Racial disparities in utilization of substance use disorder treatment among non-incarcerated adults

In 2015, almost 30% of all the estimated arrests in the U.S were related to a drug abuse violation or for driving under the influence of narcotics or alcohol (FBI’s Uniform Crime Reporting). While research has shown that there is a greater likelihood of receiving substance use disorder (SUD) treatment among individuals involved in the criminal justice system (Saloner, et al, 2016), racial and ethnic differences in sources of payment for SUD treatment have not been quantified. This study seeks to provide evidence on racial and ethnic differences in sources of payment for SUD treatment. We are able to quantify if those differences are due to observable characteristics or unexplained factors. Using data from the 2008 - 2016 National Survey of Drug Use and Health (NSDUH), this study analyze data on non-incarcerated individuals with SUD who had any criminal justice involvement in the previous 12 months. An extension of the Blinder-Oaxaca decomposition method for non-linear models (Fairlie, 2017) is implemented to determine to what extent differences in SUD treatment utilization across non-Hispanics blacks, non-Hispanics white and Hispanics are explained by observed and measurable characteristics or unobserved factors. We find that non-Hispanic whites are about 50% more likely to receive a SUD treatment paid by a family member than non-Hispanic Blacks while Hispanics are about 30% less likely to receive a SUD treatment paid by a family member than non-Hispanic whites. We also find that non-Hispanics blacks are 23% more likely to receive a SUD treatment paid by a court, but only 2.5% can be explained by observed factors and the rest is attributed to unobserved factors. This technique is extremely useful to inform policymakers on how much of the differences in SUD treatment utilization among racial and ethnics groups reflect differences in observable or unobserved factors.