During the Argentine winter of 2009, I was returning from a two-hour interview with Pablo and José,Footnote 1 sharing a taxi as we headed back to Buenos Aires from La Plata. As soon as we got into the car, both men started making phone calls. One of those conversations went as follows: “How many?” asked José; someone replied on the other end of the line. “Great! Thanks!” he responded in excitement and hung up. Then Pablo asked, “So? How many?” “Fifteen!” replied José, with obvious satisfaction. He continued, naming potential recipients, “María, Cecilia, Susana …, ” while counting on his fingers. Then, looking in my direction, he added, “You see? This is political activism – live! (¿Ves? Esto es militancia ¡En vivo y en directo!).”Footnote 2 In my most innocent voice, I asked, “How many what?” While Pablo seemed quite uncomfortable to disclose the information in my presence, José quickly replied, “Social welfare benefits! (¡Planes sociales!).”
Pablo and José are public sector employees, and both are also important Peronist political brokers (punteros or referentes).Footnote 3 Both obtained their jobs in the public sector through political networks. José became politically active in 1995, at age eighteen, when he “knocked on the door at a base unit” in his neighborhood.Footnote 4 He was an activist “for free, for a long time, out of vocation (milité bastante tiempo gratis, por vocación).” In 1999, he left his job in the private sector and lived on unemployment benefits for more than a year, “waiting to be able to start having a salary, quote, unquote (entre comillas), from politics, from the administration.” In 2000, he finally got a public sector job. Pablo, who “has always been a Peronist,” became more involved in politics when he married in 1994. His mother-in-law ran a base unit. At the time, he was working in “nothing related to politics, no public employment, nothing (nada que ver con la política, ni empleo público, ni nada).” He eventually obtained a public sector job as well, through a politician his mother-in-law worked with.
Along with working their regular jobs in the public administration, José and Pablo are active participants in politics: they help with electoral campaigns, disseminate the party message, mobilize voters for elections and rallies, help their clients solve problems, distribute handouts and food, and, as the opening anecdote illustrates, participate in the discretionary distribution of welfare benefits to the poor. They are, in other words, paid political workers. They occupy public sector positions that they would not have received had they not been politically connected in the first place. In return, they work for the machine.Footnote 5 In José’s words: “The activists (militantes), we get a salary. We work the hours we are supposed to work, we are not ñoquis, but I still feel that I owe my salary to politics.”Footnote 6 Of course, not all (or even most) public employees are political brokers, but many other public employees are also involved in the provision of political services – they too obtained their public sector jobs with the understanding that they would provide political support in return. While not all public employees are involved in providing these services, the use of public employment to fund the salaries of political workers is certainly not unusual. The goal of this book is to shed light on this issue: patronage, or the exchange of public sector jobs for political support.Footnote 7 In the following pages, I provide a detailed description of what patronage employees do in exchange for their jobs, as well as a novel explanation of why they do it. In sum, this book aims to understand the specific mechanisms behind the electoral returns to patronage politics.
1.1 Patronage, Electoral Competition, and the Quality of Democracy
The story of Pablo and José, two public sector employees who engage in political work in exchange for their jobs, illustrates some of the challenges facing new democracies. As a consequence of the “third wave” of democratization (Reference Iacoviello and EchebarríaHuntington 1991), more people have the right to vote today than at any other time in history. However, even across democracies, the extent to which elections are free and fair varies enormously.Footnote 8 In Latin America and elsewhere, it is not unusual to find democratic countries where the electoral playing field is skewed in favor of the incumbent party. The mechanisms supporting this imbalance are diverse, ranging from openly illegal practices such as government-sponsored election violence (Reference HassanHafner-Burton, Hyde, and Jablonski 2014) and voter intimidation by state employees (Reference Mares and YoungMares and Young 2018, Reference Mares and Young2019a) to more subtle ones, such as the use of government advertising to reward and punish the media (Reference BrownBrown 2011), the manipulation of the electoral calendar (Reference Oliveros and SchusterOliveros and Scherlis 2004), electoral reforms (Reference Calvo and MicozziCalvo and Micozzi 2005), and party finance regulations (Reference Grzymala-BusseGrzymala-Busse 2003, Reference Grzymala-Busse2007) that benefit the incumbent party. When such mechanisms exist, elections might still be competitive, but they are less fair for the opposition and less free for voters.
One of the mechanisms that undermine the quality of democratic politics in many young democracies is the political use of public employment, or patronage.Footnote 9 In general, the use of any state resource for electoral benefit gives the incumbent party an unfair advantage at the polls. But patronage is particularly troubling for two other reasons. First, patronage appointments create opportunities for other illegal uses of state resources, such as political corruption, pork-barrel spending, and other forms of clientelism. Appointing friends to jobs in the public administration makes it easier for politicians to access resources and distribute them according to political criteria. Second, patronage has the potential to affect both the quality of electoral competition and the quality of democracy more broadly. Like other forms of clientelism, patronage arrangements are intended to generate political support. When successful, they affect the quality of political competition, generating the same negative consequences as other forms of clientelism in terms of accountability and representation. But the fact that patronage exchanges necessarily involve the use of state resources makes them fundamentally different.Footnote 10 Political bias in the distribution of public sector jobs affects the independence of public administration and raises serious concerns about unequal treatment by the state. Patronage can thus have a more damaging effect than other forms of clientelism on the quality of democracy, which suffers because of the effect of patronage on both political competition and on access to the state.Footnote 11
This book describes what public employees do that affects both electoral competition and the quality of democracy, and also establishes why they do it. I argue that patronage jobs are distributed to supporters in exchange for a wide range of political services – such as helping with campaigns and electoral mobilization – that are essential for attracting and maintaining electoral support. Yet those who receive public sector jobs with the understanding that they will provide political services in return can easily renege on their side of the agreement after getting the job. Why would public sector employees comply with their side of the patronage contract once the job is in hand? Existing explanations are based either on fear of punishment (clients comply with their side of the agreement because they are afraid the patron will cut off the benefit if they fail to do so) or feelings of reciprocity (clients comply with the agreement because they want to help the person who has helped them).
By treating both patrons and clients – in this case, politicians and public employees – as equally sophisticated and self-interested individuals, my argument departs from existing accounts that tend to portray clients as passive, nonstrategic, or myopic actors.Footnote 12 What makes patronage contracts self-sustaining without punishment or reciprocity is that patronage jobs are distributed to supporters whose fates are tied to the political fate of the politician who has hired them. An incumbent politician (the patron) will maintain patronage jobs held by supporters (and protect their working conditions) while a competing politician will not, because supporters of the incumbent cannot credibly commit to provide political services for the opposition. Because their jobs and/or working conditions are on the line, supporters have a big incentive to help the incumbent stay in power by providing political services, which makes their original commitment to provide such services credible. This alignment of interests between patrons and clients (or politicians and patronage employees) makes patronage contracts incentive compatible and therefore self-sustaining.
I test the empirical implications of this theory using individual-level data from Argentina. By combining in-depth interviews with public sector employees, political brokers, and politicians with list and survey experiments embedded in a survey of 1,200 local public sector employees, I provide a detailed picture of how patronage works. I describe what public employees do in exchange for their jobs and explain why they do it. I find that a large proportion of public sector employees – particularly those who support the incumbent – are involved in political activities because they believe their jobs and working conditions are tied to the political success of the incumbent who hired them.
1.2 Defining Clientelism and Patronage
Throughout this book, I define clientelism as the personalized and discretionary exchange of goods or favors for political support.Footnote 13 I refer to patronage as a subtype of clientelism, in which the good that is exchanged is a public sector job. In a clientelistic exchange, either the patron promises to deliver the benefit after the client provides political support or the patron provides the benefit with the expectation that the client will provide support in the future. The actual exchange only needs to be expected for the clientelistic arrangement to exist. It is always possible for clients to take advantage of the benefit and then renege on the clientelistic agreement by not providing the promised support. In turn, patrons can also fail to comply with their side of the bargain by requesting political support in advance and then failing to deliver the promised benefit.
Political support involves a number of different political activities (or services), ranging from the simple act of voting to participating in political meetings, attending rallies, or helping with electoral campaigns. In return, patrons provide different types of goods, ranging from handouts, favors, and money to unemployment benefits, housing subsidies, and public sector jobs.Footnote 14 The type of good or favor itself does not determine the existence of a clientelistic exchange.Footnote 15 Although all the examples here are private goods, it is of course possible to distribute private goods through programmatic politics rather than through clientelism. In fact, as Reference Strazza and LongoStokes (2009) notes, in advanced democracies most welfare-state policies are targeted yet programmatic.Footnote 16 In programmatic politics, a set of rules stipulates the conditions under which citizens should receive certain benefits; all citizens who satisfy those requirements actually receive the benefit, without any bias or discretion. The group of people is defined in abstract terms – for instance, the poor or the unemployed – and all people thus defined receive the benefit. No member of that abstractly defined group of beneficiaries is excluded.
The characterization of the clientelistic exchange as personalized or individualized (and often face-to-face) helps distinguish clientelism from other forms of distributive politics, such as pork-barreling, in which the exchange involves a group of voters. In contrast to pork-barrel politics, where everyone who lives in a certain area would receive the benefit (e.g., electricity, access to water, or a road repair), in a clientelistic exchange, specific individuals can be targeted for benefits, and those who are not part of the exchange can be excluded. To be clear, politicians often target benefits to particular segments of the population with the hope or expectation that members of the benefiting groups will be more likely to vote for them. Such targeting is not personal, however, but rather focuses on a group that can be defined in abstract terms (as in programmatic politics) or by geography (as in the case of pork-barrel politics). In a clientelistic exchange, the benefit is not merely targeted, but targeted at the individual level.
At the same time, patrons know who receives the benefit. In the modern version of clientelism, clients usually have little or no personal contact with the patron. Rather, the personal connection is sustained by political brokers or some other type of middlemen “who provide targeted benefits and solve problems for their followers” (Reference SzwarcbergStokes et al. 2013, 75). These intermediaries make sure to emphasize the personal nature of the connection. They routinely and publicly emphasize their “service to the people” and stress “their particular efforts to obtain the goods … thus creating the appearance that were they not there, the benefits would not be delivered” (Reference Auyero, Lapegna and PomaAuyero, Lapegna, and Poma 2009, 5). The fact that the goods distributed through clientelism are often private, rather than public, goods makes credit claiming comparatively easier (Reference Desposato and SchafferDesposato 2007; Reference MazzucaMayhew 1974; Reference MuñozMüller 2007). However, credit claiming might be harder in certain cases, such as in the distribution of targeted government programs (Reference WolffWeitz-Shapiro 2014). In those cases, patrons and brokers have to make an additional effort to convince the clients of the personalized and discretionary character of the benefit.Footnote 17 This personal dimension of the exchange implies that clientelistic relations are ongoing, usually sustained over long periods of time.Footnote 18
The clientelistic exchange is discretionary because patrons (or brokers) enjoy considerable discretion about who receives the benefit – or at least they can make clients believe they have such discretion. Most of the contemporary literature chooses to characterize the clientelistic exchange as contingent on clients’ behavior rather than simply discretionary. Indeed, most definitions of clientelism emphasize the quid-pro-quo nature of the exchange in a very specific way: the provision of the goods and services is contingent upon the actions of clients. For instance, for Reference SzwarcbergStokes et al. (2013, 13), “the party offers material benefits only on the condition that the recipient returns the favor with a vote or other form of political support.”Footnote 19 From this perspective, two more conditions are necessary before an exchange can be classified as clientelistic. First, the patron can know, infer, or, at the very least, be able to make clients believe that it is possible to monitor political behavior. In the case of electoral support, this might involve some mechanism that violates the secret ballot or makes clients believe that such a violation is a real possibility. Second, clients should believe that they could be punished if they renege on their side of the agreement.Footnote 20 For a clientelistic exchange to take place, the patron or the broker should be able to identify non-compliers and credibly commit to punish them.Footnote 21
My understanding of clientelism is different. I do not assume that clientelistic exchanges require monitoring of specific voting decisions or political behavior, or that clients’ fear of punishment is their main reason for fulfilling their side of the agreement. Whether monitoring and punishment (or the threat thereof) actually occur are both questions subject to empirical research and not an intrinsic part of the definition of the concept. In a clientelistic agreement both sides expect to obtain something from the other, but whether and why the agreement is in fact respected and patrons and clients actually obtain what they want are empirical questions.
To be sure, I am not arguing that monitoring of political behavior and fear of punishment are never present in clientelistic exchanges. In my definition of clientelism, the term “discretionary” includes such cases. These are not, however, necessary characteristics of these types of arrangements and therefore should not be considered intrinsic parts of the concept’s definition.Footnote 22 As I argue, there is another explanation for why clients comply with their side of the agreement. Without being pressured to do so, clients can make an active and conscious decision to support their patron because they think it is in their best interest to do so. If clients believe that their political support is important to maintaining the status of the patron or broker from whom the discretionary benefits flow, clients have strong incentives to provide such support.
Patronage is a subtype of clientelism in which the good received in exchange for political support is a public sector job.Footnote 23 I use the term “patronage contract” informally to denote that patrons and clients engage in contract-like exchange relationships in which politicians provide public sector jobs in exchange for political support. Patronage contracts are implicit or explicit agreements between those who get (or expect to get) a patronage job – the client or employee – and those who get (or expect to get) political support in return – the patron. The term refers to the provision of political support both in the expectation of obtaining a job and in the aftermath of obtaining it; this is, of course, a type of clientelistic exchange only available to those in government. Furthermore, because “a steady and secure income” is a very valuable reward, patronage jobs are exchanged “not for a single vote but for broader electoral support” (Reference Strazza and LongoStokes 2009, 15). In fact, Argentinean patronage employees, for instance, engage in political services that go far beyond the mere act of voting, such as helping with campaigns, attending rallies, monitoring elections, and granting favors to voters (see Chapters 4 and 5).
This definition of patronage does not imply anything about the characteristics of the client (the employee or job recipient). The general understanding in the literature is that patronage jobs are distributed to supporters – which is consistent with the findings of this book – but it is not necessary to be a supporter to be involved in a patronage exchange.Footnote 24 Compared to people outside the partisan network, supporters are good candidates for patronage positions because they are more likely to be willing to provide political services (see Chapter 2). But an appointee who gets a patronage job through purely personal (nonpartisan) connections is also likely to be willing to provide such services (Reference KramonKopecký, Scherlis, and Spirova 2008; Reference Kopecký, Sahling, Panizza, Scherlis, Schuster and SpirovaKopecký, Mair, and Spirova 2012; Reference Kopecký, Scherlis and SpirovaKopecký et al. 2016; Reference SchiumeriniScherlis 2010, Reference Schneider2013). In other words, the only characteristic of clients that matters for this definition is that, in exchange for a job, they implicitly or explicitly agree to provide political services for the patron.
It is also important to note that the types of political services that employees provide for their patrons are very diverse.Footnote 25 The services studied in this book are some of the most common among mid- and low-level positions in the bureaucracy, and as such, are quite different from those provided by political appointees in high-level positions, which generally involve control over public policy. Politicians usually wish to appoint people whom they trust to areas they consider important to their administration, ensuring that the bureaucrat will both implement the policies preferred by the politician and make the implementation easier. Moreover, high-level patronage appointments make it easier for parties and politicians to get public employees to do them favors, make exceptions, or look the other way. Such appointments may also facilitate political corruption (related to the financing of politics), clientelism (especially in the form of manipulation of targeted public programs), and pork-barreling.
1.3 Why Study Patronage?
Why focus on public employment? Why focus on the exchange of public sector jobs for political support as opposed to some other form of clientelism? Over the past two decades, political clientelism in developing countries has become one of the most-studied topics in political science.Footnote 26 Most research on clientelism, however, has focused on the exchange of goods for votes (vote buying) rather than on understanding how the exchange of jobs for political support (patronage) works. As a consequence, we still know very little about the specific mechanisms that explain the relationship between patronage contracts and political competition. We know even less about what sustains these contracts.
Patronage jobs are generally assumed to be distributed to an incumbent’s supporters in exchange for political support. Thus, conventional wisdom posits that patronage significantly increases a party’s chances of winning elections and staying in power. Yet we have surprisingly little systematic evidence about the type of political support or political services that patronage employees provide in exchange for their jobs. We lack any precise assessment of the kind of support that employees provide, which employees provide this support, or the extent of this practice in public administration. Moreover, we still do not have a sound explanation for why public employees provide such political support.
The limited attention that the comparative politics literature has given in recent years to patronage politics is particularly surprising considering that patronage is widespread and has been found in Western and Eastern Europe (Reference ChubbChubb 1981, Reference Chubb1982; Reference GoldenGolden 2003; Reference Grzymala-BusseGrzymala-Busse 2003; Reference Kopecký, Sahling, Panizza, Scherlis, Schuster and SpirovaKopecký, Mair, and Spirova 2012; Reference Pepinsky, Pierskalla and SacksPasotti 2010), the United States (Reference Banfield and WilsonBanfield and Wilson 1963; Reference Jones and HwangJohnston 1979; Reference WolfingerWilson 1961; Reference XuWolfinger 1972), Asia (Reference Callen, Gulzar, Hasanain and KhanCallen et al. 2016; Reference JamesIyer and Mani 2012; Reference PollockPierskalla and Sacks 2020), Africa (Brierley 2021; Reference DriscollDriscoll 2018; Reference Kopecký and MairKopecký 2011), and Latin America (Reference Brollo, Forquesato and GozziBrollo, Forquesato, and Gozzi 2017; Reference Colonnelli, Prem and TesoColonnelli, Prem, and Teso 2018; Reference Grzymala-BusseGrindle 2012). Not only is patronage a widespread phenomenon about which we know very little, but it is one that has significant economic and political consequences.
It is easy to see how the politically motivated distribution of public sector jobs could affect the quality of public administration and generate economic inefficiencies. Given that the criterion for selecting new employees for patronage contracts is their willingness or capacity to deliver political services rather than their skills, education, or ability to do the job, there is no formal mechanism to prevent unqualified individuals from getting hired, leading to poor public administration. The effect of hiring potentially unqualified workers, however, is limited by the fact that many jobs in the public sector do not require very sophisticated skills. Indeed, in today’s polities, “most clientelistic intercessions operate above the fulfillment of minimal capacity requirements for entry into the administration” (Reference RosenzweigRoniger 2004, 366). Moreover, political appointments can be restricted to low-ranking positions, while top positions that are more consequential to the functioning of the administration are filled using meritocratic criteria (Brierley 2021).
However, when partisan bias – even if limited – in hiring does affect the quality of the administration, economic resources end up being wasted. If the employees hired are unqualified for the job, public sector productivity drops, perhaps as a product of lack of proper qualifications or, if political appointees devote some of their working hours to the provision of political services, as a product of devoting less time to ordinary on-the-job tasks. Patronage may also reduce the incentives for discretionarily hired employees to perform on the job (Reference ZarazagaXu 2018). Additionally, high rates of administrative turnover, which often occur when hiring and firing decisions are politicized, can result in less-experienced public servants who have shorter time horizons and are often unwilling to engage in projects left over from previous administrations (Reference CornellCornell 2014). Bureaucratic turnover can also disrupt the process of public service delivery, reducing the quality of services (Reference Akhtari, Moreira and TruccoAkhtari, Moreira, and Trucco 2017; Reference Valdebenito Pedrero, Yáñez and RojasToral 2020). Additionally, if the purpose of hiring is to obtain political support, the size of the public administration may expand beyond its objective needs to instead reflect the political needs of the patron (Reference Kemahlıoğlu and BayerKemahlioğlu 2012). Spending on patronage, in turn, leads to the under-provision of public goods. When public money is used for political gain, less is available for public goods (Reference Lizzeri and PersicoLizzeri and Persico 2001; Reference Mares, Muntean and PetrovaMagaloni, Díaz-Cayeros, and Estévez 2007; Reference Piattoni and PiattoniPersson and Tabellini 1999).Footnote 27
At the same time, following Reference Weitz-ShapiroWeber’s (1978) foundational ideas on how bureaucratic structures shape bureaucratic behavior, empirical studies have found a number of positive outcomes associated with professional bureaucracies in which hiring, promotions, and dismissals are insulated from electoral politics. Indeed, Weberian bureaucracies have been associated with economic growth (Reference Evans and RauchEvans and Rauch 1999), poverty reduction (Reference HerreraHenderson et al. 2007), lower corruption (Reference Bersch, Praça and TaylorBersch, Praça, and Taylor 2017; Reference Charron, Dahlström, Fazekas and LapuenteCharron et al. 2017; Reference Dahlström, Lapuente and TeorellDahlström, Lapuente, and Teorell 2012; Reference Meyer-Sahling, Schuster and MikkelsenMeyer-Sahling and Mikkelsen 2016; Reference OlsonOliveros and Schuster 2018; Reference Robinson and VerdierRauch and Evans 2000), less disruptive protests (Reference Cornell and GrimesCornell and Grimes 2015), more investment in infrastructure (Reference RemmerRauch 1995), improvements in health outcomes (Reference Cingolani, Thomsson and DenisCingolani, Thomsson, and de Crombrugghe 2015), and more generous welfare systems (Reference RuedaRothstein, Samanni, and Teorell 2012).Footnote 28
Despite all the negative outcomes linked to patronage politics, some scholars have argued that completely insulating the bureaucracy from politicians is not necessarily desirable. For instance, Reference MuñozMüller (2007, 258) argues that, far from generating inefficiencies, patronage can be used to increase policy-making capacity, writing: “(b)y planting their trustees in the administration and the public sector more generally, political parties can make their policies better informed and smooth their implementation.” From the principal–agent perspective dominant in the study of bureaucracies in the developed world (Reference 239Persson and TabelliniPepinsky, Pierskalla, and Sacks 2017), one of the main problems in the relationship between politicians (principals) and bureaucrats (agents) is how to make sure that politicians can delegate responsibility to bureaucrats and still obtain their preferred outcome.Footnote 29 Patronage appointments are a possible solution to this problem; with full freedom to appoint supporters, the likelihood that the interests of the bureaucrat mirror those of the politician increases considerably. In this vein, Reference Grzymala-BusseGrindle (2012, 154) argues that in some Latin American countries “patronage systems encouraged the responsiveness of bureaucratic actors to executive policy leadership.” Although these arguments have traditionally been applied to the study of the behavior of high-level public officials who determine policy implementation and not the middle- and low-level public employees who are the focus of this book, political appointees at any level may be more enthusiastic, responsive, and loyal workers when they share the preferences of politicians. For instance, Reference Toro, Luna and MardonesToral (2019) argues that, under certain conditions, political appointments at mid-levels of an administration (i.e., school directors) can improve public service delivery by ensuring social and political connections between politicians and bureaucrats.Footnote 30
The distribution of patronage jobs also has a number of serious political consequences that can reduce political competition and the quality of democracy. Control of patronage jobs increases an incumbent’s likelihood of re-election, reducing the level of electoral competition (e.g., Reference Calvo and MurilloCalvo and Murillo 2004; Reference Folke, Hirano and SnyderFolke, Hirano, and Snyder 2011).Footnote 31 Beyond the principled reasons – discussed below – for opposing this unfair advantage, a decrease in political competition is also costly because competitive elections are associated with a number of positive outcomes. Empirical studies show that increased political competition improves government performance by making politicians more accountable and responsive to their constituencies. Competitiveness has been associated with greater spending for primary education in Mexico (Reference Helmke and LevitskyHecock 2006) and in Africa (Reference Stokes, Boix and StokesStasavage 2005), higher economic growth in US states (Reference Besley, Persson and SturmBesley, Persson, and Sturm 2010), stronger rule of law in Argentina (Reference ChávezChávez 2003, Reference Chávez2007), better access to local public goods in Tanzania (Reference RuedaRosenzweig 2015) and in Mexico (Reference HollandHiskey 2003), less politicization of the state in Eastern Europe (Reference Grzymala-BusseGrzymala-Busse 2003, Reference Grzymala-Busse2007), and bureaucratic reforms involving meritocratic recruitment in Latin America (Reference GeddesGeddes 1994) and the United States (Reference ToralTing et al. 2013).
Crucially, in countries with “widespread institutional weakness” (Reference Levitsky and MurilloLevitsky and Murillo 2005, Reference Navarro and Marcelo2009), such as Argentina and many other young democracies, political competition is an even more important mechanism of accountability and responsiveness. Politicians in democracies characterized by weak institutions are subject to low levels of horizontal accountability and enjoy ample opportunities to break and change the rules, but they still need to win elections to stay in power. In fact, institutional strength is the fundamental difference between young democracies with low levels of competition, and what Reference Persson and TabelliniPempel (1990) has called “uncommon democracies,” advanced democracies dominated by one party. According to Reference Persson and TabelliniPempel (1990, 7), one undesirable consequence of single-party dominance is that “the longer a party is in power, the greater the opportunity it has to use state resources to shape and reshape its following.” Such opportunities are considerably more numerous when institutions are weak, so low levels of electoral competition have more serious consequences in such contexts.Footnote 32
Moreover, the use of any state resource for electoral competition provides an unfair advantage to the incumbent party. In this sense, patronage has much in common with the targeted manipulation of public programs and pork-barrel spending. In all these cases the electoral playing field is skewed in favor of the incumbent party. Elections may still be competitive, but they are less fair for the opposition because of the political use of state resources by the incumbent party. Furthermore, the use of state resources to finance political workers generates perverse incentives among politicians. If their success at the polls strongly depends on the political services provided by public employees, politicians have “little reason to care about the formulation of policies, the construction of programmatic parties, and practices of accountability” (Reference ScherlisSchaffer 2007, 11).
As a consequence of this unfair incumbency advantage, democratic accountability also suffers. When competition is fair, elections provide information about voters’ preferred policies, which makes it easier for these preferences to be represented and translated into public policy. If incumbents have an unfair advantage in electoral competition, elections tend to provide less information about the distribution of voters’ preferences. An uneven playing field can also affect levels of satisfaction with the functioning of democracy and decrease the legitimacy of elected leaders. In a study of party competition in Latin America, Reference Kitschelt, Wilkinson, Kitschelt and WilkinsonKitschelt, Luna, and Zechmeister (2010) show that citizens express the least confidence in democratic practices in countries where programmatic party competition is weak. Similarly, empirical studies have shown that higher levels of (perceived) corruption (e.g., Reference Anderson and TverdovaAnderson and Tverdova 2003; Reference Rothstein, Samanni and TeorellRose-Ackerman 1999; Reference ShefterSeligson 2002; Reference Weitz-ShapiroWeitz-Shapiro 2008) and clientelism (Reference Kitschelt, Altamirano, Carlin, Singer and ZechmeisterKitschelt 2007) reduce confidence and trust in democratic institutions.
It is also important to understand how patronage works because it directly affects opportunities for other forms of clientelism as well as pork-barrel spending and corruption. Bureaucrats have control over a wealth of resources that can be used and abused for political and personal gain. When politicians appoint “friends” to certain positions, they will often find it easier to get such patronage employees to do them favors, make exceptions, and provide the resources required for clientelistic exchanges and political corruption. Without patronage – without the possibility of discretionarily appointing “loyal” public employees – misuse of state resources for partisan or personal gain becomes more difficult.Footnote 33 Put differently, public sector employment is “both a source of rents and a mechanism of rent distribution” (Reference Hafner-Burton, Hyde and JablonskiGrzymala-Busse 2008, 659).
Public sector jobs are also often the main resource used to pay for the work of brokers and activists. Although this is hardly ever the main focus of analysis, most studies of clientelism refer to public jobs as one of the principal ways of financing political work (Reference Mares and YoungMares and Young 2016). The machine politics literature in the United States has long explicitly recognized and studied the importance of “jobs for the boys” to the sustenance of political machines. In the words of Reference Kiewiet de Jonge and NickersonKey (1956, 382): “The operation of a party organization requires the services of many men and women … Indirectly, a considerable part of party expense is met by the public treasury, and the chief means of channeling public funds to party support is through the appointment of party workers to public office.” Although there are few studies of machine politics outside the United States that focus specifically on public employment, there is considerable agreement that public sector jobs are an important component of electoral machines.Footnote 34
In addition to the economic inefficiencies associated with patronage, its erosion of political competition and accountability, and its likely enabling of other forms of clientelism and corruption, patronage has yet another serious effect on the quality of democracy. The biased distribution of public sector jobs raises serious questions about the independence of the public administration and the possibility of equal access to the state. Without clear rules for hiring and promotion, public employees owe their jobs to the patron who makes hiring decisions. The bureaucracy thus lacks independence, leading to a possible lack of impartiality vis-à-vis citizens. In patronage democracies, “proximity to a state official increases a voter’s chances of obtaining valued state resources and services” (Reference Chandra, Kitschelt and WilkinsonChandra 2007, 87). What happens to the many citizens who lack connections to a state official?
At the same time, as Reference ChubbChubb (1981, 120) points out, “an able administrator can … transform a routine bureaucratic procedure into a personal favor.” In settings with widespread patronage, public positions that channel those favors are filled with partisan workers who provide political support to their patrons. Would a citizen who is clearly identified with the opposition get equal treatment? Would a citizen with no personal connections to anyone on the public sector payroll be discouraged from even asking for help? When the public administration is filled with patronage employees, ordinary citizens may feel excluded. Whether this impression of uneven access to the state is accurate or not, it could be enough to keep citizens without connections away.Footnote 35 When those without personal connections are the poor – as is often the case – patronage is linked not only to uneven access to the state and therefore weak rule of law, but also to the reproduction of existing inequalities.
1.4 Challenges to the Study of Patronage
Although patronage is often perfectly legal, it is particularly difficult to study because it constitutes a “gray area” of acceptable practice (Reference Vommaro and CombesVan de Walle 2007, 52). The challenges are sufficiently serious that, according to Reference GeddesGeddes (1994, 105): “In the real world, there is no way to measure amounts of patronage or how much influence on the vote it has.” Long before Geddes, Reference WolfingerWilson (1961, 372) wrote: “So much secrecy is maintained in city politics that no exact data on patronage may ever be obtained in cities of any size.” Since then, contemporary researchers have used different strategies in an attempt to generate more systematic data for the study of patronage politics.
Taking advantage of the fact that civil service rules are weak or nonexistent in some contexts, such as Argentina and many young democracies (e.g., Reference Grzymala-BusseGrindle 2012; Reference IacovielloIacoviello 2006), a number of studies have tried to get around the problem of measuring patronage by using proxies such as the number of public sector employees (Reference Calvo and MurilloCalvo and Murillo 2004; Reference Grzymala-BusseGrzymala-Busse 2003; Reference OliverosO’Dwyer 2004), the number of temporary employees (Reference Brollo and TroianoBrollo and Troiano 2016; Reference FerraroFerraro 2006; Reference Kemahlıoğlu and BayerKemahlioğlu 2012), the number of low-level employees (Reference DriscollDriscoll 2018), or spending on personnel (Reference GottliebGordin 2002; Reference NichterNazareno, Stokes, and Brusco 2006; Reference Rocca RivarolaRemmer 2007; Reference SchusterSchiumerini 2018). However, cross-national analysis finds similar levels of meritocracy in actual hiring procedures among countries with and without legal merit requirements (Reference ScottSchuster 2017). While strict and clear civil service rules might be a good indicator of low levels of patronage, the absence of such rules does not necessarily mean that public sector jobs are distributed through patronage contracts. Neither a large number of public sector jobs nor high spending on personnel is a necessary or sufficient condition for patronage; those figures could easily hide (or not) patronage appointments. Both of those measures, for instance, would remain stable in a context in which new incumbents fire employees hired by the previous administration and replace them with their own supporters without increasing the number of employees or spending on salaries (Reference KramonKopecký, Scherlis, and Spirova 2008; Reference Meyer-Sahling and MikkelsenMeyer-Sahling 2006).Footnote 36
A few recent studies in developing countries have managed to obtain unique public employee-level administrative data. In Ghana, for instance, Brierley (2021) combines subnational bureaucrat-level data on time of hiring with bureaucrats’ ethnicity and home regions (to proxy for partisanship) to study variation in patronage hiring across administrative ranks. In Brazil, scholars have combined an unprecedented dataset on the universe of the Brazilian public sector with a dataset on party membership (Reference Brollo, Forquesato and GozziBrollo, Forquesato, and Gozzi 2017) and, in another study, with candidates and campaign donors (Reference Colonnelli, Prem and TesoColonnelli, Prem, and Teso 2018) to show partisan bias in hiring decisions.Footnote 37 In cases where it even exists, this type of fine-grained data is very hard to obtain in most developing countries (Brierley 2021), and Argentina, where data like this is simply not available, is no exception.Footnote 38
The challenge, moreover, is not only to be able to assess whether public sector jobs are distributed with partisan bias or which jobs are distributed this way but to study the type of political support that patronage employees provide to their bosses in exchange for those jobs as well as the reasons behind this support. Because both politicians and public sector employees have incentives to misreport clientelistic exchanges, asking either of them directly about these political services is also problematic.Footnote 39 Employees could be reluctant to admit that they hold a patronage job because that could imply that they did not really “earn” it. From the point of view of the patron, admitting the existence of patronage jobs is even more problematic because it implies the use of state resources for political purposes. Even when patronage is often not illegal, politicians may view inquiries about it “as valence questions on which they suspect most citizens and observers to be on one side of the issue (against)” (Reference Kitschelt, Wilkinson, Kitschelt and WilkinsonKitschelt and Wilkinson 2007a, 326).Footnote 40
To measure the types and extent of the political services that employees hired through patronage contracts provide to their patrons,Footnote 41 I avoid the use of proxies and mitigate the problems associated with asking sensitive questions directly to the actors involved. Here, I test the empirical implications of my theory of self-enforcing patronage with a face-to-face survey of 1,200 local public sector employees (see Chapter 3). To minimize social response bias, I use a specific survey methodology, the list experiment, to guarantee respondents full anonymity. Another research tool, the survey experiment, allows me to assess why public sector employees comply with their side of the patronage contract. I complement the analysis of the survey results with in-depth interviews of public sector employees, political brokers, and politicians.
While the theory developed and tested in this book applies to many countries without strict civil service rules, Argentina offers an excellent setting in which to test it. Once infamous for its political instability,Footnote 42 Argentina is today infamous for its clientelistic politics. Fairly or unfairly, clientelism has received more scholarly attention in Argentina than in any other single country.Footnote 43 Besides its well-documented and widespread clientelism, Argentina provides an especially good setting for studying the mechanisms of patronage for several reasons. The country lacks stable civil service rules and has a large public sector with “well-developed patronage systems” (Reference Calvo and UjhelyiCalvo and Ujhelyi 2012). It has an extensive level of decentralization that results in significant variation in the size and characteristics of public employment across provinces and municipalities.Footnote 44 In addition, although some provincial regulations apply to municipal public employment, control over local personnel is the exclusive responsibility of local governments, which increases the variation in the distribution of patronage even more.
The Argentine Constitution has guaranteed job stability for public employees since 1949.Footnote 45 National, provincial, and local public sector employees often gain job security after a year and cannot be fired once they have attained permanent positions. Job security rules, however, are systematically bypassed through the increasing use of temporary contracts. According to estimates by Reference Pradhanawati, Ikbal Tawakkal and GarnerPomares, Gasparin, and Deleersnyder (2013), from 2003 to 2012 the proportion of federal public employees on temporary contracts increased from 15 percent to 30 percent. At the local level, temporary contracts are even more common (e.g., Reference SchiumeriniScherlis 2010). Furthermore, even though there have been attempts to create a meritocratic system of recruitment, most appointments at all levels of the administration are still discretionary (e.g., Reference Ferraro and MasseyFerraro 2011; Reference Grzymala-BusseGrindle 2012; Reference Iacoviello, Zuvanic and EchebarríaIacoviello and Zuvanic 2006a). The government can appoint supporters to public jobs either by enlarging the public sector or by replacing temporary employees. As long as permanent personnel are not touched, every new president, governor, and mayor enjoys considerable discretion to appoint new personnel. This discretion is used (and abused) extensively to distribute patronage contracts, especially at the local level.Footnote 46
Despite all the attention given it in the literature, Argentina is not an outlier among Latin American countries for the specific type of clientelistic exchange that is the focus of this book. An Inter-American Development Bank (IDB) evaluation of bureaucracy and civil service systems across Latin America classifies Argentina as a country with intermediate civil service development, a designation also given to Colombia, Mexico, and Uruguay (Reference EchebarríaEchebarría 2006; see also Reference Zuvanic, Iacoviello, Gusta, Scartascini, Stein and TommasiZuvanic, Iacoviello, and Gusta 2010). Even though all Latin American countries have established professional civil services, the extent to which merit is the only criterion used for hiring, firing, and promotion varies enormously across countries. Across Latin America, civil service systems exist parallel to political criteria (Reference Grzymala-BusseGrindle 2012; Reference IacovielloIacoviello 2006; Reference Zuvanic, Iacoviello, Gusta, Scartascini, Stein and TommasiZuvanic, Iacoviello, and Gusta 2010).Footnote 47
All in all, the combination of well-documented and widespread patronage and high levels of decentralization – leading to significant variation in the economic and political settings in which patronage contracts arise – makes Argentina a particularly useful laboratory for studying patronage politics. Because the purpose of this book is to study how patronage works and not to establish whether patronage exists or how it varies, the choice of a country well known for its patronage politics seems appropriate. At the same time, Argentina’s non-outlier status in terms of patronage appointments, at least by Latin American standards, makes the findings of this book relevant for other Latin American cases and other young democracies. Finally, Argentina’s substantial subnational variation allows me to test the theory of self-enforcing patronage in very different settings without losing the advantages of conducting research in a single country.Footnote 48 Subnational variation also makes the portability of the results to other settings more plausible.
1.6 The Argument in Brief
Why does the control of patronage increase an incumbent’s chances of staying in power? What are the specific mechanisms behind the electoral returns to patronage politics? What do patronage employees do? What motivates them to do it? In line with other scholars, I argue that public sector jobs are disproportionally distributed to political supporters, i.e., those who are more closely connected to party networks. Since such government jobs are extremely valuable, the type of political support that is expected in exchange for public sector employment goes far beyond the simple act of voting. In fact, politicians distribute many low- and mid-level positions in the bureaucracy with the intent of maintaining a network of political workers on the ground to perform various political activities – such as helping with electoral campaigns or organizing and attending political rallies – that are essential for attracting and maintaining electoral support. Patronage jobs provide politicians in power with a “free” supply of political workers.
Patronage contracts, however, are difficult to enforce and both politicians and employees face a commitment problem. Politicians risk “wasting” jobs on employees who, once hired, may not comply with their side of the contract. For their part, potential employees (clients) promised a future job risk providing services for candidates who, once in office, might renege on their side of the agreement, withholding the job offer. This book focuses on the first type of commitment problem.
Why would public employees uphold the agreement and provide political services even after receiving the benefit of the job? How can the patron make sure not to “waste” jobs on citizens who will not fulfill their side of the contract?Footnote 49 Existing explanations are based on either feelings of reciprocity or fear of punishment. According to the first set of theories, clients (patronage employees) comply with their side of the agreement and provide the promised political services because they want to help the patron who helped them (e.g., Reference Finan and SchechterFinan and Schechter 2012; Reference LazarLawson and Greene 2014). According to the second set of theories, clients comply because they are afraid that the patron will cut off the benefit (fire them, in this case) if they fail to do so (e.g., Reference Brusco, Nazareno and StokesBrusco, Nazareno, and Stokes 2004; Reference Ronconi and ZarazagaRobinson and Verdier 2013; Reference StokesStokes 2005, Reference Stokes, Dunning, Nazareno and Brusco2007).Footnote 50 Departing from these accounts, I argue that there is a third explanation for why public employees uphold the patronage contract. Patronage employees (clients) engage in political activities that support politicians (patrons) because their fates are tied to the political fate of their patrons.
Politicians use personal and partisan connections to screen potential patronage employees and separate perceived supporters from non-supporters. Perceived supporters may support the politician for ideological or personal reasons or because of past or expected benefits; they may have real connections with the party, or they may just wish to be perceived as supporters in order to obtain benefits. Indeed, perceived support is not about ideological affinity but rather about proximity to the party’s network (Reference Calvo and MurilloCalvo and Murillo 2019). All things being equal, perceived supporters are more likely to comply with the agreement and provide political services because they are more likely than non-supporters to truly support the party or politician. This makes the provision of political services less costly for them.
Perceptions, however, can be misleading. In fact, once voters expect partisan bias in hiring decisions, they have an incentive – the prospect of a job – to move closer to the partisan networks and to pretend to have certain political preferences. Once hired, patronage employees may also change their minds about their political preferences or simply reduce the effort they are willing to devote to political services. Patronage employees still have the option of going back on their agreement after getting the job. Hiring supporters – or, more accurately, perceived supporters – is not in itself enough to solve the commitment problem.
Why, then, are patronage agreements respected? Patronage contracts are respected because patronage employees believe that their jobs and working conditions are tied to the political success of their patron. Public employees under patronage contracts believe that if the incumbent loses the election, their own jobs could be in jeopardy. Knowing that they are branded as supporters of the incumbent, patronage employees worry about keeping their jobs and/or working conditions if the opposition were to win. Just as the ousted incumbent did previously, a newly elected official will want to distribute patronage jobs to new employees who will be likely to provide political services for him or her. Because only his or her supporters can credibly commit to do that in the future, old employees who are known or perceived to be supporters of the incumbent will be replaced or demoted. Supporters of the incumbent thus have an enormous incentive to engage in political services aimed at keeping the incumbent in power. Patronage contracts are thus incentive compatible, and the commitment problem associated with the sequenced nature of the job-for-services exchange disappears.Footnote 51
By providing a direct assessment of the specific mechanisms that explain the effect of patronage on political competition, this book makes two main contributions. First, it is the first systematic attempt to measure the types and extent of the political services that employees hired through patronage contracts provide for their patrons. Importantly, as highlighted in this book, political brokers are not the only ones involved in these activities. Most of the literature has focused on brokers as central actors in machine politics, but many public employees who are not political brokers are nevertheless involved in the provision of political services. We know very little about the activities and motivations of these less studied and less influential but still critical political actors.
Second, I provide a novel explanation of the enforcement of these patronage contracts that has implications for understanding the functioning of clientelistic exchanges more broadly. By treating both patrons and clients – in this case, patronage employees – as equally sophisticated and self-interested individuals, my argument departs from existing accounts that emphasize either feelings of reciprocity or fear of punishment. Public sector employees comply with the patronage contract because their interests are aligned with those of their patrons. Clientelistic contracts need not imply passive clients behaving against their interests out of fear of being punished, nor need they imply particularly noble clients who always reciprocate favors. Clients can make sophisticated calculations when deciding whether to comply with the agreement or renege. At the same time, if patronage contracts, and clientelistic arrangements more broadly, are self-sustaining without punishment, clientelism becomes cheaper, harder to detect, and even harder to curb. If costly political machines are not needed to identify and punish non-compliers, clientelism becomes cheaper; if illegal threats, coercion, and monitoring devices are not necessary to convince clients to provide political support, clientelism becomes a lot more difficult to uncover and eliminate.Footnote 52
1.7 Plan of the Book
The rest of the book is organized as follows. Chapter 2 introduces my theory of self-enforcing patronage and its empirical implications. Following a discussion of the relationship between patronage jobs and political competition and a description of the types of political services that different types of public employees provide for their bosses, I lay out a new theory that seeks to understand the reasons behind the provision of these political services. Chapters 3 through 6 offer the main empirical tests of the theory. After describing the empirical strategy, the case selection, the interviews, and the survey, Chapter 3 shows that patronage jobs are disproportionally distributed to those more closely connected to the partisan network (supporters). Chapter 4 uses list experiments embedded in an original survey of 1,200 public employees to measure the types and extent of the political services that patronage employees provide for their patrons during elections – namely, attending rallies, helping with electoral campaigns, and monitoring elections. Chapter 5 focuses on one political service that employees perform in between elections: providing favors. Both Chapters 4 and 5 show that, as predicted by the theory, supporters are more involved than non-supporters in the provision of political services. Chapter 6 tests the theory’s central implication, that public sector employees fulfill their side of the patronage contract by providing political services in pursuit of their own personal interests. Using survey experiments, this chapter shows that patronage employees believe that their fates are tied to the political fate of their patron, providing a strong incentive for them to help the patron remain in power. Chapter 7 presents evidence from other Latin American countries as an out-of-sample test of the theory in order to build more confidence about the external validity of the argument. Beginning with a general discussion of the characteristics of civil service systems in Latin America, I then discuss the cases of Argentina, Bolivia, and Chile. The final chapter reviews the major findings and discusses their implications for our understanding of the dynamics of clientelism and the possibilities for curbing it. The research findings draw particular attention to the interests and strategic behavior of clients, to what these interests and behavior mean for the eradication of patronage and clientelism, and to the particularly damaging effects of patronage for the quality of democracy and equal access to the state.