In the past two decades, low- and middle-income countries (LMICs),especially in Latin America and Asia, have seen a rise in ideological andaffective polarization, along with democratic backsliding. Yet, research onthe drivers of such polarization’s rise and persistence remains limited. Weidentify three general informational drivers of polarization systematicallypresent in LMICs. First, echo chambers foster and reinforce false beliefsabout opposing groups. Second, while diverse media could counterpolarization, the dominance of biased traditional, digital and social mediahas deepened it. Third, divisive political elites using hostile rhetoric andincivility exacerbate polarization by amplifying social and political divides.We highlight opportunities to further our understanding of the empiricalrelevance of these mechanisms in LMICs, as well as interventions informedby these mechanisms to counter polarization.

Political polarization can have detrimental effects on democratic governance by undermining electoral accountability due to biases in how voters consume and process information. A growing literature, mostly from high-income countries, has identified promising interventions for reducing negative feelings towards out-partisans, and for increasing ideologically opposed information internalization by prompting citizens to be aware of biases stemming from emotions or partisanship. To assess whether these types of interventions can contribute to counterattitudinal information consumption and internalization, we evaluate the effectiveness of different depolarization treatments combined with a bias alert within a survey of nearly 2,500 respondents from a nationally representative panel in the heavily polarized context of Mexico. Videos designed to foster empathy and emphasize shared identities reduce partisan polarization by improving attitudes toward those from opposing parties. However, these depolarizing treatments do not increase the consumption and internalization of ideologically opposed information. A nudge encouraging individuals to adopt an open mindset before engaging with ideologically opposed information, in contrast, does lead to greater information internalization. These findings provide insights into the potential and shortcomings of scalable interventions that can help citizens process politically relevant information, regardless of their partisan affiliations.

We study the political behavior of Americans with immigrant backgrounds, a growing group that accounts for over 20% of eligible voters. Using voter registration records from three states that report birthplace, we characterize voting registration patterns among 7 million naturalized and second-generation immigrants relative to U.S.-born whites between 2014 and 2024. We document a broad Republican realignment among immigrant-background Americans. We then examine whether rising unauthorized immigration, a salient development of the past decade, contributed to this shift. Using Mexican consular data on migrant flows from Mexican municipalities to U.S. counties and origin push shocks, we construct predicted county-level inflows of unauthorized migrants. We find that counties with larger undocumented inflows experienced greater increases in Republican registration among naturalized and second-generation voters relative to U.S.-born whites in the same county and year. These effects are strongest among Hispanic voters. We provide suggestive evidence that social mechanisms related to status and fairness account for these patterns, while economic competition appears to play a more limited role.

We study the political and socioeconomic impacts of unauthorized Mexican migration to the United States. Our shift-share identification strategy combines variation in migration inflows and migrant networks using novel administrative data to capture unauthorized migration. With rich county-level data, we document conservative electoral, legislative, and policy responses. Recent unauthorized migration increases the vote share of Republican candidates and induces more support for conservative legislation. It decreases total public expenditure, prompting reallocation away from education toward support for law and order. Three mechanisms partially explain these effects: job losses in “migrant-intensive” sectors; White flight and residential sorting; and higher out-group bias.

Press coverage: How the surge in migration to the U.S. might affect 2024 (The Washington Post), Immigration Backlash (Marginal Revolution), Why Biden Needs an Immigration Deal (The New York Times), The two sides of immigration backlash (The Science of Politics)

Awards: Robert H. Durr and co-winner of the Best Paper in Political Behavior (MPSA)

We study how unemployment shocks in the US affect Mexican households’ migration decisions. We emphasize households at origin (as opposed to individuals) as the migration decision-making units. We show that negative changes in US labor market conditions, which diffuse through household members at destination to those at origin, generate a stronger reaction from more skilled Mexican households that have members abroad than less skilled households. We argue that the relative magnitudes of income and substitution effects drive this differential response across households with varying skill levels. These results also inform the literature on selection patterns in international migration, which suggests a new channel through which negative shocks in the host economy negatively affect the skill composition of subsequent migrants.

We estimate the impact of the foreign aid program UGDP on electoral outcomes and civic attitudes in Bangladeshi Upazilas (sub-districts). We exploit the selection rules by implementing a regression discontinuity design, comparing otherwise identical sub-districts during the program’s first years. Participating in UGDP has an imprecise null effect on voting behavior in parliamentary elections and a significant and negative effect on civic attitudes, suggesting capture by the dominant party.