Performing Foundational Research

Performing Foundational Research

Across the criminal justice field, there is widespread agreement about the need for more rigorous research to answer foundational questions about the way the system currently functions and how it can be improved. In the pretrial phase specifically, decision makers often lack basic information about who is in their jails; why those individuals are being detained; and the ways in which being held before trial affect the defendant, various case outcomes, the system itself, and the community as a whole.

In order to gain a better understanding of the impacts of pretrial decision making and provide officials with information that can inform their criminal justice policies, LJAF funds research and monitors the existing literature on key pretrial issues. While additional research is still needed, the evidence to date shows that the initial decision about whether to release or detain a defendant can have a significant impact on what ultimately happens in his or her case.

Several recent independent studies examined a variety of factors, including the impact of pretrial incarceration on convictions, sentencing, and other outcomes such as future employment and recidivism. Here is a brief synopsis of each project:

  • A study conducted last year by Megan Stevenson looked at all 331,615 criminal cases in Philadelphia between 2006 and 2013. Because the magistrate judges who determined bail in these cases worked on a rotating schedule and were significantly different in their propensity to detain defendants prior to trial, it was effectively as if some defendants were randomized to be detained more often than others. Stevenson found that defendants who are detained before trial are 13 percent more likely to be convicted than those who are released. This is due to the fact that people who are held in jail are more likely to plead guilty—regardless of whether they actually are—in order to get out. In addition, Stevenson found that individuals who are detained before trial receive sentences that are about five months longer than those who are not held.

And in cases where the evidence of a person’s guilt is weaker, the negative impacts of pretrial incarceration are even more pronounced—in these cases, defendants are 30 percent more likely to plead guilty and receive sentences that are roughly 18 months longer. Stevenson suggests that detention may place “undue coercive pressure” on the defendant to take a guilty plea rather than risk trial. Additional possibilities are that detained defendants are less able to gather evidence or prepare their case, and moreover are unable to demonstrate to the judge that they can stay out of trouble.

  • Another well-conducted study released in 2016 reported similar findings. Will Dobbie, Jacob Goldin, and Crystal Yang looked at 328,492 cases from Philadelphia and 97,538 cases from Miami-Dade County. In both places, judges worked on a rotating schedule, making bail determinations effectively random. The researchers found that defendants who are released before trial are 27 percent less likely to be convicted than those who are detained. Again, the difference can largely be attributed to the fact that defendants who are released pretrial do not plead guilty as often as those who remain behind bars.

Dobbie, Goldin, and Yang also studied other important outcomes. They found that pretrial release or detention does not affect whether a person will commit another crime up to four years in the future. However, when a person is released before trial, he or she is 14 percent more likely to file a tax return three or four years later—and, therefore, to be employed—than a defendant who was held. Pretrial release also increases a person’s unemployment insurance benefits by more than 100 percent and his or her Earned Income Tax Credit benefits by 66 percent. In addition, the data suggest that pretrial release leads to a 27 percent improvement in employment over the next three to four years. These indicators demonstrate that when a defendant is detained before trial rather than released, he or she tends to get separated from the more formal job market and associated government programs.

  • In a forthcoming article, Paul Heaton, Sandra Mayson, and Megan Stevenson use detailed data on 380,689 misdemeanor cases filed between 2008 and 2013 in Harris County, Texas, to measure the effects of pretrial detention on case outcomes and future crime. With this dataset, they were able to control for “uniquely detailed information,” including “defendant demographics, extensive criminal history variables, wealth measures (ZIP code and claims of indigence), judge effects, and 121 different categories of charged offense.” The results show that defendants who are detained before trial are 25 percent more likely to plead guilty and 43 percent more likely to be sentenced to jail than defendants who are released. The detained defendants’ jail sentences are more than twice as long on average. In addition, those who are held are 20 to 30 percent more likely to commit new crimes during the 18 months following the initial bail hearing. This result may be because detention itself can result in the loss of jobs or housing, thus leaving defendants with fewer options for survival.
  • A study conducted by Emily Leslie and Nolan Pope looked at nearly a million cases brought before 212 judges in New York City between 2009 and 2013. The researchers noted that the judges varied in how likely they were to set bail, and that which judge a defendant saw was close to random, depending, for the most part, on the time of day the person appeared before the court. Leslie and Pope found that “pretrial detention increases the probability that a felony defendant will be convicted by at least 13 percentage points; effects for misdemeanor defendants are larger than seven percentage points.” The researchers observed that this effect arises because defendants who are detained are more likely to accept a plea deal. Study results showed that felony defendants who were detained received sentences that were more than 150 days longer, and that these defendants were substantially less likely to have their charges reduced.
  • And yet another rigorous study, conducted by Arpit Gupta, Christopher Hansman, and Ethan Frenchman, looked at 203,188 cases between 2010 and 2015 in Philadelphia and Pittsburgh. It found that assigning money bail results in a 12 percent increase in convictions and a 6 to 9 percent increase in recidivism. This result is due to cases where defendants cannot make bail and are therefore detained. Although Pittsburgh does not assign defendants to judges in a way that resembles randomization, the empirical effect on conviction rates there was the same as in Philadelphia, which does use a randomized process.

The way that pretrial incarceration impacts future criminal behavior remains an open question. As mentioned, Dobbie et al. found that pretrial release or detention has no significant effect on whether a person will commit a crime up to four years in the future, while both Gupta et al. and Heaton et al. concluded that pretrial detainees are more likely to commit future crimes.

Given that much is still unknown about the effects of pretrial incarceration, our team will continue to support and monitor research in this and other areas of the criminal justice system. High-quality research is essential in order to develop effective reforms that will ensure that the system operates in a fair, effective, cost-efficient manner.