Bail tech?

It seems that every week, a new viral video from HBO’s Last Week Tonight with John Oliver makes its way into the (inter)national discussion about very serious issues. And it appears that the show’s airing its treatment even has some immediate real-world impact – the “John Oliver effect” (TIME).

The list of topics Last Week has addressed this season include three that should be under consideration as important parts of criminal justice reform (CJR), which has been the subject of our TechTables with Sen. Cory Booker, and is one of our current focuses: judicial elections (February 22, 2015), bail (June 7, 2015), and mandatory minimum sentences (July 26, 2015). Of course, those aren’t the only three things that constitute CJR, and are not equivalent in terms of likelihood of being addressed (there seems to be a lot of momentum on the minimum sentences front, not so much on judicial elections); however, I would like to focus on the area where I see tech playing a major role: bail.

The system in the United States—with bail bondsmen and bounty hunters playing an important role—is almost unique in the world, with only the Philippines having a similar one. Said John Goldkamp, a professor of criminal justice at Temple University, of the commercial bail bond system, “It’s a very American invention… It’s really the only place in the criminal justice system where a liberty decision is governed by a profit-making businessman who will or will not take your business” (NY Times). However, the system remains in place, mostly due to political influence of bail bonds agencies and their underwriters – insurance companies – through lobbying and campaign contributions (Justice Policy Institute).

Now, why is the current bail system bad? There are four main reasons:

1. It disproportionately affects low-income communities. If you and/or your cosigners are unable to pay your full bail, you have two options: stay in jail while awaiting trial—which brings with it huge opportunity costs – or obtain a bail bond. In the latter case, the fee charged by the bondsman – typically 10% of bail amount – will not be refunded even if you show up to your court date. Furthermore, a bail bondsman can refuse to bail out anyone, for any reason.

2. It keeps people incarcerated when they needn’t be, crowding jails. Data show that approximately 25% more people could be released pretrial without increasing offenses or failures to appear (FTA) (Texas Law Review).

3. It does not necessarily increase, and may in fact decrease, community safety. Dangerous defendants may be able to raise a full bail amount, regardless of a judge’s attempt to make it high enough to prevent their release. Also, a bail bondsman might take a risk on a very high bail (which would indicate the judge thinks the defendant is high-risk), because the 10% cut would be a major windfall.

4. It has adverse effects on those charged with offenses (who have not yet been convicted of anything!). For instance, an ability to pay bail might lead a defendant to plead guilty to a crime he or she did not commit, just to get out of jail. Not only will that lead to problems relating to that person having a criminal history, but it will also leave the actual guilty person unaccounted for. Also, high false conviction rates will skew research that is used to predict who is at risk for committing future crimes or harming the community. Instead, researchers will merely figure out who is most likely to plead guilty, regardless of actual innocence.

So, what can tech do about it? First, the literature shows that good risk assessment drastically reduces the rate of pretrial misconduct (FTA or harm to community). However, many jurisdictions have not adopted risk-assessment tools because they don’t believe it will work, or it would be cost-prohibitive. To me, this seems like an area ripe for a tech innovation built upon big data analytics. There is more than 30 years of research in this area, and a lot of data about pretrial misconduct…

Second, modeling software could be used/developed to allow jurisdictions to test changes to their pretrial system before implementing them, increasing the chance that they will make the decision to reform.

Third, the actual implementation of a pretrial supervision plan, including tracking of defendants, checkups, and reminder systems, could be implemented using software, automating the process, and reducing FTA.

Let me know if you have any ideas or knowledge of work being done in this area. Thank you!

Note: Facts without explicit citations are from: Neal, Melissa. Bail Fail: Why the U.S. Should End the Practice of Using Money for Bail. Justice Policy Institute. September 11, 2012.