Estimating the Effects of European Union Membership on Income Using Synthetic Control Methods

Abstract: The effect on GDP per Capita of joining the European Union is estimated using synthetic control methods, finding an average effect of roughly a half percent increase in the ten years after joining the union, varying considerably between countries, with both large positive and negative estimates present. When limiting the sample to countries with longer pre-treatment periods, and thus presumably more accurate estimations, the average effect is 3.57%. These estimates are considerably lower than previous estimates using similar methods, which are shown to inadvertently ignore important predictors in the estimation, indicating results are sensitive to methodological specification. Furthermore, results for most countries are not robust in comparisons to placebo-tests in space and time, raising questions about the suitability of the methodology in evaluating some of the EU-entries.