Alcohol Availability and Alcohol Consumption: New Evidence from Sunday Sales Restrictions in Canada

Many US states and Canadian provinces have recently lifted bans on off-premise retail liquor sales on Sundays, yet there is no evidence on how such provisions affect drinking. We investigate this issue using unique information on day-specific alcohol consumption from Canada’s National Population Health Surveys (1994-1999) and variation across provinces in the presence of a Sunday sales restriction. In a cross-sectional regression approach, we find that allowing Sunday alcohol sales is associated with more drinking on Sundays but no difference in drinking on other days of the week. To account for unobserved factors that may be associated with both the presence of a Sunday sales restriction and lower Sunday drinking rates (such as religiosity), we use quasiexperimental variation induced by Ontario’s 1997 repeal of its Sunday sales ban. Our difference-in-differences estimates confirm that Sunday sales policies have important effects: Ontario’s Sunday sales liberalization increased drinking on Sundays by about 10 percent and had modest spillovers to consumption on Mondays and Tuesdays. We do not find strong evidence, however, that allowing Sunday sales affects overall population drinking; this suggests that recent state repeals of Sunday sales prohibitions are unlikely to result in large increases in tax revenues as many states have hoped. Our results add new evidence to the large body of research showing how alcohol availability affects point-in-time alcohol consumption.