You may have seen this chart on social media recently — after Sean McElwee’s tweet about it went viral:
Charts like this have been fairly common in criminal justice writing over the past few years, with media outlets like the Economist, Mother Jones and, yes, Vox referencing something like it in previous articles.
But as criminal justice expert John Pfaff points out, depending on how you read this chart, it can be very misleading:
1. This graph is really misleading. It looks like we’ve simply moved the mentally ill out of hospitals into prisons. It’s NOT saying that. https://t.co/ychg5i2dcI— John Pfaff (@JohnFPfaff) October 18, 2016
2A. “Prison” line ISN’T “mentally ill in prison,” which title suggests. It’s JUST total prison pop, mentally ill or no. I mean, seriously?— John Pfaff (@JohnFPfaff) October 18, 2016
2B. The best work I’ve seen on deinstitutionalization and prisons is this, which say effect is about 4% to 7% of prison growth from deinst. pic.twitter.com/cz4RyfLYcZ— John Pfaff (@JohnFPfaff) October 18, 2016
3. Also, hospital and prison pops were very different: hospital patients were older, whiter, and more female. Not just moving the same ppl.— John Pfaff (@JohnFPfaff) October 18, 2016
4. Graph seems to imply we just emptied the hospitals into the prisons. What really happened is far more complex, harder to disentangle.— John Pfaff (@JohnFPfaff) October 18, 2016
5A. Obviously attitudes about treating various populations changed during the 1970s; NOT saying lines aren’t causally linked.— John Pfaff (@JohnFPfaff) October 18, 2016
5B. But the link is far more complex than what that simple one-line-down-one-line-up graph suggests.— John Pfaff (@JohnFPfaff) October 18, 2016
6A. Sadly, last rigorous analysis of mental health problems among inmates was conducted in 2006—a DECADE ago: https://t.co/ARXNDIVR8u— John Pfaff (@JohnFPfaff) October 18, 2016
6B. FWIW, here are its core findings. pic.twitter.com/GxYaeJMpPp— John Pfaff (@JohnFPfaff) October 18, 2016
7A. Finally, a stats challenge w prison-vs-civilian pop mental health comparisons: all ppl screened for mental illness when entering prison.— John Pfaff (@JohnFPfaff) October 18, 2016
7B. Imagine this will lead us to overestimate the frequency of mental health problems in prison vs. outside: census vs. imperfect sample.— John Pfaff (@JohnFPfaff) October 18, 2016
7C. That’s not saying “it’s not such a big deal.” If anything, it likely signals how much undiagnosed mental illness exists outside prison.— John Pfaff (@JohnFPfaff) October 18, 2016
8A. And to be clear, think we seriously mistreat mentally ill ppl convicted of crimes and overlook mental health harms of prison. pic.twitter.com/yfoVdBjl4T— John Pfaff (@JohnFPfaff) October 18, 2016
8B. But that doesn’t change the fact that the Economist graph tells far too simplistic a story about prisons and mental health.— John Pfaff (@JohnFPfaff) October 18, 2016
In short, the chart isn’t totally wrong, but it oversimplifies a rather complicated story: While deinstitutionalization did contribute to mass incarceration, it wasn’t the whole cause or even a big one.
Beyond what Pfaff noted, there are a couple of other problems with the chart.
For one, it suggests that people with mental illnesses are violent and have to be locked up somewhere. That’s not true: People with mental illness are more likely to be victims — not perpetrators — of violence. And only about 3 to 5 percent of violent acts in the US are carried out by people with serious mental illnesses, while about 4.2 percent of adults in the US experience a serious mental illness in a given year that substantially interferes with or limits their major life activities.
The chart can also be read to suggest that the deinstitutionalization of mental health care was a bad thing. While state-run mental hospitals are still underfunded, the asylums of old were, experts now widely agree, horrifically abusive. There’s a very good reason why popular depictions of these old hospitals tend to be all about abuse (like in One Flew Over the Cuckoo's Nest) or literal horror movies and shows (like in American Horror Story). The move to shut down these places don was a good thing — even if the better forms of treatment developed since then, such as community clinics, are frequently underfunded and unaffordable.
Still, it’s also true that the criminal justice system often mistreats people with mental illnesses. A 2015 report by the Treatment Advocacy Center found that someone with an untreated mental illness is 16 times more likely to be killed by police than other civilians approached or stopped by law enforcement. And 2 million people with a mental illness cycle through US jails each year, while at least 83 percent of jail inmates with a mental illness do not have access to adequate treatment, according to the National Alliance on Mental Illness.
But those problems can be signaled and studied without simplified graphs. Like many things in the policy world, the connection between mental illness and prison time is a bit more complicated than one chart can suggest. Getting that nuance right is crucial to understanding and fixing the problems surrounding mental health and the criminal justice system today.
For more on this topic, check out Vox’s in-depth explainer.