Do bans on carrying firearms work for violence reduction? Evidence from a department-level ban in Colombia.

This paper develops a theoretical setting that models gun interactions between workers and
delinquents under a rational expectations equilibrium. We show that a gun ban increases the
gun carrying costs of both workers and delinquents, decreasing unambiguously the fraction of armed delinquents while weakly lowering (not necessarily) the fraction of armed workers. We then evaluate the impact of a temporary ban on gun-carrying licenses in Colombia during December of 2009 up to February of 2010 at the department level to verify our theoretical prediction. The gun control intervention operated by extending law enforcement activities that target gun carriers across territories and periods, thus increasing gun costs for all gun carriers especially for illegal guns. Under a common trend assumption between treated and untreated departments, which is then empirically veriÖed, we exploit regional and temporal variations of the gun ban Önding a large and signiÖcant violence reduction impact, both in terms of fatal (gun homicides drop by approximately a 23% on average) as well as non-fatal gun related intentional injuries (gun injuries drop a 53% on average). Moreover, we do not Önd evidence of an increase in homicides and injuries with non-Örearms, suggesting that the gun ban did not generate a substitution of weapons by potential attackers. Nonetheless, e§ects for only gun homicides seem to diminish as time passes by since the e§ect starts to deteriorate after 40 days after the implementation of the ban even though the enforcement of the ban did not diminish over time.