New study highlights the damages of jailing people before they’ve ever been convicted of a crime.
By Matt Keyser
National Partnership for Pretrial Justice
March 24, 2022
It was 2013 and Christopher Lowenkamp had just published a report on the hidden costs of pretrial detention in Kentucky.
Armed with data collected by the state, Lowenkamp’s original study found that people who were detained three days or more saw their failure rates sharply increase when compared to those who spent less time in pretrial detention.
But he wanted to dig deeper with more granular data — measuring detention in terms of hours, not days, for example — and use advanced statistical models to show pretrial detention’s harms.
Around 2018, he received that chance with support from Arnold Ventures which allowed him to investigate further.
By looking at nearly 1.5 million cases of people booked into Kentucky jails between 2009-2018, the effects of pretrial detention became all too clear.
“The longer you spend in pretrial detention,” Lowenkamp says, “the poorer outcomes you have.”
We spoke with Lowenkamp about his findings and in what ways they can be used to create systemic change.
This interview has been edited for length and clarity.
NPPJ: Chris, I want to get into the finer points of your report, but there’s one of your findings that struck me immediately. Pretrial detention, regardless of time spent detained, is damaging to a person. Why is that?
Christopher Lowenkamp: There are a couple of things to consider. Pretrial detention seems to really increase the likelihood someone gets rearrested once they’re released. So being locked up for even a short period of time seems to really impact whether you get rearrested.
To me, though, what’s more important from all that is the fact that people who are detained pretrial and who are detained for longer periods of time end up having a greater likelihood of being sentenced to incarceration — whether that’s jail or prison — and they are sentenced to longer periods of incarceration compared to people who were released pretrial.
This is consistent with what we found in 2013. It’s consistent with what other researchers and other data have found. But to the extent that we’re replicating this with different data from different times and different places starts to point you in one direction that shows pretrial detention can be problematic and should be used very judiciously.
NPPJ: Coming back to that point, you first began studying Kentucky’s pretrial system around 2012 and even released a report back then. Why did you decide to revisit the data?
CL: All projects have parameters. When we finished that research, we weren’t completely satisfied with the analysis that we did. We wanted to do more, but we didn’t have time or funding to do more. And then about three or four years later, we started talking to Arnold Ventures about replicating that study with more detailed and newer data.
NPPJ: Why Kentucky specifically?
CL: Part of it was because they were fairly progressive at the time [in 2012] in terms of risk assessment. They’re a state system. And they had available data to start looking at some of these questions and getting answers at pretrial. Pretrial research is fairly new. It doesn’t have 100 years of research at this point. So to have that data is really important.
NPPJ: What were some of those things that you weren’t satisfied with within the original report that you wanted to dig deeper into?
CL: One is more advanced statistical models. We wanted to have greater precision and measuring how long someone actually spent in pretrial detention. In the original data all we had was the date that they were booked and the date they were released.
In the more recent data, we have the date and time that they were booked, as well as the date and time that they were released. So we can now calculate hours in detention rather than dates.
NPPJ: Why was it important for you to have that more granular data?
CL: I think what was important about it is one of the things that came out of the earlier report was what became known as the “three-day rule,” which found that failure rates increased steeply when people were detained three days or more. I think that finding in itself was an artifact of how the data we had at the time, meaning when you look at it hour by hour, there doesn’t seem to be this magical break point in three days. In fact, there doesn’t seem to be a magical break point anywhere.
What we saw with the more granular data is a gradual increase of failure rates almost every hour someone was locked up, and it begins from day one of someone being detained. If you’re a defense attorney, you would want to get your client out as soon as possible and not wait around thinking you have three days before any damages begin.
Again, the longer you spend in pretrial detention, the poorer the outcomes you have.
NPPJ: Given there’s now a growing body of research showing that pretrial detention is harmful from the moment a person is detained, what’s it going to take to create change in the system?
CL: I wish I had an answer. I think it’s going to take some education. I think having actors in the system look at their data and evaluate how quickly people are getting out and what are the barriers to getting them out quickly. Leaders in other jurisdictions should look at their data and see what’s really happening, because using their own data is always more convincing that looking at someone else’s.