August 9, 2023
Transitional housing and recidivism
Logan Lee discusses the impact of residential housing programs on reincarceration rates.
The United States spends over a billion dollars a year on housing programs that give recently released prisoners a place to stay and modest support before reintegration into society. Yet there is little causal evidence that these programs work.
In a paper in the American Economic Journal: Applied Economics, author Logan M. Lee estimated whether residential housing programs in Iowa kept prisoners from returning to prison. He found that instead of reducing recidivism, prisoners assigned to halfway houses appeared to have higher rates of reincarceration than those who were paroled.
Lee recently spoke with Tyler Smith about how he arrived at his estimates and whether or not residential housing programs should be scaled back in the United States.
The edited highlights of that conversation are below, and the full interview can be heard using the podcast player.
Tyler Smith: Can you just walk me through how you became interested in this question about the impact of halfway homes on reincarceration and in particular why you thought it was important to get an answer to that question?
Logan Lee: Nationally, we spend more than one billion dollars a year on these residential housing programs or on halfway house programs. And there's really no causal evidence out there about whether they're effective. We just don't know. There's lots of reasons to think maybe these are a good idea. There's some evidence that suggests that a lack of housing is a huge reason why people come back to prison. On the other hand, you're also creating a situation where they have a lot more opportunities to fail. So there was a big need for some causal evidence on that.
Smith: What does a typical residential housing program look like? What kind of image should listeners have in their mind?
Lee: I think this is really important because when you hear “halfway house,” you think of a regular house. But this is more like an institution. And in that sense, residential housing is a better term. You can think more like dormitories. The average one in Iowa is housing 80 to 100 people, like a college dormitory, although a very high-security college dormitory. It's maybe more like incarceration, and also the fact that you're living with like four people in a dorm-sized room. This is not a comfortable house.
Smith: How did you go about trying to estimate the impact of these housing facilities?
Lee: When you walk into prison in Iowa, you're going to immediately be assigned a case manager, and they will work with you throughout your incarceration. They're going to tell you when your release date is, and they're going to prepare you for release. The most important thing, though, that these people do is determine whether you go to parole or whether you go to residential housing. The reason why that's so important in this context is that those individuals are not all the same as we might expect. When you're randomly assigned to a case manager, it's going to dramatically affect whether you eventually go to residential housing or not.
This is the kind of thing that economists really like because it gives us a clear opportunity to use the randomness that's already inherent in this system. We can compare two individuals that look almost identical on paper. One happened to be assigned to a parole-favoring case manager, and one happened to be assigned to a residential-housing-favoring case manager. We can see then how their outcomes differ. When you think about doing that with thousands and thousands of people, it should be that these groups have the same outcomes except for the impact of that residential housing placement or not.
Smith: What do you find when you compare these prisoners who were assigned different case managers?
Lee: The basic finding is that if you happen to be assigned to a case manager that was more likely to give you residential housing and you end up going to residential housing, you certainly are not recidivating less. And in fact, in the first six months after you're released, which would be the time that you're actually living in one of these residential housing facilities, you're more likely to return. You're more likely to recidivate and go back to prison.
You're more likely to go back to prison because of a technical violation. The idea is that you haven't committed a new crime, but you've done something that was bad enough to send you back to prison. So you see a big increase in those technical violations, but you also see a big statistically significant increase in violent crime. Finding that increase in violent crime associated with being assigned to residential housing is quite disconcerting and suggests that we're making this big investment and actually making things worse.
Nationally, we spend more than one billion dollars a year on these residential housing programs or on halfway house programs. And there's really no causal evidence out there about whether they're effective. We just don't know.
Smith: That seems like a really counterintuitive result. Did it surprise you?
Lee: The technical violations I definitely expected. If you give someone more opportunities to fail, then they’re more likely to fail. We do see small decreases in both drug and property crimes. And again, that kind of makes sense. If you're not out in the world, it's harder to commit a property crime. But the increase in violent crime was a big surprise. That's why I spend quite a bit of time in the paper trying to figure out what's going on there beyond just saying violent crime goes up.
Smith: Do you think these housing programs should be scaled back to make room for other approaches to reintegrating prisoners back into society?
Lee: Obviously, this is a really important question. I think that—looking more broadly at the literature—one of the takeaways that definitely dovetails nicely with my research is that lower levels of supervision seem to lead to better outcomes. In that sense, prison is the worst, residential housing is slightly better, parole is better still, but the more you can kind of move along that line toward low-supervision strategies, that seems to be preferable.
I'm not comfortable with my results saying that we should get rid of residential housing programs entirely. There are probably people for whom residential housing is important and for whom it's even worth that high cost. It's just that there aren't nearly as many of them as are being assigned in the current system. So I think that my results suggest scaling back, not getting rid of entirely, and just generally looking for opportunities to reduce the level of supervision where possible.
“Halfway Home? Residential Housing and Reincarceration” appears in the July 2023 issue of the American Economic Journal: Applied Economics. Music in the audio is by Podington Bear.