Read on Twitter
How "risky" does someone need to be to justify pretrial detention under current law? @sandy_mayson and I lay out the legal & empirical framework to answer this. We show that jail is so harmful that virtually no one is "dangerous" enough to warrant it. 1/
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3787018
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3787018
Each year millions of people are detained pretrial, many ostensibly due to crime risk. But existing law and scholarship provides no answers to the level of crime risk that could legally justify pretrial detention. 2/
If locking up 10 people for 30 days each is expected to avert a serious crime, is it "worth it"? Despite increasingly using quantified estimates of crime-risk - risk assessment tools - in determining pretrial freedom, there are no serious attempts to answer that question. 3/
We aim to fill the gap by 1) developing a conceptual framework in law and 2) implementing a novel empirical technique to estimate the minimum risk threshold at which purely preventive detention could be justified. 4/
Governing law basically says that in order to be constitutional, pretrial detention must pass a cost-benefit test. It must avert more harm (crime) than it generates (lost liberty). 5/
(Of course flight risk is an issue as well, but a distinctly secondary one in practice. Our paper focuses on the question of when detention for the purposes of preventing crime is justified.) 6/
Furthermore, prohibitions against pretrial punishment means that we must take the well-being of the arrestee fully into account. We can't discount the harms they experience due to mere allegations of crime, past or future. 7/
So: how to weigh the relative harms of crime and incarceration? The literature provides lots of estimates of the cost of crime, but no workable ones of the cost of incarceration.* 8/
We implement a novel method of estimating the relative costs of incarceration and crime that we call "Rawlsian cost-benefit analysis". Formally, it's a type of contingent valuation, but most would recognize it as a game of "Would you rather?" 9/
In it, we ask respondents questions like "If you had to choose between being the victim of a robbery and spending a month in jail, which would you choose?" 10/
We find that respondents across the board are simply terrified of jail. Most would rather be the victim of a burglary than spend even a day in jail; rather be the victim of a robbery than spend three days in jail; rather suffer a serious assault than spend 30 days in jail. 11/
Incarceration is simply horrible, according to our respondents. As such, a person must pose an extraordinarily high risk of serious crime in order to justify it. In order to justify detaining someone for one day you must expect to avert crimes as grave as burglary. 12/
No existing risk assessment tool is accurate enough to identify those who pose such a high risk. In one leading tool, defendants in the highest risk category only pose a 2.5% chance of rearrest for a violent offense within a month. 13/
So here's a puzzle: current legal doctrine suggests that pretrial detention on the basis of dangerousness should be extraordinarily rate. But it's not. There are likely millions of peope detained each year because they are perceived of as too "dangerous" to release. 14/
Bail courts appear to be discounting the wellbeing of arrestees relative to potential crime victims when determining when to set affordable bail. This is probably partly because arrestees are disproportionately Black, Brown and poor. 15/
This is also probably because bail magistrates hold an arrestee culpable for their alleged crime or for the risk they pose. They may be legally presumed-innocent, but practically presumed-guilty. 16/
Our current rates of pretrial detention look an awful lot like widespread pretrial punishment. What's the path forward? 17/
At a very least, current law and policy needs to reckon with the fact that the consequentialist principles used to justify preventive detention don't remotely track onto current practice. 18/
Our study also raises tricky questions about how law *should* proceed when the human impulse varies so widely from its legal justification. 19/
It's clearly morally wrong and legally impermissible to discount someone's wellbeing based on race or socioeconomic status. 20/
But we acknowledge that there is a very strong human impulse to discount the wellbeing of people who are perceived to have done wrongful acts in the recent past, relative to an innocent potential crime victim. 21/
There is also a small amount of legal precedent for this (Scott v. Harris). We don't resolve this issue, but rather leave it as one of the many interesting and thorny issues that are exposed when consequentialist justifications are carefully probed. 22/22
* @davidsabrams and Chris Rohlfs have a cool paper that estimates the value of freedom. But, as the authors acknowledge, the estimates are capturing ability-to-pay of an indigent and credit-constrained group, and are likely biased downward.
