Anisha Ghosh*, Marco Matani**

*This study investigates whether Ramadan adherence reduces domestic violence against children. According to UNICEF (2024), two-thirds of children regularly experience violent punishment at home. High fertility rates in predominantly Christian and Muslim countries, together with the Islamic “youth bulge” also suggest that by 2050, the majority of young people will have been raised in a religious household (Pew Research Center, 2025). By leveraging the rotating timing of Ramadan, I evaluate how ritual preparation and intensity affect reports of child punishment across a large sample of Muslim adherents, non-adherents, and non-Muslims across Egypt, Nigeria, and the United States. Satellite-based night-time light data from the Visible Infrared Imaging Radiometer Suite (VIIRS) provides an additional measure of Ramadan adherence by capturing the shift of daytime activities to nighttime hours. Estimating the pre-Ramadan period separately also helps disentangle the effects of fasting-related fatigue from those of religiously motivated restraint or forgiveness. This approach builds on recent work using Ramadan timing as a source of exogenous variation in behavior and decision-making, while extending it by incorporating a high-resolution spatial measure of religious activity.

**The present project frames a (revisited) proximity-concentration trade-off of multinational firms within a Ricardian-Heckscher-Ohlin model with variable markups (arising from VES preferences) and multiproduct firms. If frictions on new varieties are higher when producing abroad, firms choosing between foreign subsidiaries and exports weigh the resulting increase in marginal costs against the benefits of avoided trade frictions and possible input price differences. The varieties produced as multinationals are those closer to the firms’ core competency, that is, those produced more efficiently. When multinational operations require additional fixed costs, a novel mechanism obtains under VES compared to CES. Variable markups imply that, for the less productive firms, smaller market shares are compensated by more salient marginal cost reductions. Multinational production is thus optimal for firms that are both productive enough to hold sufficiently large market shares and unproductive enough for cost reductions to significantly affect markups on foreign sales. The most and least productive firms with access to the foreign market remain pure exporters. To the best of my knowledge, this is the first model to characterize firms’ location choices in a setting of monopolistic competition with variable markups. The implications are both theoretical and empirical. While domestic firm selection is still driven by Ricardian comparative advantage, the role of relative input prices becomes less straightforward than in trade-only models. Empirically, the model calls for alternative approaches to studying multinational status and productivity beyond pooled linear regressions.