Read on Twitter
I’m going to geek out on this thread more than normal b/c new Science paper is WTF bananas awesome, and I don’t want it to get lost in the news cycle. In criminal justice policy, there’s a bad habit of solving problems by increasing punishment. https://science.sciencemag.org/content/early/2020/10/07/science.abb6591">https://science.sciencemag.org/content/e... 1/
And so if people aren’t showing up to their summons hearings, increase the fines, or put out an arrest warrant. The logic is that people don’t want to comply so force them to. Authors had a different idea. Maybe the authorities are just poorly communicating critical info 2/
Working w/NYC, authors redesigned summons ticket. Old summons (left) focused on describing you and what you did. New summons (right) focuses on court location/date, states in bold on front an arrest warrant will be issued if you miss. (note my screenshot resolution not great) 3/
The redesign reduces failures to appear by 6.2 percentage points, or by 13.2% relative to the 47% baseline. In regression discontinuity graph, you can see the drop-off following new form intro. Robustness checks for seasonality, etc in SM 4/
Next, the authors added simple text message reminders to the redesign for defendants who opted to provide phone# a week, a few days, and one day before date. Messages vary a little, but on average reduce failures to appear by an additional 8 percentage points 5/
Authors estimate interventions prevented 30K arrest warrants over 3 years! Of these, 2/3 cases dismissed; That’s 20K people whose lives went from arrest warrant to case dismissed! Interventions also cheap and saved city a ton of money in paperwork/labor. 6/
In lab experiments, authors confirm that participants are faster to identify relevant info and have better recall w/new form. Thus, small changes in communication are far more effective and more humane. So why do punitive solutions persist in criminal justice policy? 7/
In a series of folk psych studies, study participants believe failure to comply in criminal domain is more intentional and less accidental than other domains (e.g. not making a doctor’s appointment). Also less supportive of nudges in criminal domain. 8/
These beliefs are malleable. When asked to think of a reason why someone might miss court on purpose or by accident, support for nudges goes up in the accident condition. Note that the intentional condition is same as control, suggesting it’s closer to default. 9/
When experts (prosecutors, judges, defense lawyers) are shown the old and new forms they clearly think redesign will improve recall and court appearance. Laypeople show no preference, suggesting criminal justice policy is closer to lay intuitions than expert experience. 10/
Nudges take a lot of heat for implicitly favoring small touch-ups at the risk of avoiding structural change. Authors are aware of this. In a final exploratory analysis, they look at SES/race effects. Big takeaway- Poor minorities get way more summons tickets 11/
This means that while redesign/text reminders aren’t any more effective across race/SES, they could help poor minorities the most simply b/c their baserate of being issues them by police is highest. But while nudges can help, they cannot solve this core inequality. 12/
In an awesome perspective piece, Issa Kohler-Hausmann suggests this vast difference in who is getting the summons tickets is at the root of why ineffective punishment persists over effective policies. Absolutely necessary reading in my view 13/
Hausmann points out that there’s a lot of grey between intentionally avoiding court and completely forgetting about it. And yet, our folk psychology for poor minorities is that they don’t get the same leeway for excuses as those not constantly under state surveillance. 14/
Summons tickets are another call by the state, where the whole purpose is to show up, jump through hoops in order to signal deference to white supremacy (my words). Its not about intention, its about who the law sees as worthy of full citizenship. 15/
In reflecting on Hausmann, I wonder about effect persistence. Normally, a critique of nudge work is that people will adapt. Putting salads at the front of the food bar works at first, but then people learn to start in the middle. Here, the problem is a bit different 16/
If the purpose of the system is not to facilitate law but rather to make life hard on poor minorities, then if nudges are created to make life easier on them, how does the system react? Does the system find a new avenue to implement ‘sludge’ beuracracy on poor minorities? 17/
Much to chew on. For now, we talk a lot these days about basic theory vs. applied policy work in social science. This paper is an exemplar of how you can achieve both. Congrats to authors @alissafishbane @AurelieOuss Anuj Shah. Even the funders have my kudos this time :)
p.s. noticed I forgot the link to Kohler-Hausmann https://science.sciencemag.org/content/early/2020/10/07/science.abc2495/tab-article-info">https://science.sciencemag.org/content/e...
