The first chapter uses five rounds of a household dataset for four major Colombian cities; it finds that after controlling for various socio-economic characteristics, districts with higher inequality are associated with a higher probability of kidnap victimization among household and with a significant increase in decisions to mi -grate abroad. Evidence on positive selection of urban migrants from Colombia is also found; contradicting Borjas's "negative selection theory" and supporting the idea that inequality is fuelling a rain drain" from the country