Guillermo Toral


"The benefits of patronage: How the political appointment of bureaucrats can enhance their accountability and effectiveness"


The political appointment of bureaucrats, a ubiquitous phenomenon around the world, is typically seen as a rent-seeking strategy whereby politicians sustain clientelistic networks and manipulate public administration to their advantage. I argue that political appointments can also increase bureaucratic accountability and effectiveness in public service delivery because they provide political and social connections between bureaucrats and politicians. These connections provide access to material and immaterial resources, enhance monitoring, facilitate the application of sanctions and rewards, align priorities and incentives, and increase mutual trust. Patronage therefore works as a governance technology. In certain conditions, especially in developing contexts where politicians value the delivery of public services but cannot access other tools to motivate bureaucrats to perform, the benefits of political appointments may outweigh the costs. I test this theory with data on municipal governments in Brazil, leveraging two quasi-experiments with administrative data for schools in the whole country (a difference-in-discontinuities and a regression discontinuity); two original surveys including conjoint experiments in one state (a face-to-face survey of 926 street-level managers and an online survey of 755 politicians); and 121 in-depth interviews with bureaucrats, politicians and anti-corruption agents. The findings challenge the traditional view of patronage as universally detrimental for development, and draw attention to how bureaucrats and politicians can leverage political appointments and connections for public service delivery.

2020 Instituto Tecnológico Autónomo de México (ITAM, Mexico City, online)
2020 Harvard Ash Center for Democratic Governance and Innovation (Cambridge, MA)
2019 Rutgers University (Newark, NJ)
2019 Centro de Investigación y Docencia Económicas (CIDE, Mexico City)
2019 Yale-NUS College (Singapore)
2019 Vanderbilt University (Nashville, TN)
2019 New York University Abu Dhabi (NYUAD, online)
2019 Tecnológico de Monterrey (TEC, Monterrey, Mexico)
2019 Northeast Workshop in Empirical Political Science (NEWEPS, New York City)
2019 Conference of the Society for Political Methodology (PolMeth, Cambridge)
2018 American Political Science Association Congress (APSA, Boston)
2018 MIT Workshop on the Politics of Education (Cambridge, MA)

"Turnover: How electoral accountability disrupts the bureaucracy and service delivery"


Electoral accountability is fundamental to representative democracy. Yet, it can also be costly for governance because it causes turnover among bureaucrats (not just elected officials) and disruptions in the delivery of public services. This article advances a theory of political turnover as a process that, starting the moment election results are published, leads to bureaucratic shuffles and depresses service delivery. I demonstrate these turnover dynamics through a close-races regression discontinuity design, using administrative data on public employment and on healthcare service delivery in Brazilian municipalities. The results show that an electoral defeat of the incumbent causes increases in dismissals, hires, and resignations of bureaucrats, and declines in public service delivery in the months following the election. These findings draw attention to the political strategies of lame-duck politicians, and highlight the intense and consequential bureaucratic politics that follow elections.

2020 American Political Science Association Congress (APSA, online)
2018 MIT GOV/LAB Political Behavior of Development Conference (Cambridge, MA)

"Political bureaucratic cycles: How politicians' responses to electoral incentives and anti-corruption policies disrupt the bureaucracy and service delivery around elections"


A vast literature has studied political cycles in economic outcomes and economic policy tools (political business and political budget cycles, respectively). I identify a related phenomenon, which I call political bureaucratic cycles: electoral cycles in the hiring and firing of bureaucrats and in the activities of public employees, which emerge as a result of the combination of electoral incentives and legal rules imposed to limit the use of public employment for electioneering. Empirically, I leverage administrative, identified, contract-level data on the universe of municipal employees in Brazil between 2002 and 2016 to measure political bureaucratic cycles. Hires and dismissals of municipal personnel show markedly cyclical patterns around elections, which are shaped by both incumbents' electoral incentives and their reaction to anti-corruption policies that constrain hiring and firing around elections. Cycles are most pronounced for temporary bureaucrats but are also detectable for civil service bureaucrats, which counters the received wisdom that civil service regimes isolate bureaucrats from political dynamics. Hiring and firing around elections are targeted at less educated people, which is consistent with political bureaucratic cycles partly responding to clientelistic strategies. Consistent with the clientelistic use of public employment, and the legal rigidities imposed on hiring around elections, pre-natal check-ups (a key output of the healthcare bureaucracy) are systematically lower around elections. Findings are grounded on, and complemented with, in-depth interviews with prosecutors, politicians and bureaucrats conducted in 7 states. The paper contributes to bridging the gap between the literatures on political budget/business cycles and on clientelism, two fields that have rarely been linked before.

2019 American Political Science Association Annual Meeting (APSA, Washington DC)
2019 Red para la Economía Política de América Latina Annual Meeting (REPAL, New Orleans)
2018 Latin American Studies Association Congress (LASA, Barcelona)

"Information, oversight, and compliance: A field experiment on horizontal accountability in Brazil"


Pre-Analysis Plan

Does the provision of information about local bureaucracies to the politicians who oversee them decrease irregularities and improve bureaucratic effectiveness? Information interventions are appealing because of their solid microeconomic foundations and their relatively low costs. However, recent experimental studies of information campaigns aimed at fostering vertical ac- countability (between voters and politicians) have found mixed results. Providing information to politicians directly could be more powerful, given politicians’ direct responsibility for allo- cating and managing resources. Information may be particularly effective when provided by auditing institutions, given politicians’ susceptibility to sanctions by these horizontal account- ability actors. I partnered with the audit court of the Brazilian state of Rio Grande do Norte to experimentally study the effects of informing local politicians (both in government and in the opposition) about irregularities and performance in the bureaucracies they oversee. Outcomes are measured using administrative payroll data, a face-to-face survey of bureaucrats, and an online survey of politicians. Preliminary results suggest the treatment reduced the share of workers hired under temporary contracts, increased knowledge about rules among politicians, and changed politicians’ sense of accountability pressure from the state audit court.

2019 Federal University of Rio Grande do Norte (Natal, Brazil)
2019 State Audit Court of Rio Grande do Norte (Natal, Brazil)
2019 Latin American Studies Association Congress (LASA, Boston)
2019 Midwest Political Science Association Annual Meeting (MPSA, Chicago)

"The electoral consequences of EU voting rights: Evidence from Spain" (with Jeremy Ferwerda)

2016 Midwest Political Science Association Conference (MPSA, Chicago)

"Franchise reforms in the age of migration: Why do governments enfranchise noncitizens?"


This article examines the enfranchisement of noncitizens and seeks to explain governments’ decisions about whether to include or exclude them from the vote. By focusing on the incentives behind incumbents’ decisions, the article argues that partisanship, inequality, and immigration are the factors driving the (dis)enfranchisement of noncitizens. The theory leads to a number of testable hypotheses that are then subjected to regression analyses using an original dataset of 33 democracies in the period 1960-2010. The results indicate that while franchise reforms to include noncitizens are more likely to be passed by left-wing governments, noncitizen voting rights are highest at intermediate levels of immigration. The findings are relevant for an emerging literature of quantitative studies of immigrant rights, as well as for the literature on franchise extensions along lines of class and gender.

2015 Midwest Political Science Association Conference (MPSA, Chicago)
2011 Developments in Contemporary Citizenship Conference (King's College London)