Consumption of alcohol products imposes substantial costs on individuals and society. Alcohol is among the top three leading causes of preventable disease and death in the U.S. The economic cost of excessive drinking was estimated $223.5 billion in 2006 due to lost productivity, healthcare, criminal justice, and other costs. Policymakers have enacted myriad policies to reduce consumption and consequently harms including taxation, Sunday sale bans, outlet density limits, alcohol control systems and age limits, among others. But the findings with respect to the effects of these policies are often mixed with some studies finding large effects while others find marginal or no significant effects. The inconsistency in the findings across studies is not surprising: Evaluating these relationships is empirically challenging in large part due to concerns about selection and heterogeneity. This study exploits a rare natural experiment to estimate the effects of alcohol policies on consumption and harms while simultaneously addressing selection and heterogeneity. Individuals select into neighborhoods, and hence policy environments, based on preferences for health and other unobserved factors. As a result, selection may bias estimates of policy effects when using observational data. Such bias may even undermine longitudinal studies if policy changes are influenced by changes in preferences. Because random assignment to neighborhoods is generally infeasible, evidence from natural experiments is desperately needed. An additional concern with respect to inference is heterogeneity. Alcohol policies may not affect all individuals equally and so may average out to small or null effects, thereby yielding misleading conclusions. An emerging literature suggests that risk preferences may play a role. Risk preferences vary considerably and have been linked with health behaviors. And perceptions of the risk from substance use are correlated with attitudes toward control policies. It seems likely then that individuals with greater risk tolerance might be differentially impacted by policies. However, this has not yet been examined empirically. Our innovative study will exploit a unique natural experiment, compulsory reassignment of military personnel to new installations and hence different alcohol policies, to examine the effects on consumption and harms as well as whether risk preferences moderate these relationships. Military personnel's compulsory relocations are a result of the military's needs and so exogenous to individuals' health-related behaviors. To leverage this natural experiment, we utilize three waves of the Department of Defense Survey of Health-Related Behaviors, which collects detailed data on substance use and harms. Importantly, the survey also collects information on individual's risk preferences, which allows us to explore the important yet vastly understudied issue of how preferences moderate responses to policy efforts. While there is often a trade-off between causality and generalizability with natural experiments, military personnel exhibit the same troubling patterns of substance use as civilians, potentially indicating that generalizability may be a limited concern. This study will have a high impact because it addresses one of the most critical and costly health issues and because it is highly innovative in both its utilization of plausibly exogenous variation in policy environments and in its focus on the role of risk preferences in understanding heterogeneity in policy effects. The findings will yield important policy implications not only about ?what works?, but also ?for whom? it may work best.

Public Health Relevance

The goal of this study is to utilize rarely-available exogenous variation in alcohol control policies generated by a unique natural experiment to provide a more rigorous assessment of how those policies influence consumption and harms overall and across socioeconomic groups and to examine whether risk preferences moderate these relationships. Gaining a better understanding how policies affect substance use has the potential to inform efforts to reduce alcohol-related harms as well as the potential to address health disparities.

Agency
National Institute of Health (NIH)
Institute
National Institute on Alcohol Abuse and Alcoholism (NIAAA)
Type
Research Project (R01)
Project #
5R01AA025622-02
Application #
9556966
Study Section
Special Emphasis Panel (ZRG1)
Program Officer
Bloss, Gregory
Project Start
2017-09-05
Project End
2019-05-31
Budget Start
2018-06-01
Budget End
2019-05-31
Support Year
2
Fiscal Year
2018
Total Cost
Indirect Cost
Name
Rand Corporation
Department
Type
DUNS #
006914071
City
Santa Monica
State
CA
Country
United States
Zip Code
90401