NRPLUS MEMBER ARTICLE L ast week brought awful news from the CDC, and for a change it wasn’t about COVID-19: After dipping slightly in 2018, raising hopes that the drug epidemic had finally started to wane, fatal overdoses hit a record high in 2019, climbing above 70,000, according to the agency’s preliminary tally.
Then, this Monday, the National Bureau of Economic Research launched another salvo in the war over who’s to blame.
The title of the study, by economists Janet Currie and Hannes Schwandt, gets right to the point: “The Opioid Epidemic Was Not Caused by Economic Distress but by Factors That Could Be More Rapidly Addressed.” The paper is framed as an assault on the economically grounded analysis of researchers such as Anne Case and Angus Deaton. Case and Deaton fit the opioid epidemic into a broader phenomenon of “deaths of despair,” pointing out that suicides and alcohol-related deaths are rising among middle-aged white Americans along with drug overdoses, though not as dramatically — and arguing that a bunch of economic trends beginning around 1970, such as weak wage growth, profoundly affected the cohort that has entered middle age over the past 20 years.
Currie and Schwandt, by contrast, point out that economic changes can’t come close to explaining the stunning rise of drug overdoses we’ve seen since the turn of the century. A simple way of thinking about it is that drug overdoses have tripled in the past 20 years, but economic distress has obviously not gotten three times as bad over the same time. Certainly, the drug epidemic has hit poor and working-class Americans the hardest, with a stark dividing line between those with and without a bachelor’s degree, but that’s true of countless social maladies. And while drug abuse increases when jobs disappear, that correlation is only one small part of the story.
I wrote a skeptical review of Case and Deaton’s book earlier this year, so I’m glad to see Currie and Schwandt pushing back on their arguments. But honestly, the “warring researchers” aspect of the new paper is not what’s most important about it. Currie and Schwandt don’t deny that economic conditions can affect drug overdoses, and Case and Deaton don’t deny the role of other factors, so the two analyses differ in their emphasis more than they contradict each other.
Instead, the most important claim made by Currie and Schwandt is the promise that tackling the problem of drug-overdose deaths directly will be easier and quicker than focusing on the underlying economic factors stressed by Case and Deaton. This claim is true — it is a lot easier to directly combat drug overdoses than it is to reengineer the entire economy. Yet even the former is easier said than done.
A quick recap of how we got here: During the 1990s, doctors became much more serious about treating pain, and the drug industry stepped up to supply new opioids such as Purdue Pharma’s OxyContin, which released its dose slowly so that it could kill pain for a longer time, but also was vulnerable to abuse, particularly by those who crushed and snorted it. These opioids were prescribed in ever-increasing quantities thanks to aggressive marketing efforts and the broad deference given to doctors in the U.S. Many pills made their way to illicit users, and some legitimate users got hooked.
Once the epidemic got out of control, efforts to confront it only backfired. Though Purdue made OxyContin crush-resistant and states started cracking down on over-prescribing, many addicts responded by switching to heroin, which is chemically similar but far more dangerous. In recent years, as the number of opioid prescriptions has fallen, prescription-opioid deaths have leveled off — but deaths from heroin, fentanyl, cocaine, meth, and other drugs have gone up to keep the overall death toll rising:
In short, viewed through the lens of deaths, prescription opioids are certainly still a part of the epidemic, but they’re no longer the dominant driver of it. Yet prescription drugs play a big role in the Currie and Schwandt narrative, and a big plank of their reform platform involves further restrictions on prescribing.
Currie and Schwandt write that “it is important to look at prescription opioids because most people who abuse opioids began with legally prescribed medications. . . . For example, 80% of heroin users began by misusing prescription opioids.” This gets at one of the trickier problems of studying the drug epidemic. Death data tell us what happened at the end of an addiction — what mix of drugs someone had in his system when he died. But they don’t tell us how the addiction started, and therefore can obscure important details. If prescription opioids are overwhelmingly starting the addictions that result in deaths by other drugs, efforts to address prescribing could be a huge part of a solution.
But I don’t think that is clearly the case. For instance, while a 2013 study did find that 80 percent of heroin users had previously abused prescription medications, more recent work suggests that’s changing. A 2017 study of addicts seeking treatment found that while a large majority of those who’d started using opioids in 2005 had first gotten hooked on oxycodone or hydrocodone, the percentage had fallen to about half by 2015, “with no evidence of stabilization.”
It’s also less than clear how people are getting hooked on prescription opioids: At this point, should reforms target drugs that are likely to seep into the black market, or should they aim to further reduce prescriptions in general, even at the risk of denying pain relief to patients who need it? The 80 percent statistic pertains to heroin addicts who previously used prescription drugs non-medically, meaning the drugs were prescribed to someone else or were taken only to get high, and despite years of searching I have not found great data on what percentage of opioid addicts started by taking their own legitimate prescriptions. This small, non-representative, ten-year-old survey puts it at 40 percent, this more recent one at 25 to 40 percent depending on the sample analyzed. (Looking at this issue from the other angle, Currie and Schwandt cite estimates that “21 to 29% of patients prescribed opioids for chronic pain misuse them, between 8 and 12% develop an opioid use disorder, and 4 to 6% of those who misuse prescription opioids start taking heroin.”) That represents a lot of deaths, but not the core of the epidemic.
As for solutions, Currie and Schwandt are fans of prescription-drug monitoring programs — which are already in effect in every state except Missouri, though their strictness varies — and suggest new efforts to identify doctors who overprescribe. That’s fine, but I’m not sure it goes too far toward fulfilling their promise of rapid solutions to the drug epidemic.
They also advocate more help for existing addicts, including easy access to overdose-reversal drugs and medication-assisted treatment. On laws making overdose-reversal drugs more accessible, which a majority of states also already have, they cite a favorable study finding a roughly 10 percent reduction in deaths but don’t mention a less favorable one. On drug treatment, they don’t provide any kind of hard estimate of how many deaths we could prevent through increased funding and access. There’s some promising data showing that treatment reduces the chance of an overdose by half or so when it’s available and an addict takes advantage of it. But a lot of addicts won’t do so.
We could save a lot of lives by combating prescription-drug abuse and treating addiction. But those are only partial solutions to the overall epidemic and would have to be paired with other approaches — including ones directed at the illegal drugs that now cause most of the deaths — to have any chance of truly bringing the problem under control.
To kick off the paper’s conclusion, Currie and Schwandt write that “the implementation of mandatory [prescription-drug monitoring programs], new guidelines for opioid prescribing, and laws promoting naloxone have already made a dent in the epidemic,” and cite the drop in overdoses between 2017 and 2018, the one we now know has disappeared. The timing of this paper’s release was unfortunate, and any sense that we are making tangible progress by implementing known solutions is sadly unjustified.
The crisis — and the debate over who’s to blame — rage on.