Local Liquor Control from 1934 to 1970

Abstract

This work considers the state liquor policy over the period 1934 to 1970 as a case study of decentralization. While historical analyses of liquor control have tended to focus on the Prohibition period (Miron and Zweibel, 1991), the period following re-legalization in 1933 remains largely unexplored. The 21 st Amendment explicitly assigns to states the power to regulate liquor, and state policies can be grouped into three categories: legalization throughout the state, prohibition throughout the state or local option. Under local option, local governments, such as municipalities or counties, set their own liquor legalization policy typically through a referendum. States with local option have decentralized liquor policy to local governments. The traditional economic theory of federalism posits that more heterogeneous preferences result in more decentralized policy-making (Oates, 1972). This suggests that a state will select local option/decentralization when citizens on both sides of the legalization issue have intense preferences. We test this presumption by comparing preferences in centralized states with preferences in decentralized states. In states with local option, we can observe how local characteristics such as demographics and religious affiliation influence the probability that a county's residents choose to legalize liquor. We interpret this local policy choice as a measure for the unobserved local preferences over liquor policies. In centralized states, no local liquor policies are observed but we observe the local characteristics (such as religion) whose relationship to local liquor policy is known from the decentralized states. In this manner, we generate for every state the distribution of local tastes and use these to test the theoretical prediction. Our results suggest that states do decentralize liquor policies when there are intense preferences on both sides of the issue, just as theory predicts. While this is just a case-study, the lessons of this paper can likely be applied to other policy-settings. The intuition that preference heterogeneity induces decentralization immediately applies to any other binary policy issue such as whether to permit the death penalty or to legalize abortion. Under certain conditions this logic can also be extended to the more typical case of continuous policy issues. And while this paper focuses on state decision-making, the reasoning also holds for central governments. This suggests one of the forces contributing towards the