The results show that between the years of 1983 and 2002, controlling for the impact of arrests, police, the minimum wage, the number of the AFDC cases [Aid to Families with Dependent Children], and the size of the population ages 15 to 49, an increase in alcohol consumption has a positive impact on assaults, rapes and grand larcenies, but alcohol has no impact on murders, robberies, burglaries and motor-vehicle thefts. On the other hand, murder arrests deter murders, robbery arrests deter robberies. Similarly, an increase in own-arrests generate reduction in burglaries, motor-vehicle thefts and grand larcenies. The magnitude of the impact of arrests, in terms of elasticities, are greater than the impact of alcohol, except for rape and assault. P. 20.
From a new paper by Corman & Mocan (January 2013). The title is "Alcohol Consumption, Deterrence and Crime in New York City."
You should see figure 1 in page 24.