Research

My current research focuses on the bureaucratic politics of anti-deforestation enforcement, with a regional emphasis in Latin America. I also have ongoing interests in causal inference, individual-level adaptation to climate change, as well as the broader field of environmental politics. In the past I’ve also worked on topics related to direct democracy, citizenship, and distributive politics.

Working papers

Forbearance forgone: the electoral cost of enforcing the law. Latest version: October, 2025. Draft submitted for review available here.

In a democracy, lawbreakers can punish politicians who support crackdowns. Yet despite findings that officeholders manipulate sanctions for political gain, both theory and evidence supporting that enforcement is electorally costly remain scarce. Consequently, I propose three mechanisms through which backlash emerges—individual pocketbook, network solidarity, and collective grievance—and specify the conditions under which losses are offset by audience rewards. I focus on anti-deforestation enforcement in the Brazilian Amazon, and, through multiple research designs, find that the Workers’ Party—supportive of the crackdown—experienced significant losses in targeted areas, while Jair Bolsonaro—critical of it—reaped electoral rewards. I provide the first individual-level test of enforcement’s effects on political preferences using novel administrative microdata on half a million farmers: while targeted individuals shifted toward rival parties, there was no audience effect among pro-environment individuals. These findings underscore democracy’s tension between the rule of law and the incentives to manipulate enforcement.

Zealots unbound: the bureaucratic microfoundations of the rule of law. Latest version: May, 2026. Job market paper; available upon request.

When politicians try to subvert the rule of law, who resists? I propose an alternative explanation for why law enforcement continues despite top-down pressures: bureaucrats’ commitment to their mission. In contrast to other civil servants, zealots—those with strong beliefs about the contents and procedures that govern policymaking—are more likely to resist attempts to undermine bureaucratic autonomy. I focus on environmental enforcement in Brazil and propose a novel measurement strategy of bureaucratic attitudes leveraging administrative records of over ten thousand employees of two federal agencies, including their partisan, academic, and employment histories. Using quasi-experimental designs and insights from more than fifty interviews, I show that while others decreased enforcement after pressures from the Bolsonaro government, zealots continued to carry out their duty. Their persistence curbed deforestation and influenced non-zealots in their networks. Taken together, these findings illustrate the role of bureaucratic motivation in slowing the erosion of impartial administration.

Swimming in the flood: ideology and individual-level climate adaptation. with Preston Johnston. Latest version: May, 2026. Post pilot survey draft available here.

How do people adapt to climate change? Although individuals have incentives to reduce their exposure to climate risks, research has shown that risk perceptions and tolerance are socially constructed. We propose that political attitudes may affect the adaptive actions individuals take in response to climate risk. Studying the case of Brazil, a country that faces severe climate impacts, we combine experimental and observational data to examine whether ideology explains variation in adaptation behaviors and whether it moderates responses to risk exposure. First, we field a pilot survey experiment on participants sampled from flood-exposed municipalities, in which we administer an informational treatment revealing respondents’ local flood risk level. We measure how exposure to risk information alters stated and revealed adaptation behaviors, as well as support for adaptation-relevant public policies. Second, we leverage a large, government-administered survey of Brazilian rural households. Using this data, we develop a novel typology of self-reported behavioral responses to climate change. We then examine whether respondents in areas with more right-wing electoral preferences self-report different adaptation behaviors and whether they exhibit lower elasticities of such behaviors to climate risks such as droughts, floods, and landslides. We present pilot findings and discuss design considerations for a larger, pre-registered follow-up survey.

Seeing with blind eyes: scope and limits of politicized enforcement. Latest version: February, 2026. Available upon request.

How does political control shape the weaponization of enforcement? Prior research has uncovered politicized sanctions, yet their institutional determinants remain unclear. I propose that while politicians with formal jurisdiction over agencies can influence the level of enforcement, particularly after appointing new bureaucrats, authorities at other levels have a weaker impact. Moreover, I argue that principal-agent dynamics limit principals’ ability to protect allies from enforcement. I test this argument by studying anti-deforestation enforcement in Brazil. I estimate the elasticity of enforcement to lawbreaking through remote sensed measures and find that opposed principals can decrease its responsiveness only when they have formal control over bureaucrats. Furthermore, I leverage novel administrative data on two million property owners to test of how political connections influence the probability of punishment, finding no evidence of quid-pro-quo dynamics. These findings underscore the institutional constraints that politicians face when they attempt to subvert the rule of law.

Published work

Two paths towards the exceptional extension of national voting rights to non-citizen residents. (2023). with David Altman and Sergio Huertas-Hernández. Journal of Ethnic and Migration Studies, 49(10), pp. 2541-2560. DOI: 10.1080/1369183X.2023.2182713

Only five countries have extended universal voting rights to non-citizen residents for all political spheres (local, intermediate and national): Uruguay in 1934, New Zealand in 1975, Chile in 1980, Malawi in 1994, and Ecuador in 2008. These cases constitute a unique intercontinental medley and an opportunity to study the conditions behind such revolutionary change. Through a calibrated comparative strategy based on most similar system designs (inspired by Mill’s method of difference) using QCA, this paper finds that the extension of national voting rights to non-citizen residents transpired in two distinct scenarios. The first setting (Chile, New Zealand, and Uruguay) took place within unitary states with already-existing local voting rights for non-citizen residents and settler trajectories, but that were not undergoing a liberalization process. On the other hand, the second configuration (Ecuador and Malawi) developed within unitary states that recognized nationality by ius soli and were going through a process of liberalization, but without previous local voting rights for non-citizen residents or a settler trajectory. To our best knowledge, this paper offers the first cross-national explanation that involves all cases that have broadened their respective political communities (demoi) to include national voting rights to all non-citizen residents.

Citizens at the polls: direct democracy in the world, 2020. (2021). with David Altman. Taiwan Journal of Democracy, 17(2), pp. 85-106.

There is suggestive evidence that the growth of democracy has stagnated, and even some signs indicate that democracy is in retreat. In such a context, one might have expected to witness an increase in experimentation with democratic innovations such as direct democracy. This is not the case. While there is a spectacular and statistically significant increase in the uses of mechanisms of direct democracy (MDDs) since the early 1990s, 2020 remained notably similar to the previous years in terms of the level of direct democracy worldwide. In 2019, we witnessed less than half of the MDDs we saw in 2018 (eighteen vs. fifty), but, in 2020, the count bounced back to thirty. The COVID-19 pandemic did not halt the march of direct democracy, although it delayed some of its events. Beyond the specific number of popular votes in 2020, direct democracy still tracks almost perfectly with global electoral democracy trends. When all was said and done, however, thirty MDDs were held in 2020: fourteen obligatory referendums (eight in Liberia, two in Chile, and one in Algeria, Italy, Palau, and Northern Cyprus, respectively), six plebiscites (two in New Zealand, two in Liechtenstein, one in Russia, and one in Guinea), five popular initiatives (four in Switzerland and one in Liechtenstein), and five rejective referendums (all in Switzerland).