João Manoel Pinho de Mello

The Pharmacological Channel Revisited: Alcohol Sales Restrictions and Crime in Bogotá

This paper attempts to evaluate the impact on crime of the restriction of late-night alcohol sales in Bogota and quantify the causal effect of problematic alcohol consumption on different crime categories. It is found that the restriction reduced deaths and injuries in car accidents and batteries. The results are stronger in areas where the restriction was actually binding and are highly heterogeneous depending on the number of liquor stores restricted at the block level. Finally, the paper measures the impact of the restriction on alcohol consumption (the first stage, or mechanism), and quantifies the causal pharmacological impact of alcohol consumption on crime using the restriction as an instrument for problematic alcohol consumption (the second stage). The results indicate that a one standard deviation (s. d. ) increase in problematic alcohol consumption increases deaths and injuries in car accidents by 0. 51 s. d and 0. 82 s. d. respectively, and batteries by 1. 27 s. d.

Short-Selling Restrictions and Returns: A Natural Experiment

Restrictions on short-selling may impede market participants from fully expressing their opinions about an asset, causing departures from price efficiency (Miller (1977)). Measuring the impact of short-selling restriction on returns has been elusive because the decision to sell short reflects expectations on returns. We measure the causal impact of short-selling restrictions on returns by taking advantage of an unique dataset and an unique source of exogenous variation in rental fees. In Brazil during the 2010-2013 period rental transaction from individual investors to mutual funds carried an implicit tax discount on days of distribution of Interest on Net Equity (IoNE). The possibility of tax arbitrage produces an exogenous spike in rental fees and short interest during the days surrounding IoNE distribution, making it prohibitively expensive to short-sell for speculative reasons. Our data contains all rental transaction and the identity of the parts, thus allowing us identify transactions for tax arbitrage. We find that the variation of rental fees induced by the tax arbitrage operations has a large impact on abnormal returns, corroborating Millers hypothesis.