The Welfare Effects of Blue Laws: Evidence from Pennsylvania Liquor Retailing

I study the efficiency consequences of how many days that retail stores operate. Just as free entry of firms can lead to inefficient outcomes, free choices of days of operation can also lead to inefficient outcomes. While consumers are unambiguously better off with more days and hours of operation, producers can be either better or worse off depending on consumer preferences and market structure. I study the liquor market in Pennsylvania and quantify consumer gains from changes in the number of Sunday-opening stores and evaluate the welfare effects of government policies regulating Sunday openings. Pennsylvania relaxed its ban on Sunday sales, allowing me to document the impact of Sunday operation on consumption patterns. Using data on before and after the policy change, I document that an increase in the number of Sunday-opening stores generates substantial “new” sales and leads to consumer substitution, both between stores and between days of week within a store. To directly quantify welfare benefits, I estimate a random coefficient model of demand using detailed store-level sales data. The demand estimates allow me to simulate consumer choices and I find that opening all stores on Sundays would increase consumer surplus by 3.6% compared to the actual policy in Pennsylvania with a quarter of stores open on Sundays. I also use the model to estimate the welfare-maximizing number of Sunday stores and find that the actual number is 70% below the optimal level. Finally, I find that private stores, whether atomistic or monopoly, would deliver more Sunday stores than the observed number.