Alcohol and Mortality: Insights from South Dakota's 24/7 Sobriety Project

The CDC estimates that approximately 80,000 deaths annually are attributable to excessive alcohol use, which makes alcohol use the third-leading lifestyle-related cause of death in the US.   While the link between alcohol and mortality has been well-examined, there is little evidence on whether programs that target heavy drinking can affect mortality.  In this paper, we evaluate whether South Dakota’s 24/7 Sobriety Project, which targeted heavy drinking among repeat DUI and other alcohol-related offenders, reduced mortality due to conditions commonly associated with alcohol. 

Alcohol use, particularly among those in the criminal justice system, imposes tremendous costs on individuals and society (Miller et al. 2006).  Policymakers are searching for ways to mitigate the threat alcohol poses to public health and safety even as budgetary pressures force them to consider keeping low-level offenders with alcohol problems in the community (rather than incarcerated).  With more than 5 million individuals under community supervision on any given day, approaches that generate reductions in alcohol consumption have the potential to improve public health and safety.  Monitoring and enforcing restrictions on their alcohol use has proven difficult, in part, because alcohol passes through the body more quickly than other substances.  This makes traditional supervision models, which test infrequently, less likely to detect a violation for alcohol.  Moreover, traditional models often impose sanctions only after a fairly serious violation or series of minor violations and often not until weeks after the precipitating offense.  The lack of certainty and celerity has called into question whether such methods can be effective.  Emerging research from neurobiology, psychology, and economics suggest that punishment certainty is a stronger deterrent than severity (Witte 1980; Grogger 1991; Eide 2000; Durlauf and Nagin 2011) and similarly that individuals value immediate rewards more strongly than delayed rewards (Ainslie and Haslam 1992; McClure et al., 2004).  This tendency is particularly pronounced among alcohol-abusing populations (Vuchinich and Simpson 1998; Petry 2001). 

South Dakota’s 24/7 Sobriety Project first applied the combination of frequent monitoring with swift and certain sanctions among repeat drunk driving offenders and eventually among other alcohol-related offenders.  The program requires that offenders submit to breathalyzer tests twice per day or wear continuous monitoring devices. Those testing positive or missing a test face an immediate, but brief, jail term.  The vast majority of tests were clean (98%), which implies that participants did not drink heavily on most days.  From 2005 through mid-2012, 24/7 Sobriety’s 21,000 participants racked up almost 3.4 million days without a detected alcohol violation. Kilmer et al (2013) found that after 24/7 was operational in a county, repeat DUI arrests fell by nearly 12 percent and arrests for domestic violence decreased by nine percent.  While criminal justice outcomes and budgetary concerns were the primary motivations for the program, this paper explores the potential public health benefits.   Specifically, this paper provides quasi-experimental evidence on whether 24/7 influenced mortality by estimating differences-in-differences models comparing changes in mortality (by cause) among those counties with and without the program in operation.