Compliance with mandated health measures has substantial effects on public health, particularly during epidemics. We analyze how political leaders influence such compliance by exploiting President Bolsonaro’s display of skepticism towards COVID-19 during pro-government demonstrations in Brazil. We compare trends of the disease spread around the demonstrations’ date in municipalities with different levels of support for the President. After the demonstrations, the disease spread faster in pro-Bolsonaro municipalities. The results are driven not only by people’s agglomerating during the demonstrations but also by lower compliance with social distancing. Finally, we rule out that pre-existing differences in health-related behavior explain the results.

Electoral support for far-right parties is often linked to specific geographies of discontent. We argue that public service deprivation, defined as poor access to public services at the local level, helps explain these patterns in far-right support. Public service deprivation increases the appeal of far-right parties by making people more worried about immigration and increased competition for reduced public services. We examine our argument using three studies from Italy, a country home to some of the most electorally successful far-right parties in the past decades. We examine cross-sectional data from municipalities (study 1), exploit a national reform forcing municipalities below a certain population threshold to jointly share local public services (study 2), and explore geo-coded individual-level election survey data (study 3). Our findings suggest that public service deprivation helps us better understand geographical differences in far-right support and the mechanisms underlying it.

As the climate crisis worsens, it becomes increasingly important to understand how voters respond to first-hand experience of natural disasters. Conventional wisdom holds that exposure to natural disasters fosters environmental concern, thereby increasing support for green parties and candidates. Looking at the impact of wildfires on voting behavior in Brazil, we argue instead that exposure to natural disasters increases support for green candidates only when the costs of disasters outweigh their benefits. While fires have unambiguously negative environmental and health effects, their economic implications are not necessarily negative. In areas where fires destroy natural vegetation, newly ``cleared'' land may represent an opportunity for land grabbing and the expansion of agricultural and livestock production. We source satellite, administrative, and electoral data from Brazil and use it in two different identification strategies. Our results show that exposure to fires increases support for the main green candidate only in municipalities with low shares of employment in sectors that are likely to benefit from the degradation of the natural environment. Our findings shed light on the distributional implications of environmental degradation and their political consequences.

No abstract

No abstract