The Impact of Increasing Community-Directed State Mental Health Agency Expenditures on Violent Crime

John S. Palatucci & Alan C. Monheit

Publication Date: 11/20/2021

Abstract

Violent crime remains a prevalent threat to population health within the United States. States offer varying policy approaches to prevent violent crime and support behavioral health, such as community-based programs that include substance use disorder prevention and treatment. Using state mental health agency data, we construct a panel of U.S. states over nine years and apply an instrumental variables empirical model with state and time fixed effects to adjust for policy endogeneity, omitted variable bias, and time trends. We find that a 10% increase in community-directed state mental health agency expenditures yielded nearly a 4% reduction in violent crime rates. Larger magnitude reductions in violent crime rates were associated with the presence of gun control regulations and increases in the proportion of the population completing secondary education. Policymakers should consider the added benefit of violent crime reduction when considering budgetary allocations of community-directed state mental health agency expenditures.

Introduction

Violent crime remains a prevalent threat in the United States. More than 1.2 million violent crimes (e.g., murder, aggravated assault, rape, and robbery) were committed in the U.S. in 2018 (United States Department of Justice, 2020). Such crimes also exhibit considerable geographic variation across the U.S.; in 2018, more than 800 violent crimes per 100,000 population were committed in Alaska, New Mexico, and Tennessee compared to less than 200 violent crimes per 100,000 population in Maine, New Hampshire, and Vermont. The ongoing violence, including mass shootings, has drawn media attention to the nuanced relationship between mental illness and violent crime (Metzl & MacLeish, 2015). Individuals with psychosis and/or co-occurring substance use disorders exhibit heightened risk for perpetrating violence (Brucato et al., 2019; Soliman & Reza, 2001; Steadman et al., 1998). However, population-based estimates consistently suggest that less than 5% of all violent crime can be attributed to individuals with mental illness (Fazel et al., 2009; Metzl & MacLeish, 2015; Swanson et al., 1990).

By contrast, there is empirical evidence suggesting that behavior associated with alcohol misuse and other substance use dramatically impacts rates of violent crime. In Sweden, Grann and Fazel (2004) merged hospital discharge data with conviction data from 1988 to 2000. They found that 16% of all violent crimes in Sweden were committed by individuals with a primary diagnose of alcohol misuse, and roughly one-tenth of violent crimes were committed by individuals who were diagnosed with other substance use disorders. Young et al. (2021) evaluated self-reported data from Canadian inmates who were convicted of violent crimes and found that 20% of violent crimes were attributable to alcohol, and 26% were associated with other psychoactive substances. A 2017 brief from the U.S. Department of Justice details that 61% of jail inmates incarcerated for violent crimes met DSM-IV criteria for drug dependence or abuse. Approximately one in six inmates reported committing violent offenses to obtain drugs, and 37% of inmates reported being under the influence of drugs when the offense was committed (Bronson et al., 2017). While these studies do not provide the precise number of violent crimes caused by drugs and alcohol abuse, they do invite further population-level research on the nature of violence and efforts to enhance population behavioral health. This includes determining the impact of modifying community-directed state mental health agency expenditures (hereafter, CMH).