Measuring Income Group Heterogeneous Treatment Effect of Trade Openness on Unemployment Rate at Country-Year Level

Opening up to trade has alleviated millions of Chinese people from poverty, but also took many jobs away from Americans. What about the rest of the world? Does trade hit each country the same? I estimate the Heterogeneous Treatment Effects of trade openness on the unemployment rate among 180 countries from 1990 to 2020, using the latest trade volume measures and free trade agreements as trade openness measurements. Despite the potential bias of endogeneity problems, I found a strongly significant negative treatment effect for upper-middle income nations (e.g., China and Brazil), a positive effect with small significance for lower middleincome countries (e.g., India and Vietnam), a positive effect with no significance for low-income countries (e.g., African countries like Sudan, Guinea etc.), and a small but mixed effect for high income countries (mostly OECD countries). However, the total number of Free Trade agreements is not a perfect measurement and there are severe endogeneity problems that still need to be untangled in future studies.