Read on Twitter
The way we think about trends is broken and it's affecting our reaction to COVID-19. We can learn from the research on prison history and mass incarceration. THREAD!
Point 1: When reviewing COVID-19 data, we put too much emphasis on national-level, aggregate trends (how many deaths in the US v broken down by state). But national-level trends can be driven by one or two states (like New York in this case). /2
When the national incarceration rate went down in 2011, this was big news. But the decline was mostly due to cuts to California's prison population. Many other states *increased* their prison populations. /3
https://www.tandfonline.com/doi/abs/10.1080/15564886.2015.1078185?journalCode=uvao20
https://www.tandfonline.com/doi/abs/10.1080/15564886.2015.1078185?journalCode=uvao20
National-level data also leaves out counter-trends and pockets of slow, gradual change. (FT map below.) /4
While the rest of the country was celebrating prisoner rehabilitation, building "correctional institutions," and changing the name of guards to "correctional officers, the South still relied on plantation-style prisons (like Cummins Farm below) that replicated slavery. /5
By the time national-level trends seem like they are wrapping up, these lagging states (or areas of states, like rural areas) can be just getting going on the trend. /6 https://twitter.com/DrRobDavidson/status/1249679726538625026
In the 1970s, Arizona hopped on the rehabilitative band wagon. This was around the time the California--which led the charge in favor of rehabilitation--was rejecting rehabilitation in favor of punishment. /7 https://www.sup.org/books/title/?id=17521
When these lagging states get going on the seemingly passé trend, they can be facing different conditions. Federal funding for testing is ending now but some states that didn't see a lot of cases initially (and some pockets like prisons) are seeing accelerating numbers of cases.8
This gets me to Point 2: We put history into boxes. In thinking about COVID-19, we are quick to do the same (e.g., talk of the "peak" or when this is all going to end and get back to normal). /9
Here's an example from prison history. When talking about the rise of new types of prisons, we put them into periods, e.g., 1790-1820 or 1890-1920. But the rise and spread of these prisons usually begins before and ends after these periods. /10
This is important bc late adopters (or late-emerging COVID-19 hotspots) face different conditions than innovators and early adopters (the places that saw a lot of cases early on). /11
In the case of prison adoption, different factors drove adoption in the 1810s/20s v. 1840s/50s. The former were driven more for ideological/penological reasons. They had a real need (what they saw as need) for prisons. /12 https://onlinelibrary.wiley.com/doi/abs/10.1111/lasr.12136
The latter were not. They were states that didn't exactly need a particular model of prison, or even a prison generally. They adopted these prisons to look legitimate. (This is not unique to prison adoption, btw. See link below for ex.) /13
https://www.jstor.org/stable/4623368?seq=1
https://www.jstor.org/stable/4623368?seq=1
Again, late-stage COVID-19 hotspots will be facing different conditions than early stage. Perhaps more prepared, more available data, know what to expect. But also already drained, fewer resources, etc. Plus, pre-existing differences like possibly less infrastructure. /14
Perhaps most importantly, though, our tendency to box history means people will be too quick to say this is "over." They'll do this when it's over in NYC and Seattle, but it won't be over for Mississippi or Iowa or prisons and other places flying under the radar right now. /15
This is why I'm so worried about all this talk of hitting the "peak." The data seem to be driven mostly by NYC (like California in our national prison population ex). But other places haven't peaked yet. /16
And the big consequence is if we loosen up social distancing too early and "reopen" the country bc the places getting the attention now are getting better--and the places that are several weeks behind don't have many visible cases yet. /17
