As I mentioned in yesterday’s post, one of the stronger arguments against the Obamacare-Opioid connection is that the explosion of overdoses among Medicaid expanding states was concentrated in the Northeast corner of the country, particularly northern Appalachia and New England.
Today, I noticed this map making the rounds, which shows that these areas tend to be higher in neuroticism (one of the Big Five personality traits that tend to be fairly reliably measurable and largely heritable) than other sections of the country:
The Neuroticism map looks pretty much identical to this one from 2008 made by the Boston Globe from Peter Rentfrow’s data collection:
In a subsequent article, Rentfrow describes the Northeast as one of three geographic personality clusters of the United States, the “Temperamental & Uninhibited Region.”
And here’s a 2008 article by another set of authors, studying Big Five personality traits among drug users; there’s a small n, but note that current heroin and cocaine users are highest in neuroticism and lowest in conscientiousness, which more-or-less aligns with the regional differences in which places have had huge increases in overdoses since 2010:
A 2014 study in India found a similar association between neuroticism and substance dependence, and a 2017 study with a mostly UK sample also found a particular association between neuroticism and heroin/opiate use:
So it’s at least possible that regional differences in personality are important drivers of where the opioid epidemic has spread fastest, and that these differences were confounded with where the Affordable Care Act was fully implemented in 2014. That said, I know very little about the measurement of personality and how large these regional variations really are or what they mean. Personality factors are obviously important to who becomes addicted to drugs, but so are economics, social networks, availability, law enforcement, and everything else. And regardless of these patterns, there still is the bigger question of why, apart from Ohio and West Virginia, this section of the country was not engulfed in the opiate crisis prior to 2010. There could be an underlying population-level susceptibility that was not clearly visible until higher subsidies for medical opiates made them available, or until changes in illegal drug distribution made them cheaper and easier to get. Or we could simply be seeing a process radiating outward steadily from a few centers of the epidemic, that isn’t primarily driven by population differences at all. I think we are seeing an interaction effect, where changes in health care financing, drug distribution and production, as well as increased subsidization facilitated an explosion in overdoses that was also, indisputably, due to other social and demographic factors as well. I’m bad at predictions, but I’m going to guess that when 2016 mortality data is widely available, the differences between ACA and non-ACA states becomes larger still than it was in 2015- the data that the CDC already shared with the New York Times makes me pretty confident of this. But that won’t settle these questions, or still less tell us what to do.