoritarian clientelism indicates, the state's coercive apparatuses, notably the police, performed a crucial role within the regulation of clientelist relations and the enforcement of compliance with the underlying informal rules of the game as well as the protection of post-revolutionary political order. In Mexico City, this was inseparable from the primary political interest of the Revolutionary Institutional Party leadership ‘to get control of violence and to pacify the city’ (Chevigny Reference Chevigny1995: 237) throughout the post-revolutionary state formation process. While it is beyond the scope of this article to offer an indepth discussion of post-revolutionary policing and patron–client exchanges in Mexico City, the basic pattern of policing that resulted from this predominantly politically motivated interest in safeguarding post-revolutionary order in Mexico City was one that endowed the local police forces with a high degree of autonomy from the political elites. That autonomy, in turn, was used for engaging in a variety of extra-legal policing practices that were politically tolerated as long as the capital city was pacified, politically safe and ‘calm’ (on these issues, see Davis Reference Davis2006, Reference Davis2010a, Reference Davis2010b; Martínez de Murguía Reference Martínez de Murguía1999; Müller Reference Müller2012a: 92–111; Piccato, Reference Piccato2001).
To keep the local police politically ‘in line’, the latter were inserted into the Revolutionary Institutional Party regime through a mutually reinforcing process of patronage politics and militarization. Although, historically, military commanders headed the capital city police apparatus, under the presidency of Lázaro Cárdenas (1934–40) the police were officially militarized in 1939; a decision that in the words of then-police chief Manuel J. Núñez, served to ‘incorporate the administration of the police within the army, so that it will become subject to military law and discipline’ (quoted in Davis Reference Davis2010b: 149). For the rank and file, this decision implied that they received the same legal status as members of the military, denying them the right to strike or to hold their superiors accountable for on-the-job mistreatment (Davis Reference Davis2010b: 147–9). As a consequence, the formally existing civil service career system was de facto turned into a pyramidal system of patron–client relations that made the rank and file directly dependent on the private and political interests of their superiors-cum-patrons, with the police chiefs appointed by the Mexican president at the top of this hierarchical structure. This created a form of ‘bureaucratic clientelism’ (see Lyrintzi Reference Lyrintzis1984) in which employment opportunities within the police were granted on the basis of political consideration and loyalty, not merit or qualification. Hence, historically, local police hierarchical relations ‘have been characterized by the unconditional loyalty and obedience expected of police [agents] toward their superior officers’ (Uildriks Reference Uildriks2010: 138; see also Müller 2012a: 82).
These outcomes reflect a general pattern of autonomous, authoritarian and politicized policing, prone to politically tolerated extra-legal activities that Saín (Reference Saín2010: 29–46) described as the ‘traditional model’ of Latin America policing, which could be observed throughout the region from the nineteenth century until the second half of the twentieth century. But if in the case of Mexico City, as well as in other Latin American countries, the underpinning logic of this pattern of policing was directly related to the prevalence of authoritarian and non-democratic political structures, it should be expected that the democratization of local and national politics throughout the last three decades or so would make a difference and eliminate these informal and politicized policing practices. This, however, was not the case. In fact, when Cuauhtémoc Cárdenas from the Party of the Democratic Revolution (Partido de la Revolución Democrática, PRD) was elected as the mayor of Mexico City in 1997 (the first elections for the local head of government since 1928), it soon turned out that democratization had a decisively negative impact upon the local security situation as it undermined the existing patron–client networks. The local democratization process virtually destroyed previous levels of political control and governability of Mexico City policing, leading to growing insecurity as well as seemingly uncontrollable manifestations of predatory and illegal police activities by enhancing the autonomy of the police forces vis-à-vis the Mexico City government (Davis Reference Davis2010a: 50).
The Persistence of Bureaucratic Clientelism
These circumstances represented a serious obstacle for the Party of the Democratic Revolution governments that have ruled Mexico City since 1997, in their efforts to confront local security problems that have been on the rise since the mid-1990s – and for meeting the related expectations of the local population – as they made necessary structural attempts at police reform extremely difficult. Police reform efforts, such as the attempted introduction of lie-detector tests in 1997, were frequently met with outright resistance from within the police apparatus, including veritable ‘strikes’, leading to rising crime rates in the following days and weeks (Davis Reference Davis2003: 20). Such experiences – demonstrating the existence of powerful ‘resistance enclaves’ (Bergman 2007: 90) inside the local police that were fearful of losing political and (illicit) economic privilegesFootnote 4 – made it clear to the leadership of the Party of the Democratic Revolution that a thorough reform of the Mexico City police was an unrealistic goal in the short term. Hence, behind the official police reform rhetoric, and in light of the persisting local insecurity problems, the local governments sought to establish a modus vivendi with the local police that would re-establish the political governability of the local police forces by curtailing their recently enhanced autonomy. In fact, as in other Latin American countries, the recently elected democratic incumbents and their security policies were primarily interested in establishing political control over the local police forces by regaining their political loyalty, and, as Quintana Taborga (Reference Quintana Taborga2003: 16) summed it up in a paradigmatic way for Bolivia, ‘Police clientelism was one of the most efficient tools of these patrimonial politics’.
Although it is difficult to tell if in Mexico City a consciously designed clientelist strategy was at work – or if the local decisions were of a more contingent and ad hoc nature, driven by the double need to control rising crime and regain a certain level of political control over the police – all mayors that followed Cárdenas (who stepped down from this post in 1999 in order to run as a candidate in the 2000 presidential elections) pursued a predominantly cosmetic or symbolic police reform agenda. To this end, reform mainly focused on territorial reorganization, new police training programmes, the renaming of police institutions and actions that demonstrated a publicly visible commitment to police reform and international ‘best practices’, but which left the patronage-based workings of power inside the police largely intact.Footnote 5 This contributed to the persisting ‘presence of more or less hidden informal networks and decision-making structures [that] outweigh the importance of formal bureaucratic rules and structures’ inside the local police forces (Uildriks Reference Uildriks2010: 204); it is a situation that has also been observed for other Latin American countries where ‘the pressing need to reform the police and institutionalize effective state oversight has been consistently at odds with the old patterns of patronage, clientelism and impunity and rule by fiat that continue to thrive decades after the transition from military rule’ (Hinton Reference Hinton2006: 194).
In Mexico City, the withdrawal from indepth and large-scale police reform initiatives was successful in that it ‘pacified’ the relationship between police and the local government and re-established the political loyalty of the police forces. The price that had to be paid for this ‘pacification’ was the persistence of patron–client structures inside the police, and the latter even expanded due to an increase in policing and security-related expenditures by the local government. Between 1998 and 2006, the total expenditures of the Department of the Secretary of Public Security of the Federal District (SSPDF), in charge of the Preventive Police, and of the Department of the Attorney General of the Federal District (PGJDF), in charge of the Investigative or Judicial Police, increased from 7,126 (SSPDF) and 1,938 million pesos (PGJDF) in 2000 to 13,393 (SSPDF) and 4,057 million pesos (PGJDF) in 2006.Footnote 6 Additionally, between 2001 and 2005 the number of police personnel increased from about 60,000 to about 80,000 (Uildriks Reference Uildriks2010: 197). This influx of resources substantially enhanced the pool of incentives – jobs, money and equipment – available for distribution along persisting patron–client structures, reinforcing the pattern of bureaucratic clientelism that characterized capital city policing under Revolutionary Institutional Party rule by offering employment opportunities within the police bureaucracies for political supporters; it is an outcome that has been facilitated by the arbitrary distribution of security funds in Mexico City, where neither accountability mechanisms, nor transparency, nor clearly defined schemes of expenditure exist (Alvarado Mendoza Reference Alvarado Mendoza2008: 46). As one member of the local law enforcement bureaucracy summed it up, ‘clientelism is the currency of change [moneda del cambio] and career advancement’.Footnote 7
It is this predominantly politically motivated, and not merit-based, pattern of personnel recruitment that also explains why it is not uncommon to find people with dubious ‘law enforcement’ backgrounds holding important positions in the local police apparatus (see, for example, La Crónica de hoy 2008; La Razón 2011; El Universal 2007). What makes such staffing decisions functional within the local context is that indepth knowledge about and experience with the extra-legal practices of the local police forces – both inside the police institutions and ‘on the streets’ – is essential for the maintenance of urban order. In other words, ‘managing’ urban political order requires knowing how to play the game of informal and extra-legal politics and policing that continues to dominate local law enforcement. In this regard, it is telling that in 2010, 13 years after the end of Revolutionary Institutional Party rule in Mexico City and permanent ‘democratic’ police reform proposals, the local police chief publicly declared that his institution was ridden with ‘corruption’ and ‘administrative disorder and abandonment’ (La Jornada 2010).
While police corruption should be differentiated from clientelism, for the purpose of the present article it is important to understand their close relationship, which converts police corruption in Mexico City into what, following Darden (Reference Darden2008), can be defined as a form of ‘state-strengthening graft’. This notion describes a type of corruption that reinforces and strengthens the state's bureaucracy through informal, graft-related practices. Such strengthening can be explained with the fact that in a corruption-ridden bureaucracy, more than in a strictly formal legal type of bureaucratic organization, the complete livelihood and freedom of the individual bureaucrat ‘are dependent on the continued goodwill of his superiors within the state hierarchy. Hence, the threat of exposing and enforcing his wrongdoings constitutes an enormously powerful sanction and places lower level officials in an especially vulnerable position’ (Darden Reference Darden2008: 42). However, the crucial element in ensuring such outcomes, according to Darden (Reference Darden2008: 43), is the ‘problem of monitoring and securing the loyalty of the “monitors”’, an element that he does not further address in his article. It is at this point that clientelism comes into play. While the literature on clientelism in Latin America acknowledges the centrality of monitoring clientelist exchanges, this issue has mostly been analysed from the vantage point of politician–voter relations, not within the bureaucratic realm. Nevertheless, basic insights from the analysis of electoral ‘monitoring’ practices are helpful for understanding how clientelism transforms police corruption in Mexico City into ‘state-strengthening graft’. In this regard, Stokes's argument (2005: 324) that clientelist political machines ‘use their social proximity to voters to monitor their actions and types and hence to enforce the implicit [clientelist] redistributive contract’ is especially important. Her observation regarding the centrality of social proximity can also be applied to the bureaucratic realm in general and police institutions in particular. In the latter case, research has highlighted that ‘cop cultures’ are in general marked by high levels of internal solidarity and isolation from the wider society (Reiner 2010: 122–3). In light of the earlier mentioned characteristics of bureaucratic clientelism inside the Mexico City police forces, where extreme levels of personal loyalty to and dependence from immediate superiors are the norm, social isolation and internal ‘solidarity’ convert superiors into what, following Scott (Reference Scott1972: 93–4), can be called ‘monopolist-patrons’. Their existence is indicative of highly imbalanced and coercive patron–client relations that provide patrons with powerful sanction mechanisms that assure high levels of compliance from their clients. It is this high level of compliance and discipline, stemming from the bureaucratic clientelism persisting within the local police that converts police corruption into ‘state-strengthening graft’. This, however, is not only because of the related internal bureaucratic strengthening processes as identified by Darden (Reference Darden2008). Bureaucratic clientelism also guarantees that ‘“non-budgetary” government finance’ (see Znoj Reference Znoj2008: 66–7), which is to a large extent obtained through extra-legal police activities, is not completely ‘privatized’ by the involved officers. Bureaucratic clientelism – and the underpinning informal control, discipline and sanction mechanisms – assures that these extra-legal resources are distributed inside the police apparatus in such a way that they compensate for financial and budgetary shortcomings, thus adding another dimension to graft-related extra-legal strengthening of the state (for a similar finding in Argentina, see Dewey Reference Dewey2011: 27). When considering, for example, that after 2006 the SSPDF's overall share of Mexico City's government budget decreased from 13,393 million pesos in 2006 to 11,450 million pesos in 2011,Footnote 8 in spite of a largely unchanged personnel strength, a 4 per cent salary increase in 2009 and ongoing investment in technical modernization, it seems evident that patron–client networks are an important mechanism for channelling ‘street money’ into ‘informal government expenditures’ for the basic working of the police, such as the purchase of fuel by the officers themselves.Footnote 9 In contemporary Mexico City, such street money acquisition is informally divided up among the different police forces – according to their legal authority (Alvarado Mendoza Reference Alvarado Mendoza2008: 55) – who charge a certain amount of money, frequently a fixed sum, the so-called renta, for turning a blind eye to illegal activities ranging from informal street vending of pirated consumer goods to the resale of stolen car parts to drug trafficking. This renta is distributed throughout the police apparatus (Müller Reference Müller2012a: 185–8).
It could be expected that such self-financing practices also spill over into the governing Party of the Democratic Revolution, as they did under Revolutionary Institutional Party rule (see Pimentel Reference Pimentel2000), and as has been observed for other Latin American settings, such as Argentina. In the latter case, research on the province of Buenos Aires has shown that a part of the income derived from illegal gambling, prostitution and drug trafficking also goes to finance one of the most important parties in the country: the Peronist party (Dewey Reference Dewey2011: 13–16; Saín Reference Saín2002: 90, 2004: 54–60, 2008: 126–8). Such forms of illegal party financing, however, seem to be absent in contemporary Mexico City. Although it is beyond doubt that some individuals within the police, the Party of the Democratic Revolution and the city administration derive economic benefits from such practices, a veritable extra-legal police-sponsored party-financing mechanism does not exist. This outcome, I would suggest, can be explained by recent transformations in party- and campaign-funding laws in Mexico, where legal changes throughout the democratization processes increasingly made party and campaign funding overly dependent on public subsidies that by far outmatch private funding (Rodríguez Alosno Reference Rodríguez Alosno2011), which is now also restricted by law (Instituto Federal Electoral 2011). The development substantially decreased the attractiveness, and need, of such illegal financing methods in the local context. In Argentina, on the contrary, public party financing ‘represents a modest fraction of the resources that relevant parties spend in electoral activities. Instead, campaign resources primarily come from private entities and the state machine thanks to the dual control that incumbent governors have over public contracts, and the future of political and bureaucratic careers’ (Lodola Reference Lodola2010: 185). Compared to the Mexican case, this situation is obviously more conducive to the above-mentioned ‘private’ party-financing practices through police-operated protection rackets.
Whereas the previously described situation seems to compromise the democratic commitment of the Party of the Democratic Revolution and local security policies in general, it should be recalled that they have also enabled the local governments to curtail the enhanced autonomy of the police, thereby regaining a certain level of control over local policing. As a result of this, the Mexico City government has been able to confront, or rather to ‘manage’, some critical aspects of the local insecurity problems. This seems to be one crucial element behind the fact that (so far) Mexico City, although undeniably one of the country's most important internal drug markets, has not suffered from the same levels of drug-related violence that are haunting many other parts of the country.Footnote 10 Another factor contributing to this outcome is related to efforts by local actors in the informal economy to put pressure on the Party of the Democratic Revolution to keep drug trafficking organizations out of their neighbourhoods and business spaces, or at least ‘under control’. Such efforts benefited from these actors’ privileged informal access to the Party of the Democratic Revolution, which depends on their political (and electoral) mobilization capacity and is therefore willing to provide informal and formal means of repression for containing violent entrepreneurs such as drug traffickers.Footnote 11
Security and Clientelism: The Societal Dimension
The practices and developments mentioned above not only preserved and even expanded patron–client relations inside the Mexico City police institutions; they also contributed to and were embedded in a more general move towards the societal expansions of policing-related patron–client exchanges, an outcome that has been made possible by the continuing reliance of capital city politics on clientelist relations under the Party of the Democratic Revolution (Bruhn Reference Bruhn2008: 136–7; Durand Ponte Reference Durand Ponte2007; Hilgers 2011, 2008, 2009; Schütze Reference Schütze2005). This persistence of clientelism in local politics is related to the internal structures and workings of power within the Party of the Democratic Revolution. Due to the party's origins, which emerged from a split within the then-ruling Revolutionary Institutional Party in 1988, the party initially integrated a large number of former Revolutionary Institutional Party members and Revolutionary Institutional Party-style machine-politics into internal workings of power, creating a legacy that can still be observed today. Therefore, even in a democratic context, for many Mexico City residents, ‘the dominant method of securing public aid continues to run through clientelistic exchanges’ (Bruhn Reference Bruhn2008: 136).
That the provision of security has become an important resource within such exchanges reflects the efforts by Mexico City residents to search for security provisions outside – largely inefficient, if not absent – officially sanctioned channels in an increasingly insecure urban environment, as well as demonstrating that brokers and politicians were receptive to these demands (see also Müller Reference Müller2012a: 108–10, 2012b: 852–3). Much of these policing- and security-related patron–client exchanges take place through the recent expansion of citizen participation programmes within the realm of policing. For example, the related practices include the clientelist appropriation of community policing programmes, like the so-called policía de barrio project (Müller Reference Müller2010), as well as the related accountability and coordination institutions, the so-called neighbourhood committees. The latter are legally entitled to articulate the security interests of the respective communities and have an annual budget of about 500,000 pesos at their disposal for local crime prevention activities. This converts them into highly attractive ‘booty’ for local brokers and politicians, which appropriate the committees and channel the related resources to their followers in exchange for political support. As one interview partner summed it up: ‘In reality, the committees are only a window dressing figure without any real authority or [policy] instruments at their disposal . . . As always in this country, the problem behind all of this is that the neighbourhood committees are prey for the political parties. Everything that happens inside [the committees] is because of political interests.’ These interests, he went on to explain, are centred on efforts by local politicians and functionaries interested in ‘making a political career’ to monopolize the access to local communities ‘in order to use this access for creating a political platform for their [political] ambitions’.Footnote 12 This is even implicitly acknowledged by the local government and police authorities, as the following disclaimer printed in a leaflet informing local residents about the merits of citizen participation in security matters indicates:
This programme is of public character. It is not conveyed nor promoted by any political party and its resources come from the taxes paid by all taxpayers. It is forbidden to use this programme for political and electoral goals, for pecuniary rewards or any other purposes different from those established in the programme. Anyone that will make an improper use of the programme's resources in the Federal District, will be sanctioned according to the respective law and from the responsible authority. (SSPDF 2009)
The threat of such sanctions, however, is more symbolic than real. The informal appropriation of ‘public’ security resources is a quotidian practice (see also Tello Reference Tello2012: 19) that can also include the provision of security-related infrastructure – like more street lighting or a greater police presence – through the privileged access of political brokers to the borough administrations (delegaciones), in exchange for political support (Müller Reference Müller2012b: 851–8, 2012c: 335–7). The crucial role of political brokers in mediating such security policies at the neighbourhood level has also been unofficially recognized by the respective bureaucracies. As one interview partner explained, it is common to inform and negotiate with local leaders of social movements or the informal economy over the implementation of citizen participation programmes, because without their support it would be difficult to implement such programmes successfully on the ground.Footnote 13
Additionally, brokers and local politicians also engage in more informal surveillance and security practices, for instance, the organization of informal security guards (vigilantes) that ‘police’ certain areas, in particular in marginalized neighbourhoods, in exchange for political support and money, which is to a large extent used to provide the vigilantes with some form of income. The following quote from an interview with a resident in a marginalized Mexico City neighbourhood illustrates the related practices:
They [the vigilantes] were created when Arturo SánchezFootnote 14 was president of the neighborhood. He called it a proposal for vigilance. At the beginning there were about 30 of them and they were accepted because they were more constant in their surveillance [than the regular police], but what was not good is when they became bodyguards of the low-ranking politicians . . . They still come around and some people give them some money and others don't.Footnote 15
Although some residents feel uncomfortable with such informal and politicized security arrangements – which provide patrons with multiple resources that can be traded for political support, including the provision of income, some sort of employment and protection – this is mostly because the involved security providers themselves are responsible for ‘taxing’ local residents for their services. While the related practices are sometimes perceived as threatening, it should be highlighted that I did not come across cases in which such informal arrangements were also used for coercing residents to vote for local politicians, a situation that has been observed for urban Brazil where drug traffickers and militias are engaged in such coercive forms of clientelist exchanges (Arias Reference Arias2006; Moreira Alves and Evanson Reference Moreira Alves and Evanson2011: 20, 93–9, 131–3).Footnote 16 This outcome not only reflects that drug trafficking has not penetrated the political environment in Mexico City as it has in urban Brazil, it also demonstrates the fact that coercion and the use of force ‘never formed an accepted element of clientelism in the party [the Party of the Democratic Revolution]’ (Hilgers Reference Hilgers2009: 55). Furthermore, this reluctance of the Party of the Democratic Revolution to use outright force in clientelist exchanges in Mexico City might also explain the absence of clientelist forms of collective violence based on party–broker–police linkages that have been documented by Auyero (Reference Auyero2004: 77–86, Reference Auyero2007) in his analysis of the 2001 food riots in Argentina.
While the basic deal involved in such forms of security-related patron–client exchanges in Mexico City is that the recipients are expected to offer political support to their patrons and brokers for providing services and security, it is important to stress that political support not only takes the form of classic vote buying. Although the issue of vote buying, which dominates much of the related literature, is undeniably relevant, demonstrating political support in exchange for service provision implies more everyday practices. In fact, elections are only one element within the involved patron–client exchanges and the ‘routine party politics’ that underpin the quotidian workings of political clientelism through which brokers solve everyday problems of their clients and in doing so ‘accumulate political capital that helps them advance in the political field’ (Auyero Reference Auyero2007: 55–71). Such political capital accumulation is inseparable from public performances that demonstrate that the broker is capable of mobilizing a loyal political following: for instance, through the latter's participation in political rallies, political festivities and/or, most frequently, by putting their signatures on a list compiled by their broker that he can show to politicians as evidence of his support base. These signs of support are used by brokers and/or politicians for enhancing their standing and position either inside the faction-ridden Party of the Democratic Revolution or, in case of an ‘independent’ broker, by offering support to competing Party of the Democratic Revolution politicians in exchange for access to jobs, public services and/or money.Footnote 17 It is through these more quotidian activities rather than through direct (and due to the secret ballot hard to control) electoral means that resources, including security, are provided and political power is accumulated through patron–client exchanges. Reflecting the ‘double-edged sword’ of fiscal decentralization in Mexico (Cornelius Reference Cornelius1999: 12), this development has been facilitated by Mexico City's growing budget throughout the last decade, as well as the expansion of a variety of comparatively decentralized citizen participation programmes. Both developments provide a permanent flow of resources – to a substantial degree financed by federal fiscal transfers that increased from 5,956 million pesos in 2000 to 12,627 million pesos in 2006Footnote 18 – that are largely independent of immediate electoral results, but which enable political entrepreneurs at the centres of resource distribution to become political patrons through the provision of jobs and access to security in return for political support.
Within the faction-ridden Party of the Democratic Revolution, this political support is the most important resource for upward political mobility and survival, which is most visible in the fact that ‘candidates are courted purely on a calculation of who heads the best clientelist networks’ (Hilgers Reference Hilgers2008: 152). Accordingly, it has been the expansion of decisively local government programmes that has facilitated this tendential decoupling of patron–client exchanges from immediate electoral results – although a ‘good’ neighbourhood broker is still capable of delivering votes – and converting them into important means through which inner-party factional conflicts are fought out and power relations modified.
It is, however, understandable that local residents are receptive to such forms of security-related patron–client exchanges provided through Party of the Democratic Revolution channels, when considering the incapacity of the Mexico City government to provide impartial and efficient public security. In this context, ‘Citizens unable to make such clientelist alliances are clearly at a disadvantage’ (Hilgers Reference Hilgers2008: 134), a fact that people are well aware of. In the words of one interview partner: ‘Yes, there are many people who are close to the Party of the Democratic Revolution here, and I, who belong to the Party of the Democratic Revolution, well, the party gives me . . . well, it protects me’.Footnote 19 Consequently, as long as such protection also implies mediated access to ‘public’ security, joining clientelist alliances aimed at gaining access to these resources continues to be an important element within local residents’ everyday survival strategies in Mexico City's contemporary political and insecurity context.
Conclusion
From the vantage point of Mexico City, this article has demonstrated that, far from simply becoming an outdated authoritarian practice, political clientelism continues to be a core element in Latin America's democratic urban environment and has even expanded to the realm of public security provision and policing. This expansion was related to the efforts of the local government to regain previously existing levels of political control over the local police forces that had been undermined by the fragmentation of long-standing existing national patron–client structures under authoritarian rule. On the other hand, this expansion also reflects the capacity of clientelist politicians and brokers to adapt to the new security needs of the local population, largely left unsatisfied by the official security policies of the local government. Although this article has further highlighted that some dimensions of the resulting clientelism–security nexus result from idiosyncratic aspects of the local context, comparative evidence shows that the Mexico City experience is not unique and that similar processes are at work in other Latin American countries as well. In this regard, it can be argued that patron–client relations survived the transition to democracy in Latin America most of all because existing sociopolitical and institutional conditions continue to make it attractive and rational for all actors involved – bureaucrats, police agents, politicians and local residents – to participate in such exchanges, be it for the material benefits that can be derived from extra-legal policing, the accumulation of political power, or as everyday security strategies aimed at coping with the deficiencies and fragmentation of local policing. As these are mutually reinforcing processes, evidence from more than one decade of democratic governance in Mexico City demonstrates that much more than just formal democratization is needed for it to break out of this clientelist circle and its impact on policing in the near, and possibly distant, future.
