Milícias are mafia-style organizations, often composed of current and former state agents, that have rapidly expanded in areas with limited state presence and weak legal oversight. In Rio de Janeiro, their territorial control is not maintained by coercion alone, but also through strategic political alliances. This paper theorizes and tests a mechanism linking milícia expansion to electoral politics: milícias deliver concentrated electoral support to specific politicians, who in return shield their operations by influencing bureaucratic appointments and law enforcement priorities. Using original geospatial and electoral data, we show that milícia entry into a new area increases electoral concentration and disproportionately benefits milícia-aligned candidates in adjacent territories. We further demonstrate that this electoral capital is converted into political power through key bureaucratic appointments that facilitate further expansion and institutional impunity. Our findings support a theoretical framework in which elections reinforce, rather than constrain, criminal governance in democratic settings.

Presidential impeachment has traditionally been seen as a clear manifestation of accountability. Recently, however, the impeachment of some presidents has led to public backlash, with some being regarded as "legislative coups." In examining the determinants of impeachment, studies have highlighted the role played by corruption scandals, mass mobilization, and the absence of a legislative shield. We contribute to this literature by highlighting the importance of affective and ideological polarization for presidential impeachment. We theorize impeachment as a multi-hurdle process and focus on examining the determinants of overcoming two primary hurdles: a lower house vote (the first hurdle) and removal from office (the second hurdle). Using 44 successful and failed impeachments, we employ crisp-set qualitative comparative analysis (QCA) to examine this issue. The analysis supports our hypotheses, highlighting that many cases of impeachment cannot be explained unless polarization is taken into account. In particular, our analysis illustrates three types of impeachment: polarized impeachment, scandalized impeachment, and mobilized impeachment. The results withstand a wide range of robustness tests, including sensitivity ranges, consistency thresholds, fit-oriented robustness, and cluster analysis. Given the rise of affective polarization worldwide, presidential impeachment may be increasingly weaponized by polarized elites.

The power-sharing literature lacks a review that synthesizes its findings, despite spanning over 50 years since Arend Lijphart published his seminal 1969 article "Consociational Democracy". This review article contributes to the literature by introducing and analysing an original dataset, the Power Sharing Articles Dataset, which extracts data on 23 variables from 373 academic articles published between 1969 and 2018. The power-sharing literature, our analysis shows, has witnessed a boom in publications in the last two decades, more than the average publication rate in the social sciences. This review offers a synthesis of how power sharing is theorized, operationalized, and studied. We demonstrate that power-sharing has generally positive effects, regardless of institutional set-up, post-conflict transitional character, and world region. Furthermore, we highlight structural factors that are mostly associated with the success of power sharing. Finally, the review develops a research agenda to guide future scholarly work on power-sharing.

State presence is often considered weakest in a country's periphery. In contrast, we argue that small towns on the periphery have more influence on the distribution of centrally-allocated public works than is commonly understood. Infrastructure projects yield greater political returns in smaller municipalities because they are more visible, less costly, and credit-attribution is simpler. Combining extensive fieldwork with a novel dataset of over 60,000 project-specific amendments added to the Brazilian budget by national deputies (2015–2023), we show that smaller municipalities receive more projects and funds per capita than larger ones. Deputies also gain greater vote returns per Real spent in smaller towns. Yet this higher spending often takes the form of simpler, lower-quality projects, allowing critical infrastructure gaps in water, health, education, and transportation to persist. Thus, while more infrastructure flows to the periphery, it is of lower quality, reinforcing the paradox of visible state presence but limited state capacity.

A fundamental component of citizenship includes seeking protection from crime and violence. How do citizens make such claims to law enforcement when the police themselves are often predatory? The core argument is that police predation (1) discourages citizens from reporting crimes directly to the police while (2) shifting reporting towards intermediated channels. We argue that trust is the mechanism driving this reporting behavior. To test our theory, we focus on the reporting of violence against women (VAW) in Rio de Janeiro, Brazil. We build an original database of reports to the police and to an intermediary (an independent hotline), then conduct descriptive and causal tests to show that exposure to police violence redirects reporting from the police to the intermediated hotline. Together, these findings bring together the claim-making and policing literatures to demonstrate that police violence incentivizes citizens to report VAW to alternative channels and, in their absence, to stay silent.

This study examines how a single variation in electoral rules influences politicians' investment in public and targeted goods. Using a unique threshold policy introduced in Brazil in 1988, I employ a regression discontinuity design to estimate the effect of a two-round election on the distribution and quality of targeted and public goods. The results show that single-round elections are associated with greater efforts to allocate targeted goods relative to single-round elections, but I only find marginal increases in the quality of public goods delivery by mayors. Counter to previous studies, it is unclear whether two-round elections are associated with greater effort to deliver public goods. The findings suggest that previous work has focused on too narrow a set of indicators to assess the effect of this particular type of electoral rule, overlooking its impact on targeted goods provision.

Does sustained state presence erode criminal governance more effectively than violent crackdowns? We argue that permanent police deployment undermines both criminal authority and the clandestine service provision that sustains it by displacing criminal actors and expanding state capacity. We test this argument by leveraging the staggered installation of Police Pacifying Units (UPPs) in favelas in Rio de Janeiro, Brazil. We estimate a difference-in-differences design using novel data, including more than 1.45 million complaints from the anonymous tipline Disque Denúncia. We find that UPP installation significantly reduced criminal governance and clandestine provision of transportation, gas, cable TV, and internet services. In contrast, pre-UPP violent crackdowns produced no such effects, suggesting that punitive operations alone are insufficient to displace entrenched criminal governance.

Research on public goods provision remains strikingly state-centric, even though millions of urban residents across diverse world regions access goods and services by turning to a range of non-state actors. This paper focuses on one such actor: armed criminal organizations that exercise governing authority. Meanwhile, existing scholarship on criminal governance emphasizes cases of direct service provision by criminal actors that make particularly visible state absence or corruption. Yet, criminal actors’ influence on public goods access can take multiple forms beyond direct provision. Thus missing from both literatures is a broader framework for conceptualizing variation in how criminal organizations structure the modes through which citizens access public goods in settings of criminal rule. We develop a typology of public goods provision under criminal governance – what we call criminalized public goods provision – along two dimensions: the type of good that is being provided (“small-” vs. “big-ticket” goods) and the structure of territorial control (from contested to monopolistic criminal governance). Crossing these dimensions yields four ideal-type modes: consistent substitution, consistent intermediation, episodic substitution, and episodic intermediation. The type of good sets the criminal group’s role – whether it stands in for the state or structures the state’s access – while the competitive structure of criminal authority sets how reliably it can play that role. Rather than replacing the state, criminal governance reorganizes the institutional channels through which critical goods flow to citizens. We ground the typology in extensive fieldwork in São Paulo and Mexico City, drawing on qualitative interviews and focus group discussions. By linking criminal governance to variation in service delivery, this paper bridges the literature on public goods provision and non-state armed governance.

This project assesses the impact of specialized judicial institutions on women's claim-making behavior across two major metropolitan areas in Latin America: the states of Mexico and São Paulo, both marked by strong criminal group presence and uneven state provision. Using a survey experiment informed by extensive fieldwork, the research examines how women in these settings choose to report gendered crimes, whether to formal authorities (e.g., police), criminal organizations, or non-governmental organizations, in municipalities with varying levels of state presence and judicial specialization.