Research

How does religious observance affect political violence and protests? I argue that heightened observance raises the return to political actions that an actor’s doctrine can frame as worship, while raising the opportunity cost of actions that compete with worship for time. I test this using Ramadan, which rotates through the solar year. In predominantly Muslim states, armed conflict rises by about 4% during Ramadan, driven by religious insurgencies. Hourly military data from Afghanistan show that close combat rises nearly 9% and shifts toward periods after meals, consistent with fasting shaping the schedule of operations. A text analysis of a decade of weekly insurgent newsletters shows that their messaging turns substantially more devotional during the month, tying worship to combat. On the civilian side, protests fall by nearly 7%, and Muslims surveyed during Ramadan pray more and report less political interest and willingness to protest, with no comparable changes detected among non-Muslims. State-scripted religious messaging turns more devotional but no more political. The same shock mobilizes religious militants, for whom combat is worship, and restrains citizens, for whom protest competes with prayer.

How does geopolitical fragmentation reshape urban growth? Using difference-in-differences on 1,396 cities over 1970–2021, I study the 1991 Soviet collapse. In post-Soviet Europe, annual population growth near new internal and old external borders fell by 0.4 and 0.6 percentage points relative to interior cities, implying populations 11–17% smaller after 30 years. Central Asian cities near previously sealed frontiers grew about one point faster. Manufacturing exposure amplifies internal-border losses; coastal access mitigates externalborder losses. These patterns support a market-access interpretation in which fragmentation disrupts production links and market transition exposes locational disadvantages along existing frontiers.

Nation states offer a cohesive identity to a diverse citizenry, at once minimizing their differences and elevating their commonalities. Collective experiences like military service and public schooling often uniformly promote national identity, but others like international sport and war hinge on outcomes: collective success in these arenas may strengthen national identity, but collective failure may undo it. We examine this theory in the context of international cricket matches and religious violence in India. On average, Indian cricket victories reduce the incidence of communal violence, irrespective of the opponent faced. In contrast, Indian cricket losses increase communal violence and this effect is driven by losses against Pakistan. These average effects mask important spatial heterogeneity: districts with fewer Muslims experience a reduction in violence after both victories and defeats irrespective of the opponent’s identity. In districts with more Muslims, however, these effects are reversed: losses raise violence on average and this effect is quadrupled when the opponent is Pakistan, while even wins against Pakistan increase communal violence. Our results suggest that nation building reliant on stochastic outcomes is risky, and can undermine the very cohesion it seeks to build.