The Opioid Epidemic and Homicide in the United States

• Author: Randolph Roth,Joel Wallman,Richard Rosenfeld
• DOI: 10.1177/0022427820986848
• Date: 01 August 2021
• Published date: 01 August 2021
• Subject Matter: Articles

Article

The Opioid Epidemic

and Homicide in the

United States

Richard Rosenfeld

1

, Joel Wallman

2

,

and Randolph Roth

3

Abstract

Objectives: Evaluate the relationship between the opioid epidemic and homi-

cide rates in the United States. Methods: A county-level cross-sectional

analysis covering the period 1999 to 2015. The race-specific homicide rate

and the race-specific opioid-related overdose death rate are regressed on

demographic, social, and ec onomic covariates. Results : The race-specific

opioid-related overdose death rate is positively associated with race-

specific homicide rates, net of controls. The results are generally robust

across alternative samples and model specifications. Conclusions: We inter-

pret the results as reflecting the violent dynamics of street drug markets,

although more research is needed to draw definitive conclusions about the

mechanisms linking opioid demand and homicide.

Keywords

violence, drug abuse, drugs, communities and crime, causes/correlates,

crime

1

University of Missouri, St. Louis, MO, USA

2

The Harry Frank Guggenheim Foundation, New York, NY, USA

3

Ohio State University, Columbus, OH, USA

Corresponding Author:

Richard Rosenfeld, Department of Criminology and Criminal Justice, University of Missouri,

St. Louis, MO, USA.

Email: richard_rosenfeld@umsl.edu

Journal of Research in Crime and

Delinquency

2021, Vol. 58(5) 545-590

ªThe Author(s) 2021

Article reuse guidelines:

sagepub.com/journals-permissions

DOI: 10.1177/0022427820986848

journals.sagepub.com/home/jrc

Explanations of homicide in the United States focus primarily on commu-

nity conditions such as economic disadvantage, racial segregation, and

racial inequality as causal factors (Peterson and Krivo 2010). Some expla-

nations also highlight the role of illicit drug markets in generating violence,

with special attention devoted to the inner-city crack-cocaine markets of

the 1980s and 1990s. By contrast, researchers have devoted little attention

to a possible connection between homicide and the current opioid epidemic.

The crack markets elevated homicide rates (Blumstein 1995). We propose

that the opioid epidemic, too, has exerted upward pressure on homicide

rates.

Three facts about the opioid crisis inform our analysis of its impact on

homicide: its sheer magnitude, demographic distribution, and drug type.

Opioids were involved in 47,600 overdose deaths in 2017, according to the

CDC (cdc.gov/drugoverdose/data/statedeaths.html). The opioid epidemic is

disproportionately concentrated in the Non-Hispanic white population

(Hansen and Netherland 2016; Jones et al. 2015). Users typically began

by acquiring synthetic opioids, such as Oxycontin and Vicodin, through

their own prescriptions or those of family and friends. As these sources of

supply ran out, some users then turned to street drug markets to acquire

synthetic opioids and heroin (Kolodny et al. 2015).

Illicit drug markets are risky, “stateless” social spaces. Conflicts involv-

ing buyers and sellers over price, quantity, purity, and other terms of trade

cannot be settled by the police, courts, or other formal agents of dispute

resolution. When access to formal social control is cut off, violence

becomes a potential enforcement mechanism (Donohue and Levitt 1998).

Growing demand entices more sellers into the market, increasing competi-

tion and potential conflict between them. All else equal, expanding drug

markets should generate more drug-related violence.

That is the underlying theoretical logic of past research connecting the

crack-cocaine epidemic to mounting youth homicide rates during the 1980s

and early 1990s (e.g., Blumstein 1995; Goldstein 1985). We believe the

basic logic holds for the more recent opioid markets, but there are two

notable differences between the two drug epidemics that have important

implications for measuring the degree of violence resulting from the current

opioid crisis. One difference , as mentioned, is demographic: Th e crack

epidemic, at least in its beginning years, was concentrated among

inner-city youth. The opioid epidemic, by contrast, is more evenly spread

across age groups and big cities, suburbs, small towns, and rural areas

(Hansen and Netherland 2016). This is one reason the impact of the opioid

546 Journal of Research in Crime and Delinquency 58(5)

crisis on homicide rates may not be as readily apparent as the impact of the

crack markets.

The other difference between the two drug epidemics concerns the leth-

ality of the substances involved, with overdose risk for users of illicit opioids

considerably higher than it was for consumers of crack cocaine. In this study

we use variation in opioid-related death rates to measure variation in the

prevalence of opioid use. Opioid-related deaths, however, may reflect not

only the number of users but also the lethality of the drugs being consumed.

As discussed below, we are reasonably confident that opioid-related deaths

constitute a valid measure of opioid use in the United States.

We evaluate the connection between homicide and the demand for

opioids in a cross-sectional study of counties and county clusters between

1999, the beginning of the current opioid epidemic, and 2015. Our analysis

controls for multiple demographic, social, and economic conditions shown

in prior research to have significant effects on U.S. homicide rates. We also

estimate the association between the opioid death rate and the county cov-

ariates to describe the sociodemographic conditions that characterize geo-

graphic areas with differing levels of opioid demand. Because the opioid

epidemic has been disproportionately concentrated among whites, whereas

homicide rates are higher in the black population, we estimate race-specific

models of homicide rates and opioid death rates. Small population counts

and missing data precluded county-level estimates of homicide and opioid

death rates for other races and Hispanics.

To our knowledge, this is the first study of the relationship between the

opioid epidemic and homicide.

1

For that reason alone, the study should be

considered exploratory. In addition, the analysis is cross-sectional and

causal inferences are therefore uncertain. County-by-year data sparseness

prevented us from conducting a panel analysis. Finally, the analysis does

not include a direct measure of illicit drug market activity, the presumed

connecting link between opioid demand and homicide, and so any associ-

ation between the two should be viewed as warranting further investigation.

The following sections present the rationale for the study, data and methods,

results, and conclusions.

Background

An extensive body of research documents the social sources of homicide in

the United States. Homicide rates are elevated in cities and states with high

levels of economic deprivation and population density (Land, McCall, and

Cohen 1990; McCall, Land, and Parker 2010; Messner and Rosenfeld

Rosenfeld et al. 547