Corruption: Transcending borders

Is corruption capable of spreading across national borders? This paper uses panel data for 120 countries from 1995 to 2012 to evaluate whether the corruption levels ofneighboring countries, as weighted by the relative joint border length, affects domestic corruption. Including country fixed effects allows us to control for unobservablecountry specifc aspects and our results suggest a positive and statistically signifcant relationship. In general, a ten point increase in the weighted freedom from corruption index of neighboring countries is associated with a one point increase of the domestic freedom from corruption index. This result is robust to a variety of alternative specifcations, such as a GMM estimation or including additional control variables. The proposed effect becomes stronger as income increases and the relationship is only positive for countries with a GDP per capita above US$1,600 (in 2000 US$). For the richest countries, the estimated coe cient rises up to 0:43.