Download
Abstract
We examine the impact of expanded local alcohol access on traffic safety, exploiting close local option elections in Texas as a source of quasi-random variation. More than 300 cities in Texas voted to legalize alcohol sales between 2003 and 2019, leading to a large increase in liquor licenses - driven almost entirely by off-premise retailers such as liquor stores and grocery stores - and a reallocation of alcohol sales from bars and restaurants toward retail outlets. Despite these sizable market changes, we find no meaningful increase in total crashes, DUI crashes, or fatal crashes, and no evidence of spillover effects on neighboring cities. We also show that on-premise licenses are positively associated with crash rates while off-premise licenses are not, consistent with the absence of a crash effect when legalization primarily expands retail access.