There’s a common story told in the media about the opioid epidemic: In this telling, it’s not a coincidence that many of the communities that have been hit hard by the drug overdose crisis happen to be Rust Belt and Appalachian towns that have seen jobs leave over the past few years. This economic hardship has fed what some have characterized as, borrowing a term from groundbreaking research by economists Anne Case and Angus Deaton, “deaths of despair.”
A new working paper distributed by the National Bureau of Economic Research, however, pushes back on this common story. Its central finding: Changing economic conditions in the past decade and a half explain less than 10 percent of the rise in drug overdose deaths between 1999, near the beginning of the opioid crisis, and 2015, the latest year with applicable data at the time of the study.
In short, recent economic hardship seems to explain only a small portion of the opioid crisis — and Christopher Ruhm, the study’s author, argued that even his findings may overestimate the impact of worsening economic conditions on the epidemic.
Read more:
What Serena Williams’s scary childbirth story says about medical treatment of black women
Trump’s first full presidential physical exam, explained
“Climate change” and “global warming” are disappearing from government websites
Case and Deaton, whose report addressed opioid deaths as well as suicide and alcohol-related deaths, have already responded to the paper. They don’t dispute its findings. But they argue that it doesn’t address the original definition they use for deaths of despair, which includes deaths that were likely caused not solely by recently changing economic conditions but also shifting social and cultural forces as well as longer-term economic outcomes.
Ruhm, an economist at the University of Virginia in Charlottesville, told me that it’s true Case and Deaton may not have intended their deaths of despair hypothesis to be interpreted as a solely economic phenomenon, but it’s often been interpreted as a one anyway in much of the media and public discussion. At the very least, his paper underlines that recent changes in economic conditions do not explain why the opioid epidemic became America’s worst overdose crisis in US history, leading to tens of thousands of drug overdose deaths annually for more than a decade.
This is crucial not just to understanding the causes of the opioid epidemic, but also the solutions. If the economy isn’t to blame for the crisis, then a better economy won’t fix what currently ails us. So what will?
First, a big caveat: Ruhm’s study is a working paper. That means it has not yet gone through the formal peer review process. That process could expose problems in the research that a more rudimentary read can’t catch. Still, Ruhm is a widely respected researcher, who has done some good work on both economics and opioids before.
For his study, Ruhm looked at the link between a composite of economic conditions and drug overdose deaths as well as other deaths of despair, such as alcohol-related deaths and suicides, at the county level.
To measure this, Ruhm put death rates against a mix of five economic measures: poverty rates, median household incomes, home prices, unemployment rates, and import exposure. “I can’t claim this is completely comprehensive, but it’s far more so than has typically been done in previous works,” Ruhm said.
“You should expect to see higher rates of death in counties that have done worse [economically] than in those that have done better, but also that that should be a causal factor rather than just a correlation,” Ruhm said. His results: “The economic conditions can only explain a small amount — less than 10 percent, probably much less than 10 percent — of the change [in overdose deaths].” The link was even weaker for economic conditions and the other deaths of despair.
In a separate analysis, Ruhm also used the Gini coefficient, which is used to measure economic inequality. He found that increasing county-level inequality generally predicted slower growth in mortality rates, although the results often weren’t statistically significant. But he cautioned that it’s difficult to calculate the Gini coefficient at the county level and changes over time in the county-level Gini coefficient, and he’s working on further research to try to get a clearer overview on the effects of economic inequality on drug overdose deaths.
What’s more, Ruhm argued that the weak connection he does find between economic conditions and drug overdose deaths is likely overstated — because places with worse economic problems also correlate with all sorts of other problems that his model likely couldn’t fully control for.
“For example, people who are less educated are probably at greater risk as this drug environment gets riskier,” Ruhm said. “And in rural areas, education levels tend to be lower. So some of that correlation — which is relatively weak to start with — is not caused by economic conditions; it’s that the people at risk happen to live in areas that tended to have lower economic growth.”
Ruhm claimed that, more than the economy, the bigger driver of overdose deaths was “the broader drug environment” — meaning the expanded supply of opioid painkillers, heroin, and illicit fentanyl over the past decade and a half, which has made these drugs much more available and, therefore, easier to misuse and overdose on.
To analyze this, Ruhm’s study makes an assumption: If economic (or social) factors were behind the rise in drug overdose deaths, then the availability of certain drugs would only change which drugs are linked to overdoses rather than who is overdosing.
Imagine two simplified hypothetical worlds. In the first, economic hardship is the only cause of drug use. So people in the worst economic conditions are the only people who use drugs. (This is not remotely true in the real world, but, again, this is a very simplified hypothetical.) In this world, you would imagine that a recent spike in the supply of heroin would not make it so wealthy people start using drugs, because they’re not going to be using drugs anyway. The only thing an increase in the heroin supply would do is make the drug more accessible — and therefore more used — among the worst-off population that has all the drug users.
In the second hypothetical, the supply of drugs is the only cause of drug use. In this case, everyone in all demographics uses drugs at similar rates, but younger people only use heroin while older people only use opioid painkillers. Which group uses more drugs entirely depends on which drug is more available. (Again, not remotely true in the real world.) So an increase in the supply of heroin would lead more younger people to start using drugs since their preferred substance is more accessible, but it wouldn’t change the older population of drug users since they don’t want any heroin anyway.
Ruhm finds that the real world is closer to the second hypothetical than it is to the first — that is, greater availability of certain drugs has altered who suffers the most from drug overdoses. So as the opioid epidemic has transformed into more of a heroin and illicit fentanyl crisis than one of conventional opioid painkillers, some groups have seen disproportionate increases in overdose deaths — in particular, younger, more urban, more male, and more black populations who are historically at greater risk for illicit opioid overdose deaths than other groups.
In short, this suggests that supply matters more than economic (and social) factors.
Ruhm doesn’t take his findings to mean that economic issues or other factors besides supply played no role in causing the opioid epidemic. In his view, socioeconomic factors and other issues created a powder keg for opioid misuse, and the sudden expansion of the supply of opioids was the match that lit the fuse.
Interestingly, Anne Case and Angus Deaton, the Princeton economists who coined the phrase “deaths of despair,” have used a similar analogy — arguing that the expansion in the opioid supply “poured fuel into the fire.”
Deaton didn’t contest Ruhm’s study results, but he also told me that he doesn’t believe Ruhm’s paper disproves his and Case’s own findings.
“The term ‘deaths of despair’ comes from the fact that the three causes of deaths that were rising most rapidly were suicides, drug overdoses, and alcohol-related deaths,” Deaton said. He argued that his original paper is clear that the causes for these deaths are not just economic issues, but also other issues, including lower rates of marriage, reduced labor force participation, and higher self-reported pain. “We spend about a third of our paper saying that it’s not [just economics].”
One way of summarizing the findings, according to Deaton: “We think that deaths of despair are following from life having lost its meaning for a large section of the American population.” That can be due to economic forces, but it can also be due to all sorts of other issues, from a growing sense of social isolation to mental health issues. After all, suicides and alcohol-related deaths have increased among certain groups, based on Case and Deaton’s work — and something must explain that besides increases in the opioid supply.
Ruhm explained that, although Case and Deaton may not have intended the deaths of despair hypothesis to focus on economics, that’s how it’s been characterized by much of popular media. This is something I’ve seen during my opioid coverage as well: While Case and Deaton’s paper is fairly clear that there were multiple causes of deaths of despair, much of the public dialogue has leveraged the concept to beat up on their hobby horses — particularly economic policies that they don’t like.
Stefan Kertesz, an addiction researcher at the University of Alabama at Birmingham, put it another way: “I think that [Ruhm’s] paper makes a good argument against the public cartoon of what Case and Deaton suggested.”
To the extent economic issues are behind these trends, however, Deaton told me that there could be other economic factors behind an increasing sense of despair. As one example, it’s possible that people are looking at how much better off their parents or grandparents were — indicating a stagnation in social mobility — and that is making them despair. That wouldn’t be captured in Ruhm’s paper, which only looks at economic data going back to 1999. But it’s plausible and worth studying further, Ruhm acknowledged.
Ruhm also said he wants to find a way to measure the decline in “social capital” — a term for some of the cultural and social forces that Case and Deaton point to — and find out if that’s contributed to the rise in drug overdose deaths as well.
“In no way am I claiming that this is the last word on this subject,” Ruhm said. “In my view, it’s the first careful analysis.”
The discussion around Ruhm’s findings helps contextualize not just what caused the opioid epidemic, but how to end the crisis.
If the economy was a primary cause, then looking for ways to boost the economy in worse-off communities could be one way to address the opioid crisis. But if the economy isn’t a major cause or a cause at all, then other solutions are needed. (Of course, boosting the economy regardless of its impact on the opioid epidemic is good on its own terms.)
Ruhm’s findings suggest that solutions specific to drugs and the opioid crisis are needed here. Thankfully, we have a pretty good idea of what to do in this area, as I previously explained: Policymakers could dedicate far more resources to boosting access to treatment (in particular, highly effective medications for opioid addiction), pull back lax access to opioid painkillers while keeping them accessible to patients who truly need them, and adopt harm reduction policies that mitigate the damage caused by opioids and other drugs, particularly illicit fentanyl.
At the same time, some experts have told me that there also needs to be an effort to address the root causes of drug addiction. That’s not just the economy, but also physical, emotional, and mental health, social isolation, a lack of healthy forms of entertainment, and just about anything else that can make a person feel miserable.
As Leo Beletsky, a professor of law and health sciences at Northeastern University, previously told me, “We have a lot of complex problems in this country. Without really addressing all of those physical, emotional, and mental health problems, just focusing on the opioid supply makes no sense — because people still have those problems.”
Ruhm is on the side that a riskier drug environment — one in which drugs are too abundant while addiction treatment and prevention are not — is likely the key driver of the current overdose crisis. Case and Deaton are on the other side; they agreed that supply control and treatment would help, but argued that neither will be enough to fully stop the current overdose crisis and prevent future ones until what they view as the root causes of drug addiction are also addressed.
In an ideal world, perhaps policymakers would take an all-of-the-above approach. But in the real world, financial resources and political capital are limited. Depending on where one lands in the spectrum of this debate, one may take a very different approach with those limited resources to combat the deadliest overdose crisis in American history.
