Can well-defined access rights to publicly owned land be as effective as privatization in increasing productivity and wealth? In this paper, I evaluate the impact of public property rights using the 1934 Taylor Grazing Act, which determined secure access rights for ranchers to newly created, large grazing districts in the Western US. Using satellite-based vegetation data, I exploit spatial discontinuities across grazing district boundaries and find that public lands with well-defined access rights for ranchers are at least 10% more productive than lands without. Immediately after establishing grazing districts, ranchers inside these districts held more cattle, reported higher income and farm values than their counterparts outside. Despite ranchers being unable to invest in publicly-owned lands, these magnitudes are similar to outright privatization. Instead, I argue that secure access rights resolve uncertainty around future usage and align the incentives of ranchers and regulators, thus incentivizing sustainable and profitable usage. I provide two results supporting this hypothesis: Areas with stronger pre-reform state capacity show larger increases in vegetation. Monthly patterns on vegetation are consistent with the adoption of productivity-increasing fallowing practices. I investigate alternative explanations, and find no empirical support for differential initial productivity; negative spillovers; or systematic local manipulation of boundaries.

Does access to education facilitate the emergence of a human capital elite from which social activists, and thus, social change can emerge? Assembling a city-level panel of the political, intellectual, and economic elite throughout German history, we find that the opening of secondary schools for women increased their representation among the human capital elite. Women became recognized for living independently, for becoming authors, and for actively fighting for women's rights, gradually bringing about social change. Several other city-specific indicators of economic and gender-specific cultural change are unrelated to the increasing representation of women among the human capital elite.

What is the political legacy of state repression? Using local variation in state repression during the Khmer Rouge genocide in Cambodia, we investigate the effects of repression on political beliefs and behavior. We find that past state repression decreases votes for an authoritarian incumbent while enhancing electoral competition and support for democratic values four decades later. At the same time, individuals become more cautious in their interactions with the local community: they exhibit less trust, participate less in community organizations, and engage less with local government. Our theoretical model suggests that these opposing forces arise because experiencing repression bolsters preferences for pluralism while also heightening the perceived cost of dissent. Consequently, citizens are more likely to support the opposition in elections (voice) but engage less in civil society (exit) to avoid publicly revealing their political views. Exploring channels of persistence, we demonstrate that repression cultivates a lasting fear of violence as a societal threat, and that genocide memorials and remembrance ceremonies maintain the collective memory of the atrocities.

If trading goods is beneficial, whom do these gains accrue to? In this paper, I use the unique setting of Africa and study how the welfare gains from trading are distributed, and how this affects societal development. Exploiting a shift-share setting, in which the welfare gains from increased trade activity (shift) are assigned to each ethnic group (share), I find that the gains from trade are unequally distributed: Political elites in power redistribute from larger ethnic groups. This reduces people's satisfaction with democracy and trust in institutions, casting a shadow on the implementation of trade liberalization policies in developing countries. Instrumental variables estimating exploiting exogenous variation in trading activity confirm initial results.