Welcome to Painesville
Home to only 5% of the world’s population, the US consumes an almost unbelievable 80% of all opioids produced globally. The US is considered to be the ‘epicentre and the origin’ of the opioid crisis, holding a clear global monopoly on the consumption of pain medication. Opioid prescriptions in the US have reportedly quadrupled in the last 20 years (CDC, 2017).
The aggressive rise in opioid prescriptions fueling the drug crisis has formed the basis of more than 2,700 lawsuits, with the little town of Painesville, Ohio, becoming a landmark case that represents the true scale of the crisis. Between 2006–2014, for example, a single Rite Aid store located in Painesville (population, 19,524) recorded the sale of 4.2 million oxycodone and hydrocodone doses.
Oxycodone, colloquially known as ‘hillbilly heroin’ , is a semi-synthetic opioid that is loosely related to morphine, and constitutes the key ingredient of OxyContin, a prescription drug that has come to be regarded as ‘ground zero in an opioid epidemic that has engulfed the United States’ (Walters, 2017).Tragically, around 40% of all US opioid overdoses recorded include the use of prescribed opioids. The drug crisis does not seem limited exclusively to opioid over-prescription; a rapid rise in amphetamine based-ADHD prescriptions have also risen at an exponential rate. Between 2003 & 2015, for example, ADHD prescriptions amongst the 25–29 yr age group increased by 700%.
Over the last few years, geographic patterns in opioid death and over-prescription rates made some political observers sit up and take note. It was something that caught my interest too- at the time, I was burning the midnight oil, night after night, working on a manuscript for a political textbook that focused on the role of evolutionary biology in shaping voting preferences (a super interesting but as yet under-researched concept).
One study that piqued my interest at the time was a Medium article written back in 2016 by historian Kathleen Frydl. Frydl had noticed an unusual correlation — that all the US states affected worst by the opioid crisis had either flipped the hardest to Trump, or recorded real landslides in the Presidential Election that had taken place that year. The stats were striking enough for Frydl to write about it under the compelling title of ‘Oxy Electorate’ (Frydl, 2016)-an article that captured the attention of many political commentators. Overall, 77 of the 82 counties with very high levels of opioid deaths (15+ people per 100) voted for Trump in 2016. The Economist performed similar analyses and mirrored Frydl’s general observations.
Comparing Frydl’s 2016 observations to 2020 shows that those observations still hold. Elliott County, Kentucky, which has been referred to as ‘ground zero of the opioid crisis’, flipped hard for Trump with a historic landslide of 70% in 2016 and an even more impressive 75% in 2020 (Trump won Kentucky, one of the states hit hardest by the opioid crisis, by 62% in 2020). Similarly, neighboring West Virginia, one of the most pro-Trump states in the US, recorded the highest rate of fatal drug overdoses in any state, in addition to the highest reported rate of babies born on opioids (Joseph, 2016) (Trump swept to victory in West Virginia in 2016 and 2020 with 68.5% and 68.7% of the vote respectively — his highest overall winning margin for any state).
This year the highest opioid prescription rates in the US all occurred in red states — Kansas, Oklahoma, Arkansas, Louisiana, Mississippi, Alabama, Tennessee, South Carolina, Kentucky, Indiana, West Virginia & South Carolina. Ohio, a battleground state in the 2020 US Presidential election, has just experienced one of its deadliest years in terms of opioid overdoses, with more residents dying in May 2020 than in any month in at least 14 years (according to Associated Press statistics). Trump won the state by 53.4%.
Observers generally attributed the observed pattern between opioid use & voting preferences to a range of socio-demographic factors — poverty levels, for example, and levels of unemployment. All sage observations.
I wondered if there might also be a biological or neurological cause at play, however. I’d already noted patterns in financial markets upswings & crashes that could be at least partly (& powerfully) attributed to endocrinological factors (the focus of one of my previous textbooks). Those endocrinological factors strongly motivated emotional, as opposed to rational, trading and investing— & I was starting to observe some pretty similar behaviors amongst political voters.
It was an idea that I set aside until I happened to chance upon a medical study weeks later that focused on the neurological side-effects of drug use. And once I’d read it, I just couldn’t shake off. It is an observation that, I felt, on reflection, deserved a great deal more attention.
The study in question? ‘From painkiller to empathy killer’, published in the respected medical journal Social Cognitive and Affective Neuroscience in 2016 .
The authors of the study reported clear findings that painkillers carry the dangerous side-effect of neurally blunting a users’ ability to experience empathy, engage in moral processing & effectively recognize emotions in others, particularly fear. The authors concluded that ‘Quite literally, acetaminophen reduces one’s ability to feel another’s pain.’
That piqued my interest, because by then I’d written an entire textbook chapter that focused on the neurological & physiological reasons that right-wing populist political ad’s focused so hard on ramping up fear, anger & hate — because those emotions were statistically significantly far more likely to mobilize voters to vote and to engage in pro-party online behaviors. What if America had, as a vastly over-prescribed nation, quite literally (& unwittingly) medicated out its collective ability to experience empathy for others?
Trumps’ early campaigning, which ran from around 2015 onwards, was an extremely empathy deficient affair with his campaign strategy focusing hard on vilifying, ridiculing, fearmongering & scapegoating minorities & political opponents (who could forget the ongoing chants of ‘Lock Her Up’?). It coincided, timewise, with an explosion in the opioid crisis; 2016, the year that Trump was elected POTUS, was also a record year for opioid-related deaths (42, 000)-more than any previous year on record, according to the US Dept of Health & Human Services.
Did this observation, I wonder, hold for opioids and amphetamines as well as acetaminophen? I mined the data and yes, the medical evidence stated quite clearly that those particular side-effects held for not only painkillers & opioids, but ampthetamines & ADHD medication, too.
Now, the status of America as the most medicated nation in the world might not directly explain why crazy uncle Bob throws his beer cans at the TV every night, ranting about Birther and pizzagate conspiracies. But it might shed just a little light on what has felt, in recent years, to be a ground shift in the national consciousness of America’s ability to intuitively practice respect & kindness for his fellow man. Perhaps the country really has experienced a medically-induced slump, however small, in its collective ability to empathize.
A gargantuan 191m pain medication prescriptions were written in the US in 2017- a phenomenal amount. Hell, that’s a lot of people out there unwittingly having their empathy potentially switched off and throwing their beer cans at the TV.
At that point I turned to political ad spend data. I wondered if this kind of medical data pertaining to, say, opioid use and predicted side-effects, could be weaponized to inform political-ad micro-targeting strategies.
Consumption of political ad content in 2016 was already notable (Facebook made $2.2bn in political ad revenue in 2018 alone). It was also unbalanced; in the 2016-cycle, Trump campaign ran 5.9m Facebook ads compared to Hillary Clinton’s 66,000. By 2020, however, that figure had been completely eclipsed, increasing by 50% to hit a mammoth $10.4bn (Center for Responsible Politics).
According to The Atlantic, The Republican National Committee and the Trump campaign reportedly compiled an average 3,000 data points about every voter in America. Presumably that data includes health information of some kind. Why would it not? The Washington Post noted recently that ‘Health-care researchers say geographic data could be used to target communities with a high percentage of people with substance abuse disorders’. Targeting counties on the basis of opioid use would fall squarely into such a practice. ‘Political neurotargeting’ is, as it turns out, a real, living-and-breathing concept now commonly practiced by political consulting firms (Hegazy, 2019).
Sweating for months over the creation of a manuscript for a politics textbook in 2018–2019, I had already dived deep into the concept of the neurological targeting of voters. It was a fascinating, mind-blowing journey where truth and fiction felt like it blurred hundreds of times over, so bizarre were the events and strategies that I continued to stumble across on a day-by-day basis. Asking, by that point, if opioid side-effects could have been weaponized in a political context, seemed to me to be a logical, rather mundane question, not a fantastical one, so sensationalized had the cut & thrust of modern politics become. And I wasn’t the only person thinking about it. I stumbled across another great article around that time, published in Scientific American in 2017, that warned of how ‘Medical data, both de-identified and re-identified, could…become national security weapons’.
As the MIT Technology Review stated recently, ‘If you live in the US, you’re almost certainly being tracked by political organizations’. Why wouldn’t that tracking include the harvesting of some kind of health data or pain medication preferences (e.g. via the use of cookies, commonly-used tracking technology, harvested data, third-party app data sharing)? And what would be the point of tracking & collating it if they didn’t then use it?
As it turns out, the sale of anonymous medical data represents an extremely lucrative trade (NB. in the US, medical data has to be anonymized before it can be legally sold). Whilst anonymization of patient data at point-of-sale would prevent companies from using that data to market products & services to individual consumers, there is now a whole data-matching service industry willing to quickly and efficient rematch it. That significantly increases its usability. Little wonder, then, that global revenues for the big data business analytics sector are, in general, explosive, predicted to grow from an already impressive $122bn in 2015 to more than $187bn in 2019.
Data brokers have, in recent years, turned the sale of medical data into big business (e.g. IMS Health, a major player, was recently valued at around $9bn). It appears, in sum, to be relatively easy to de-anonymize medical data and match it to voter records (Tanner, 2016).
So, what is the likelihood that political campaigners have actually used this kind of medical data in their micro-targeting strategy?
Looking at the stats, on any given day before in the 2016 election cycle, the Trump campaign was running around 40,000–50,000 variants of its ads — which is a hell of a lot of micro-targeting. A record 175,000 variants of one Trump campaign ad ran in just one day (the day of the 3rd presidential debate (Lapowsky, 2017)), demonstrating the huge appetite of political agencies for personalized data.
Health data now constitutes one of the most highly-sought after data sets out there & its sale & purchase aren’t limited to medical companies; social media have been trying for years to get in on the act. Google, for example, recently embarked on Project Nightingale — the transfer of 50 million Americans medical records from across 21 American states from 2, 600 hospitals to Google via an agreement with health care company Ascension. Following the deal, whistle blowers spoke of the fact that many Google employees had been granted full-access to patient data, without the patient or doctor having given consent first. Google remains under scrutiny; DeepMind Health, its healthcare subsidiary, has already faced criticism for its engagement with (UK-based) NHS patient data. In 2018, Facebook held talks with numerous medical groups and hospital trusts to try to acquire anonymized patient data. The company planned to re-match these accounts to live Facebook profiles via the use of a cryptographic technique known as ‘hashing’. The project has currently been shelved, at least for now, amidst concerns over privacy (the company continues to be a leader in political ad revenue, earning $2.2bn in 2018 in political ad spend alone).
Allied Market Research recently valued the global pain management drugs market at $71,43 billion in in 2019. It is projected to reach $91,64 billion by 2027. Centers for Disease Control and Prevention (CDC) reported that in 2018, approximately 50 million US adults were affected by chronic pain, with 20 million of those experiencing high-impact chronic pain. The use of pain medication is for most Americans an intuitive choice even if they don’t fall in to a chronic pain group; who doesn’t have some painkillers in their cabinet?
So where does this story end? Recently, Purdue Pharma, the makers of OxyContin, reached an $8.3bn settlement, agreeing to plead guilty to criminal charges concerning its role in the supplying of drugs without legitimate medical purpose-a supply that critics say has fueled a damaging, seemingly unstoppable drug epidemic that has ruined families, communities and local economies. That’s an ending of sorts — and hopefully some kind of relief — for Painesville and countless other small towns like it in the US.
But the real conclusion of the story to me is this: if you are still shaking your head and feeling bewildered, asking yourself how 72 million Americans could, in 2020, have experienced some kind of collective empathy bypass that allowed them to happily cast their vote for a clearly divisive, racist, misogynistic candidate, then…maybe they have. In the battle for the soul of a nation, surprisingly, your bathroom cabinet might be the first place that you could now choose to look for a little extra explanatory power. It certainly doesn’t explain the whole story, but it might have moved the needle, just a little, in affecting political preferences. And in today’s politics-a true game of inches-sometimes that’s all that it takes.
For more info:
