2026 Theses Doctoral

Essays in the Political Economy of Immigration

Devaux, Martin Samuel

This dissertation studies the political economy of immigration through the lens of immigrant economic agency. Existing research often examines how natives respond to immigrant presence, labor-market competition, or demographic change. I shift attention to immigrants as economic actors whose choices shape local communities and whose opportunities are themselves shaped by politics. Across three essays, I ask how immigrant entrepreneurship affects native political behavior, how anti-immigration politics affects the perceptions of prospective immigrant entrepreneurs, and how researchers can better understand how and why political effects vary across people and places.

The first essay examines whether exposure to immigrant-owned businesses affects support for the far right in France. I develop a theory in which businesses shape native political behavior through four channels: as symbols, producers, employers, and spaces. Using administrative business data, name-based origin classification, geocoded establishments, and municipal election results, I construct a national measure of local exposure to immigrant-owned storefront businesses. To address the endogeneity of business location, I use a shift-share instrumental variable based on pre-period industrial composition and national trends in immigrant business creation. I find that increased exposure to immigrant-owned storefront businesses raises support for the National Rally. The effect is concentrated among visible businesses owned by immigrants from Muslim-majority countries, while non-visible businesses and visible businesses owned by other immigrant groups do not produce the same pattern. Additional evidence suggests that these businesses increase local market variety without producing clear economic harm. The results indicate that backlash is driven less by material competition than by the symbolic politics of visible immigrant economic activity.

The second essay turns from natives’ reactions to immigrants’ own economic perceptions. In partnership with Adie, a French microcredit organization, I field a survey experiment among prospective entrepreneurs. Respondents are randomly assigned to a neutral control text or to a prime about the possibility that parties favoring stricter immigration policy may come to power. The treatment significantly increases agreement that one’s origin or personal situation could complicate business creation, but does not significantly affect perceived economic difficulty, institutional trust, willingness to continue with Adie, or entrepreneurial intentions. These findings suggest that anti-immigration politics can make origin-based barriers salient before formal policy change occurs.

The third essay develops a framework for decomposing cross-site treatment-effect heterogeneity into observed composition, observed context, shared components, and residual variation. The chapter adapts commonality analysis to multisite experiments and introduces sensitivity tools for assessing whether claims about composition or context are robust to omitted moderators. Simulations and applications show that apparently strong claims about where heterogeneity comes from can be fragile with small numbers of sites, measurement error, and unobserved moderators.

Together, the essays show that immigration politics is shaped not only by demographic presence, but also by visible economic participation, anticipated exclusion, and heterogeneity across social and political contexts.