The Social and Community Consequences of the Opioid Epidemic

A Guide to the Articles

This volume begins with two foundational studies that provide important context for the other ten articles. These two studies are an important starting point for readers new to this area of study. The article by Monnat describes the geographic, racial, ethnic, and gender considerations of the opioid epidemic using 1999 to 2020 data from the Center for Disease Control and Prevention (CDC) WONDER’s (Wide-Ranging Online Data for Epidemiologic Research) online database and vital statistics information to describe time trends in opioid-related deaths by gender, age, race/ethnicity, and geography (county and region). She finds that while fatal drug overdose rates were highest among young and middle-aged adult White and Native American males and middle-age and older Black males, the gap between older Black males and older males in other racial/ethnic groups has grown dramatically since the mid-2010s. In addition, opioid death rates have been consistently high in Appalachia, but in recent years have spread throughout several regions and along the urban-rural continuum. Finally, although opioids have been the main contributor, the contributions of cocaine and methamphetamine have increased dramatically in recent years, demonstrating that our problem is bigger than opioids.

The second foundational study, by Maclean et al., reviews the extensive literature detailing the direct economic costs in the health and criminal justice domains associated with opioid use. They provide a broad overview of the literature focusing on U.S. health and applied economic studies using econometric methods that approximate the scientific method and are designed to estimate the causal effect of policies on opioid-related outcomes. These authors found that the opioid crisis has led to worsening health, increased mortality, increased healthcare use, and modest increases in crime. Further, they found that policies designed to curb opioid use and associated harm have had limited success in reversing the crisis. In particular, prescription opioid suppliers played a central role in sparking the crisis by overprescribing opioids but that state and federal policy responses have demonstrated mixed effectiveness in reducing rates of opioid misuse, overdoses, and healthcare service use.

The rest of the articles in the volume are loosely organized in several groups. We begin with three articles that examine the connection between opioids and child welfare: two of them take advantage of state-level natural experiments (one in Indiana, one in Wisconsin), and one uses national-level data. The next group of articles also focuses on children, with one study of the potential consequences of opioid-related deaths on the living arrangements of children and two articles that focus on the relationship between opioids and educational outcomes. The final group of studies focuses more explicitly on community-level consequences of the opioid epidemic: one article focuses on food insecurity and another examines the effects of the epidemic on homelessness; a third article estimates the costs of the epidemic in terms of lost labor productivity in Canada; and the final article estimates the fiscal impacts of the crisis on local and state governments.

The Individual Studies and Key Findings

Pac et al. use administrative data from Wisconsin to estimate the first causal effects of in utero opioid exposure on infant health at birth and on subsequent reports to child protective services. They focus on mother-infant dyads enrolled in Medicaid between 2010 and 2019, and their findings suggest that neonatal abstinence syndrome and neonatal intensive care unit (NICU) admission bear strong associations with prenatal opioid exposure, concentrated on illicit and medication-assisted treatment (MAT) exposure in the first and third trimesters. Their analyses also suggest that prenatal opioid exposure is associated with increased child protective service reports, low birth weight, preterm birth, and births of babies who are small for gestational age, though these measures are less sensitive with respect to the timing of exposure.

In the summer of 2018, Indiana lifted a long-standing moratorium on drug treatment programs. Bullinger, Wang, and Feder, use this change in policy in Indiana to look at what happened to child well-being in the state when the number of medication-assisted opioid treatment programs suddenly increased. They find that after new opioid treatment programs opened, opioid overdoses showing up in emergency departments in surrounding communities decreased by about 40 percent and out-of-home foster care placements fell by 22 percent. They conclude that expanding access to medication-assisted opioid treatment can have positive downstream effects on children.

With an innovative use of national-level data, Chapman investigates the relationship of opioid use to child maltreatment. He uses CDC opioid mortality data and county-specific child maltreatment reports from the National Child Abuse and Neglect Data System to show that U.S. counties with higher opioid overdose mortality rates also have higher child maltreatment rates. He also finds that poverty exacerbates the effects of the opioid crisis on the incidence of child maltreatment. His findings suggest that reducing the share of families in poverty and implementing opioid-related harm reduction programs may reduce the incidence of child maltreatment.

Caudillo, Villarreal, and Cohen use U.S. Census micro data from 2000 to 2018 to investigate the relationship between opioid use in a community and the household living arrangements of children in that area. They document how the opioid epidemic, as measured by opioid overdose deaths, was associated with statistically meaningful decreasing rates of children living with two married parents, as expected. Furthermore, geographic areas affected by the opioid epidemic have also witnessed increasing rates of children living with two cohabiting (but unmarried) parents, children living with single fathers, and children living with adults who are not their parents (e.g., grandparents). White children experienced more diversification in family structures due to the opioid epidemic than their Black or Hispanic counterparts.

Using data from the Institute of Health Metrics and Evaluation (IHME) and data on standardized test scores by county, Darolia, Owens, and Tyler examined the link between community-level exposure to the opioid crisis and children’s educational outcomes as measured by standardized math and reading test scores in grades three and eight for years 2009 to 2014. Their analysis finds that counties that experienced higher drug overdose rates from 1995 to 2014 have lower average third- and eighth-grade test scores than counties with lower overdose rates. Further, the relationship between higher overdose rates and lower test scores is particularly strong in rural counties. Finally, the places with the highest overdose rates and lowest test scores tend to be economically disadvantaged, suggesting that additional resources may be needed to address the issue.

Drescher et al. provide an analysis of the association between community opioid prevalence and children’s learning rates using national datasets of county-level opioid prescription rates and public school students’ third- to eighth-grade academic achievement. They find that students in counties with high levels of prescription opioid presence are learning at a slightly slower rate over time than their peers in counties with lower rates of opioid prescribing. The negative relationship between prescription opioid presence and student learning remains even when accounting for community and school differences. The link between prescription opioid exposure and learning rates is strongest among economically disadvantaged and Native American children and most pronounced across Appalachia and the South.

Heflin and Sun use administrative data from Wisconsin to examine potential mechanisms between drug-related mortality and food insecurity: they exploit cross-state variations in OxyContin misuse prior to reformulation of the drug to examine whether food insecurity increased as individuals with opioid use disorder transitioned from prescription to street drugs such as heroin. They find that states with higher initial OxyContin misuse rates had an increase in food insecurity after OxyContin reformulation, providing further evidence of the social consequences of the opioid crises and the negative consequences associated with drug reformulation for food security.

Between 2015 and 2018, the federal government awarded $115 million in grants to local entities that was aimed at supporting opioid use disorder treatment programs that integrate evidence-based medications with counseling and behavioral therapy. Sullivan and Park examine the relationship between receipt of these funds and community-level rates of unemployment and homelessness. They find that grants for MAT for prescription drug and opioid addiction was not associated with reductions in homelessness or changes in unemployment rates in the communities where they were implemented.

Focusing on the opioid epidemic in Canada, Cheung, Marchand, and Mark estimate the lost productivity in the Canadian economy due to opioids. They apply two human capital models to the problem and find that the total estimated productivity loss to Canada is at least $8.8 billion; a “value of statistical life” calculation makes the impact estimate an order of magnitude higher than $8.8 billion and challenges the notion that the opioid crisis only affects unproductive members of society.

Finally, Bifulco and Shybalkina estimate the approximate magnitude of the effect of the opioid crisis on state and local expenditures, discussing approaches that might be taken to evaluate impacts on revenues. Their results indicate that fiscal impacts of the opioid crisis on state and local governments are modest for the U.S. as a whole but are likely to be greater in states with particularly high rates of opioid-use disorders. They provide an analytical framework to inform intergovernmental aid policies that are designed to offset the fiscal impacts of opioid misuse.

Research and Data Challenges

The articles in this volume comprise a productive step forward in our knowledge of the ways in which the opioid epidemic effects individual communities and the broader society. But a research enterprise like this one also reveals impediments to high-quality and consistent investigation of these sorts of phenomena.

First is the issue of definitions. There are numerous measures that can be used to measure the intensity of the opioid epidemic, including, but not limited to, mortality, pill distribution, hospital visits, and various other opioid-related markers of health- and criminal justice-involvement. Researchers also at times categorize these measures, such as classifying communities as being “high”- or “low”-intensity areas. The same data and measurement challenges also can be true for social and community outcomes. Furthermore, there are a multitude of ways that researchers define communities, and geographic level of analysis is often reliant on what data is available. Opioid-related measures can also be difficult to link to specific areas. For example, it is well documented that pills and drugs can be exported out of the location where they were prescribed or distributed.

The diversity of data and measures and operational choices by researchers can be a strength since different measures can capture unique features or phenomena. However, researchers using different measures and data can hinder comparisons across studies that can lead to policy solutions. Researchers should therefore be specific about limitations and strengths of data and should be clear about what available data can tell us and what they cannot. Researchers should also share data and processes as much as possible.

Another set of issues are due to methodological concerns. It can be challenging to find exogenous variation in these settings, and threats related to omitted variable bias and reverse causality should be front of mind for researchers. Some creative research approaches have emerged, such as using plausibly exogenous state and county policy variation that could affect prescriptions and marketing (e.g., Alpert et al. 2022; Harris et al. 2020). However, more work is certainly needed to extend these methods into research on societal and community consequences, and researchers should continue to look for creative ways to identify causal effects. And, descriptive work should be valued for its ability to highlight areas where problems are especially acute, in addition to its ability to lay the groundwork for future causal inquiry.

Finally, temporal considerations are also important. Beyond understanding the evolving phases of the opioid epidemic and how they may differentially affect social and community outcomes, there is still much to learn about immediate versus long-term effects. This may vary across social or community domains. For example, we may be able to observe some health consequences immediately, but it may take longer to observe educational or intergenerational effects. It is also important to understand immediate versus cumulative effects of persistent trauma, as policy solutions to address each of these phenomena may differ.

Conclusions for Public Policy

A few policy takeaways are suggested from this work. First, individuals involved with opioid addiction face more than a health issue, and there is a need for holistic policies that treat the whole person. Research suggests there can be ample substitution toward or away from other substances as opioid use fluctuates. Individuals with opioid use disorder can have food and housing needs and often serve important societal roles as parents, caregivers, and employees. While their addiction-related health needs require significant focus, opioid-related policy should address their needs holistically while minimizing the disruptions to other critical dimensions of their life.

Second, individuals involved with opioid use disorder are embedded in families and communities, and the disruption caused by opioid use can cascade out to the family system across multiple generations, impacting the effectiveness of our educational system, straining social services such as child welfare, increasing food insecurity, and reducing productivity, ultimately creating fiscal burdens beyond the direct effects documented to date at the local and state level. This suggests that solutions need to also be far reaching.

Finally, given that the demographics of opioid addiction are shifting to look more like other margins of inequality, policy approaches informed by a social determinants of health frame are increasingly relevant. And given that opioid addiction is morphing once again into a fentanyl crisis, the need to focus on U.S. societal structures that address loneliness and despair is more critical now than ever.