In 2019, the Federal Bureau of Investigation reported that there were 8 million crimes a year, which all together cost U.S. citizens $3 trillion. Government policy constantly attempts to address crime and its underlying causes, and research that can identify new policies to reduce crime is paramount.
In “Losing Medicaid and crime” (National Bureau of Economic Research, Working Paper 32227, March 2024), Monica Deza, Thanh Lu, Johanna Catherine Maclean, and Alberto Ortega look at access to health insurance as a crime deterrent that can be targeted through policy. The authors study the 2005 Medicaid disenrollment that occurred in Tennessee, in which 190,000 nonelderly, nondisabled enrollees without dependents lost health insurance. This population had been eligible for Medicaid in Tennessee since 1993 and had generous coverage: preventative care, prescription coverage, and access to mental healthcare, for example. In addition to the disenrollment data, the authors looked at all reported violent and nonviolent crimes from 2002 to 2007 in Tennessee.
The authors find evidence of a sharp increase in crime rates for at least 2 years after the 2005 disenrollment. Prior to the Medicaid disenrollment, the median Tennessee county had 25.8 percent of the population enrolled in Medicaid. Following the Medicaid disenrollment, a Tennessee county with a median amount of exposure to the disenrollment had 1,030 fewer Medicaid enrollees and an additional 270 crimes. This rate of about one new crime for every four people disenrolled equates to over 11 additional crimes per 1000 residents, a 16.6-percent increase. These rates increased by 5.5 violent crimes per 1000 residents and 6 nonviolent crimes per 1000 residents, increases of 20.6 and 14.1 percent, respectively.
Not every violent or nonviolent crime went up because of the disenrollment. Assaults were responsible for the increase in violent crime, as there were 5.3 more assaults per 1000 residents after the disenrollment. Theft drove the rise in nonviolent crime with 6 more thefts per 1000 residents. The authors estimate that the median Tennessee county faced a cost of $288,811 per 1,000 residents from the additional assaults and a cost of $3,952 per 1,000 residents from the additional thefts.
The authors offer several reasons that the Medicaid disenrollment might have affected crime. In accordance with prior research, the authors found evidence of increased poverty and lower average income. Some people covered by Medicaid were close to the poverty line and may have been pushed into poverty after losing the insurance (possibly from medical debt). The authors state that a financially unstable person with decreased disposable income may have more of an incentive to engage in criminal activity. The authors cite research that found evidence of more eviction filings and more completed evictions as well as decreased child support payments after the disenrollment.
Deza, Lu, Maclean, and Ortega acknowledge that their results may not be generalizable for policymakers as it does not address the Medicaid-covered population. Also, insurance options have changed since the Affordable Care Act. Regardless, this study suggests that granting access to health insurance for a generally vulnerable segment of the population can reduce crime.
United States Department of Labor
