Gero Dasbach

Gero Dasbach

EMJMD, M.A. & Doctoral Candidate

Research

Work in Progress

Abstract: Does a common native language affect where multinational firms locate their affiliates within a country? I answer this question at the Swiss language border, which runs inside four cantons, so that laws and taxes are the same on both sides. A fuzzy spatial regression discontinuity design compares communes on the two sides of the border, with the native-language share as the treatment. The outcome is the share of a commune’s foreign ownership links held by parent firms from the matching language zone. At the border, the native share rises by 0.60 and the link share of the matching parents by 0.38. The effect survives balance tests, false borders, and the historical line of 1888. Language decides which parents a commune attracts, net of institutions. A model with a fixed and a variable cost of running an affiliate yields the mechanism tests that firm data can answer.

Abstract: Does a political tie between two democracies raise the foreign direct investment (FDI) that flows between them? Existing work measures politics at the national level, where it barely varies between consolidated democracies. We move the question to the subnational level and exploit the world’s densest twinning network: the Franco-German twin town agreements built for post-war reconciliation. We hand-code and geocode every agreement since 1950 and pair the register with greenfield FDI projects from over 2003–2024, on 9,024 pairs of French départements and German Raumordnungsregionen. In a gravity model with fixed effects on both sides, twinning raises the probability that a pair invests and its expected project count by 56 per cent, as much as doubling distance costs, while adding nothing among pairs that already invest. Only agreements signed before 1990 carry the premium. The pattern points to lower entry costs: municipal political diplomacy, long read as symbolic, moves economic flows.

Abstract: We provide the first global assessment of subnational GDP manipulation. Building on an established GDP-night light elasticity approach, we compare officially reported subnational GDP to satellite-recorded nighttime lights for 1,535 regions in 91 countries between 1992 and 2020. The elasticity is systematically larger for regions in autocratic countries and for regions with higher corruption, consistent with political incentives to inflate reported economic performance. The distortion is concentrated where regions compile their own accounts: the corruption gradient more than doubles in countries whose subnational GDP is built bottom-up, and it is statistically indistinguishable from zero where the national statistical office distributes a national total across regions. Survey-based welfare indicators from the Global Data Lab show no comparable pattern: the gradient in GNI per capita is smaller than in official GDP and vanishes under country-year effects, and in the asset-based Wealth Index it is absent altogether.

Working Papers

Abstract: Economic gains from trade liberalization are accompanied by environmental externalities of increased greenhouse gas emissions. The European Union is currently active on both trade and climate policy frontiers. The recently signed EU-India free trade agreement (FTA) and the operational Carbon Border Adjustment Mechanism (CBAM) are illustrative. By means of a new quantitative trade model, this study uncovers counterfactual changes in trade, carbon emissions, and welfare of the EU-India FTA. The FTA is assessed first as a standalone policy and then in conjunction with CBAM. Global country-industry level input-output trade data, along with data on embodied carbon are used. CBAM reduces trade volumes and carbon emissions, with costs concentrated in Indian basic metals. The interplay of CBAM and the FTA generates a small composition-driven emission leakage in CBAM-covered sectors. Once carbon damages are priced under source-based attribution, India’s welfare verdict on CBAM turns from a small loss to a small gain.