Research Projects
Research Projects
Working papers
Can immigrants shape political outcomes? I study this question in Argentina during the Age of Mass Migration, when by 1914 the foreign-born accounted for nearly 30% of the population and were predominantly European. I combine novel district-level census data on European immigrant shares with electoral returns from national elections in the 1920s to estimate the effect of exposure to immigrants on political outcomes. To mitigate migrant selection, I implement a shift-share instrumental-variable design. I find that greater exposure to European immigrants increases left-wing vote shares and reduces support for right-wing parties. In a context of political disenfranchisement, immigrants can target voters through alternative channels. Districts with greater European exposure to immigrants exhibit higher membership in worker mutual-aid associations, which amplifies the baseline effect. The political effects are stronger where the second-generation voting base was larger. The electoral results are not explained by alternative factors such as voter urbanisation, literacy or inequality. I also find evidence consistent with complementary roles for a higher number of branches and local political newspapers. The effects are not driven by Latin American immigrants, suggesting that immigrant status alone cannot account for the results. The findings show how disenfranchised migrants can shape politics through civic organisation, political mobilisation, and other indirect channels.
We study the long-term effect of early state centralization in Medieval Ethiopia. Ethiopia was the only African state that enjoyed sufficient military and fiscal capacity to maintain its independence throughout the Middle Ages and the colonial era, and produced accurate historical accounts on its political history. In a regression discontinuity design (RDD) at the 14th century boundary, we show that longer exposure to centralized institutions translated into higher living standards today, due to earlier structural transformation and more secure property rights. Early state formation also changed the ethnic geography of the country. Treated areas are more ethnically diverse today, as captured by linguistic fractionalization, and are less likely to contain a hard boundary separating ethnic homelands.
In today's connected world, migrants remain tied to their origins in ways that allow events abroad to reverberate within host-country politics. We ask whether homeland shocks that target a diaspora's identity can reshape political rhetoric in host countries. To answer this, we assemble a panel of UK parliamentary speeches from 2013 to 2024 and link them to weekly records of religious violence in India, exploiting the timing of violent events. We find that anti-Hindu violence in India leads MPs in Westminster from constituencies with larger Hindu populations to use significantly more anti-Muslim rhetoric. The effect intensifies after the rise of Hindu nationalism in India and holds regardless of whether Hindus are victims or initiators of violence. The effect is driven by diaspora engagement with homeland events, local media amplification, and donor incentives, and is electorally rewarded. Our findings highlight how transnational ties can transmit homeland ideologies into host-country politics, underscoring that debates over migration and integration cannot be understood purely through domestic dynamics.