Abstract
Objectives. To test the associations between local employment opportunities for the Black workforce and drug mortality among Black Americans, while examining the potential moderating effects of fentanyl seizure rates.
Methods. We derived data from the National Center for Health Statistics’ restricted-access Multiple Cause of Death file, linked with county-level job counts, drug supply, and other characteristics from the US Census Bureau and the Centers for Disease Control and Prevention. After examining the characteristics of counties by the magnitudes of increases in drug mortality from 2010‒2013 to 2018–2021, we conducted a first-differenced regression analysis to test the associations between the job-to-Black workforce ratio and age-adjusted drug mortality rates among Black Americans in US counties and test the moderating effects of state-level fentanyl seizure rates.
Results. One more job per 100 Black workers was associated with 0.29 fewer drug overdose deaths per 100 000 Black Americans in the county. This negative association was stronger in the counties of the states with higher increases in fentanyl seizure rates.
Conclusions. Increasing employment opportunities can be an important strategy for preventing Black Americans’ drug mortality, especially among those living in areas with higher increases in fentanyl seizure rates. (Am J Public Health. 2024;114(7):729–732. https://doi.org/10.2105/AJPH.2024.307646)
After a decade of outpacing growth in drug mortality, Black Americans surpassed White Americans’ drug mortality rates in 2020.1 This trend coincides with the beginning of the third phase of the US drug epidemic, primarily driven by synthetic opioids, including illicitly manufactured fentanyl.2 The literature on social determinants of health emphasizes the role of “deep-seated inequalities in living conditions,” such as employment, to better understand the newly emerging drug epidemic pattern.3 Research shows that disconnection from the workforce creates collective frustration and hopelessness, family disintegration, and community violence and crime, increasing drug use as a refuge from psychological distress.2,4 In particular, illicitly manufactured fentanyl is mostly street-supplied5 and disproportionately affects individuals in communities lacking employment opportunities, especially among Black populations.3
Despite the potential contribution of local employment contexts to Black Americans’ drug mortality increases during the 2010s, few studies have investigated this relationship. The present study aims to examine the associations between employment opportunities for the Black workforce and drug mortality among Black Americans throughout the 2010s, while also examining the role of an increased supply of fentanyl, which drives the emerging drug epidemic. Specifically, we linked administrative data on county-level drug mortality and job counts to estimate the associations between the job-to-Black workforce ratio and the Black population’s drug mortality rates in US counties and the moderating effects of state-level fentanyl seizure rates.
METHODS
We drew county-level drug mortality data from the National Center for Health Statistics’ restricted-access Multiple Cause of Death file, a compilation of death certificate data from US Vital Statistics jurisdictions. Drug-involved overdose deaths of any intent were identified according to the International Classification of Diseases, 10th Revision (Geneva, Switzerland: World Health Organization; 1992) codes X40‒X44 (unintentional drug poisoning), X60‒X64 (intentional drug poisoning), X85 (drug poisoning homicide), or Y10‒Y14 (drug poisoning of undetermined intent). For reliable estimation of drug mortality rates among non-Hispanic Black individuals, we pooled 4 years of mortality data (2010–2013 and 2018–2021). Then we linked the data with statistics on county-level job counts and drug supply (including state-level fentanyl seizures from the US Drug Enforcement Administration) as well as sociodemographic data from the US Census Longitudinal Employer-Household Dynamics Origin-Destination Employment Statistics, opioid prescribing rate map files from the Centers for Disease Control and Prevention (CDC), and American Community Surveys. Given the CDC’s suppression of subnational rates with fewer than 10 deaths, our analytic sample was restricted to 214 counties with at least 10 deaths among Black residents in 2010–2013 and 2018–2021, predominantly metropolitan counties with a larger Black population (Appendix Table A, available as a supplement to the online version of this article at https://ajph.org).
Measures
We examined the following measures from 2010‒2013 and 2018–2021 and computed the first differences (value in 2018–2021 minus value in 2010–2013) for our first-differenced regression models. The outcome variable was each county’s age-adjusted drug overdose mortality rate per 100 000 Black residents. The independent variable was a county’s number of jobs occupied by Black workers divided by the county’s number of Black individuals aged 18 to 64 years. As a moderator, we examined the rate of state-level fentanyl or fentanyl analog seizures per 100 000 population. County-level sociodemographic and drug-supply controls included percentage Black population, percentage males, percentage individuals aged 65 years and older, percentage individuals aged 25 years and older without a high-school diploma, percentage veterans, percentage vacant housing, percentage unemployed individuals, median household income, region (Midwest, Northeast, South, West), and opioid prescribing rates.
Analyses
We conducted the statistical analyses in 3 steps. First, we examined the characteristics of counties with varying magnitudes of drug mortality increases among Black residents from 2010‒2013 to 2018–2021. Second, we estimated first-differenced regression models to test the associations between the job-to-Black workforce ratio and age-adjusted drug mortality rates among the Black population and its interaction with state-level fentanyl seizure rates. Our first-differenced regression models removed time-constant unobserved confounders at the county level, allowing for more accurate estimation. Lastly, we estimated Black Americans’ drug mortality rates across the job-to-Black workforce ratios, separately for the counties in the states with varying levels of change in fentanyl seizure rates.
RESULTS
The increases in the drug mortality rate among the Black population from 2010‒2013 to 2018–2021 were not uniformly observed across the counties, ranging from −0.7 to 110.7 per 100 000 population, with a mean increase of 27.5. As presented in Appendix Table B, the top-tercile group—primarily Midwest and Northeast counties with a lower median household income—reported a mean increase of 48.7 deaths per 100 000, signifying a 327% increase. On the other hand, the bottom-tercile group—mostly Southern counties with higher median household income—had a mean increase of 10.6 deaths per 100 000, a 115% increase.
When we adjusted for state-level fentanyl seizure rates and county-level controls (model 2), 1 more job per 100 Black workers was associated with 0.29 fewer drug overdose deaths per 100 000 Black population (see Table 1). Model 3 shows the statistically significant moderating effect of the state fentanyl seizure rates (b = −0.01; P < .001), implying that the negative associations between the job-to-Black workforce ratio and Black Americans’ drug-related mortality were stronger in counties with higher fentanyl seizure rates. For the counties in the highest tercile of fentanyl seizure rate growths, for instance, the magnitudes of increases in drug mortality for having 50 fewer and 50 more jobs (per 100 Black workers) in 2018 to 2021 compared with 2010 to 2013 were 98.6 and 4.9 per 100 000 Black population, respectively (Appendix Figure A). On the other hand, these relationships were nonexistent in the counties of states in the lowest tercile of the fentanyl seizure rate increases.
TABLE 1—
Associations Between Job Opportunities and the Age-Adjusted Drug Overdose Mortality Rates Among Non-Hispanic Black Populations by First-Difference Regression Model: US Counties, 2010‒2013 and 2018–2021
Sociodemographic Characteristic | Model 1 | Model 2 | Model 3 | |||
b (SE) | B (SE) | b (SE) | B (SE) | b (SE) | B (SE) | |
Independent variable and moderator | ||||||
Difference in jobs per 100 Black workers | −0.22 (0.14) | −0.11 (0.08) | −0.29* (0.14) | −0.15* (0.07) | 0.07 (0.16) | 0.04 (0.09) |
Difference in state fentanyl seizures per 100 000 population | 0.09** (0.03) | 0.23** (0.07) | 0.24*** (0.05) | 0.64*** (0.13) | ||
Difference in jobs per 100 Black workers × difference in state fentanyl seizures per 100 000 population | −0.01*** (< 0.01) | −0.56*** (0.14) | ||||
Controls | ||||||
Difference in county-level percentage of | ||||||
Black residents | −4.08** (1.34) | −0.21** (0.07) | −4.24** (1.31) | −0.22** (0.07) | −3.90** (1.27) | −0.20** (0.07) |
Male residents | 2.48 (6.65) | 0.03 (0.07) | 1.66 (6.51) | 0.02 (0.07) | −0.46 (6.31) | > −0.01 (0.07) |
Residents aged ≥ 65 y | −0.17 (2.64) | > −0.01 (0.07) | −0.08 (2.58) | > −0.01 (0.07) | 0.17 (2.49) | < 0.01 (0.06) |
Residents aged ≥ 25 y without a high-school diploma | −3.53* (1.43) | −0.18* (0.07) | −3.17* (1.40) | −0.17* (0.07) | −3.67** (1.36) | −0.19** (0.07) |
Veterans | 0.03 (2.89) | < 0.01 (0.07) | 0.80 (2.83) | 0.02 (0.07) | 1.00 (2.74) | 0.02 (0.06) |
Vacant housing | −0.63 (0.63) | −0.07 (0.07) | −0.54 (0.61) | −0.06 (0.06) | −0.75 (0.59) | −0.08 (0.06) |
Difference in county-level | ||||||
Unemployment rate, % | 0.28 (0.86) | 0.02 (0.07) | 0.16 (0.84) | 0.01 (0.07) | 0.39 (0.82) | 0.03 (0.07) |
Median household income, in $1000 s | 0.03 (0.31) | 0.01 (0.08) | > −0.01 (0.31) | > −0.01 (0.08) | 0.05 (0.30) | 0.01 (0.08) |
Opioid prescribing rate, per 100 population | −0.31** (0.10) | −0.23** (0.07) | −0.25* (0.10) | −0.19* (0.07) | −0.26** (0.10) | −0.19** (0.07) |
Region | ||||||
Midwest (Ref) | 0 | 0 | 0 | 0 | 0 | 0 |
Northeast | −2.33 (3.86) | −0.05 (0.08) | −4.16 (3.82) | −0.09 (0.08) | −1.62 (3.74) | −0.03 (0.08) |
South | −13.86*** (3.61) | −0.35*** (0.09) | −11.62** (3.60) | −0.30** (0.09) | −10.90** (3.48) | −0.28** (0.09) |
West | −8.58 (4.60) | −0.15 (0.08) | −4.72 (4.66) | −0.09 (0.08) | −3.60 (4.51) | −0.06 (0.08) |
Goodness-of-fit statistics | ||||||
Adjusted R2 | 0.17 | 0.21 | 0.26 | |||
F statistics (df1, df2) | 4.44*** (13, 200) | 5.04*** (14, 199) | 6.08*** (15, 198) |
Note. The age-adjusted drug overdose mortality rate measures the number of deaths per 100 000 population. Unemployment rate is calculated as the percentage of individuals in the labor force who are unemployed within the county. The differences in the variables were obtained by taking the differences between the value in 2018 to 2021 and the value in 2010 to 2013. The analytic sample consists of 214 counties with at least 10 drug-involved overdose deaths among non-Hispanic Black residents in 2010–2013 and 2018–2021. The fentanyl seizure rates are available only at the state level, and, thus, the moderator does not account for potential intrastate variations. Estimates and standard errors denoted as “< 0.01” or “> ‒0.01” indicate absolute values smaller than 0.005.
*P < .05; **P < .01; ***P < .001.
DISCUSSION
Our findings offer a salient understanding of the increases in drug mortality rates among Black Americans in the 2010s. First, the drug mortality increases among Black Americans were highest in the Midwest and Northeast counties, especially those with a lower median household income. Economic restructuring (that led to fewer livable-wage jobs in the areas) and increasing presence of heroin and synthetic opioids are considered major drivers of drug mortality in these regions.4,6
Second, local employment opportunities were negatively associated with Black Americans’ drug mortality, consistent with previous findings that highlighted the role of employment in the US drug epidemic.7 This suggests that expanding employment opportunities for the Black workforce can be an important component of effective prevention and recovery strategies against substance use risks in the counties with a larger Black community.8,9
Lastly, the negative associations between employment opportunities for the Black workforce and drug mortality were stronger in the regions reporting higher increases in state-level fentanyl seizure rates, consistent with fentanyl’s role as a moderator of the relationship between job loss and opioid mortality in the general population.10 That is, in the areas with more active distribution and access to fentanyl, employment opportunities can be more helpful in protecting Black Americans from drug mortality.
PUBLIC HEALTH IMPLICATIONS
Reducing drug mortality among Black Americans may require geographically targeted interventions, focusing on the Midwest and Northeast counties with lower incomes. Such efforts may include improving employment opportunities11,12 for the Black workforce through job creation and workforce development (including comprehensive skills and training packages for individuals in recovery) to better align Black workforce skills with the demands of the local labor market, especially in the regions with higher increases in fentanyl seizure rates.
ACKNOWLEDGMENTS
Research reported in this article was supported by the National Institute on Drug Abuse of the National Institutes of Health under award number K01DA057514.
Note. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
CONFLICTS OF INTEREST
The authors report no conflicts of interest.
HUMAN PARTICIPANT PROTECTION
This study was determined to be exempt from the institutional review board at Arizona State University (where statistical analyses were conducted) based on 45 CFR 46.102(e1).
REFERENCES
- 1.Overdose death rates increased significantly for Black, American Indian/Alaska Native people in 2020. Atlanta, GA: Centers for Disease Control and Prevention; 2022. [Google Scholar]
- 2.Dasgupta N, Beletsky L, Ciccarone D. Opioid crisis: no easy fix to its social and economic determinants. Am J Public Health. 2018;108(2):182–186. 10.2105/AJPH.2017.304187 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Friedman JR, Hansen H. Evaluation of increases in drug overdose mortality rates in the US by race and ethnicity before and during the COVID-19 pandemic. JAMA Psychiatry. 2022;79(4):379‒381. 10.1001/jamapsychiatry.2022.0004 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Monnat SM. The contributions of socioeconomic and opioid supply factors to U.S. drug mortality rates: urban-rural and within-rural differences. J Rural Stud. 2019;68:319–335. 10.1016/j.jrurstud.2018.12.004 [DOI] [Google Scholar]
- 5.Mars SG, Rosenblum D, Ciccarone D. Illicit fentanyls in the opioid street market: desired or imposed? Addiction. 2019;114(5):774–780. 10.1111/add.14474 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Jalal H, Buchanich JM, Roberts MS, Balmert LC, Zhang K, Burke DS. Changing dynamics of the drug overdose epidemic in the United States from 1979 through 2016. Science. 2018;361(6408): eaau1184. 10.1126/science.aau1184 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Betz MR, Jones LE. Wage and employment growth in America’s drug epidemic: is all growth created equal? Am J Agric Econ. 2018;100(5): 1357–1374. 10.1093/ajae/aay069 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Park JN, Rouhani S, Beletsky L, Vincent L, Saloner B, Sherman SG. Situating the continuum of overdose risk in the social determinants of health: a new conceptual framework. Milbank Q. 2020;98(3): 700–746. 10.1111/1468-0009.12470 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Hawk K, Grau LE, Fiellin DA, et al. A qualitative study of emergency department patients who survived an opioid overdose: perspectives on treatment and unmet needs. Acad Emerg Med. 2021;28(5):542–552. 10.1111/acem.14197 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Dean A, Kimmel S. Free trade and opioid overdose death in the United States [erratum in SSM Popul Health. 2020;12:100714]. SSM Popul Health. 2019;8:100409. 10.1016/j.ssmph.2019.100409 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Manyika J, Lund S, Chui M, et al. Jobs lost, jobs gained: workforce transitions in a time of automation. McKinsey Global Institute; 2017. [Google Scholar]
- 12.Sigurdsson SO, Ring BM, O’Reilly K, Silverman K. Barriers to employment among unemployed drug users: age predicts severity. Am J Drug Alcohol Abuse. 2012;38(6):580–587. 10.3109/00952990.2011.643976 [DOI] [PMC free article] [PubMed] [Google Scholar]