Podcast Icon

Pamela Oliver on What the Numbers Say About How to Reduce Imprisonment

  • Pamela Oliver
  • January 2021
  • PC90-2021

The goal of reducing incarceration has been gaining traction for at least the last decade in the United States. In an interview we did with sociologist Pamela Oliver in late 2020, she talked about how we got to where we are today when it comes to U.S. imprisonment and the impact that different reforms would have on reducing the U.S. prison population. This is part 1 of the interview. 

View Transcript - Part 1

Chancellor: Hello and thanks for joining us for the Poverty Research and Policy Podcast from the Institute for Research on Poverty at the University of Wisconsin–Madison. I’m Dave Chancellor. The goal of reducing incarceration has been gaining traction for at least the last decade in the U.S. And often, when we hear people talk about reducing levels of incarceration, the focus is on reducing the number of people in prison for drug offenses or nonviolent offenses. But many researchers have pointed out that while these approaches would help, their actual impact would be much more modest than we might hope for. In late 2020, I had the honor of talking to Professor Pamela Oliver who is a sociologist at the University of Wisconsin–Madison. We talked about a paper she wrote for the Marquette Law Review called “What the numbers say about how to reduce imprisonment.” We are splitting the interview into two episodes for you. This is part one and we’re going to hear about how Professor Oliver came to write this paper and how we got to where we are today when it comes to U.S. imprisonment. Next week for part two, we’ll hear about the part of the paper where Oliver discusses approaches to reducing incarceration and how much of a difference those approaches would actually make if we wanted to lower imprisonment to levels that the U.S. had prior to mass incarceration. Let’s get to the interview.

Chancellor: Pam, thanks for being willing to be here today and to talk about this paper that you’ve written, but as a first question, can you introduce yourself and talk about the kind of research that you do.

Oliver: Well, I’m Pamela Oliver. I’m a newly retired professor of sociology although I’m still active doing research. My own research has actually had two lines, only one of which is relevant to this talk today. My main line of research is the study of protest and social movements and I’ve studied black protests and news coverage of protests and so forth. But about twenty years ago, I got interested in the issue of racial disparities in criminal justice, really because of the actions of community groups. And I’ve spent a lot of time doing statistical analyses of patterns of incarceration primarily in Wisconsin and around the country and sort of got involved in using data to inform policy and sort of advocate for reducing incarceration.

Chancellor: Today we’re talking about an article that you wrote for the Marquette Law Review that was actually published in June of 2020 called “What the Numbers Say about How to Reduce to Imprisonment.” Can you give us a background of how you got into writing this paper and what’s the backstory here of what we should know about how you went about this?

Oliver: So I got invited to this conference at Marquette actually called alternatives to incarceration for violent offenders. And at first I kept saying I didn’t want to go because I only studied incarceration, not it’s alternatives. But we went around and I attended. But the presentation I initially did is still in the paper, it was initially only part of the paper but it was about returns to prison for violent and nonviolent offenders who’d been released from prison. Then, when I was at the conference and there was this discussion of reducing incarceration of violent offenders, I got interested in—they were talking about people who were lifers and juveniles who were still in prison 30 years later and so forth, and I got interested in just the numbers. I knew I had the data where I could say, well, how would this kind of reform and this kind of reform affect the numbers of people in prison. When I got home and worked on revising the paper for the journal article. I just went ahead and ran all those numbers and so the paper has become kind of this long, involved review of a lot of a variety of different things where I just said, you know, what do the numbers show about what the possible impacts would be of different kinds of policies.

Chancellor: For the first part we should cover that first part of this interview, we should kind of cover that first part of the paper and just get some background on what the imprisonment situation in the United States actually looks like today. Sometimes we hear terms like mass incarceration or the prison boom. What’s the actual history on that? Can you give us a background?

Oliver: The United States had for a long time, had relatively stable incarceration rates and then after the 1960s, U.S. incarceration started going up and as far as I can tell, from the limited data we have available by race in the 1970s and early 80s, initially, incarceration was kind of going up for everybody. Then in the last 1980s with the war on drugs there was this huge spike in black incarceration that was primarily concentrated around the drug war. And then incarceration kept going up in the 1990s but what my data analysis of that period showed is incarceration patterns become much more heterogeneous. There was a huge amount of black incarceration for drugs, but then that started leveling in the 90s. Different kinds of policies kind of kicked in that increased the incarceration for violent crimes and they increased the incarceration of white people, increased the incarceration of rural people, rural white people. So that chugged along into the 2000s and then reform movements started happening in the 2000s. One of the early things that was especially a target of reform because the data had been around and showed just strong patterns of what seemed to be injustice, you started seeing steep declines in Black drug incarcerations. In a couple states it was even happening in the 1990s but it really started catching on in the 2000s. The racial patterns and the offense patterns of who was in prison started shifting a lot in the 2000s. And the reform movement started spreading so that there’s now conservative as well as liberal groups who were pushing for reduced incarceration. And as you got past 2010, more and more people were in favor of reducing incarceration. So you start to see some declines, but it’s still really different in different states and the racial patterns are different in different states. And overall, there’s been a decline, but a lot of that decline is actually just California. The pattern of what’s going on in incarceration is really variable between states and by offenses and all sorts of patterns.

Chancellor: I feel like as we talk about those patterns—you mentioned some of the policies and maybe incentives that might have led to some of these increases in incarceration. Can you tell us a little about that too so we can have that in the back of our heads as we start thinking about differences by state.

Oliver: Mass incarceration happened really not because there were more people committing crimes, but because the people who were accused of crimes spent more time in prison and this happened in a variety of ways. They were more likely to be convicted of felonies rather than misdemeanors for any given thing that they did. If it was a felony, they were more likely to be sentenced to prison than not sentenced to prison, if they were sentenced to prison, their sentences were longer. And they also stayed in prison for a higher proportion of their nominal sentence than they had in the past. Early release declined, parole declined. If they were released from prison the rate of going back to prison after being released from prison went up. All of these factors set into the rise in incarceration. In general, very early on, arrests played some role basically because there were more arrests for drug crimes in the drug war and sending people to prison, but after the 1990s, rates of going to prison were not related to arrests either. And the most recent, in the late 1990s, it was the amount of time served that was the biggest factor in how much time people were spending in prison in overall incarceration rate.

Chancellor: Earlier you talked about some of the reform efforts that we’ve been seeing, especially in the last ten years or so. Can we talk about the aim of reducing imprisonment, is that something we would want to do? And why do we put people in prison in the first place?

Oliver: There are so many different things. The very long historical answer from 150 years ago to why we put people in prison involved a shift from death penalty to prison. So initially, the use of prison was seen as an alternative to killing people. In its era it was seen as a more humane response. But more recently, the rise in incarceration involved a political shift in how people wanted to respond to crime. It was really a rise in just being more punitive. So, it’s important to keep in mind that incarceration rates were much much lower in the 1930s, the 1940s, the 1950s, the 1960s, than they became by the 1990s. And so when you ask, well, why do we put people in prison, it’s not an unvarying situation about why people go to prison and for how long they’re going to go to prison. One model which was the old model was kind of like a long time out really. People would go to prison for long enough, the theory was, that they would settle down, repent of what they’d done wrong, and when they came out again, be ready to avoid crime. As the incarceration rate boomed, one theory of imprisonment was lock ‘em up and throw away the key basically. The ideology was being spread that if somebody did something wrong you should just sort of lock them up forever and forget about them. There was also—it’s hard to even think about what the justification for the mass incarceration around drug offenses was, except that there’d been kind of what we would call a moral panic where in the late 1980s and early 1990s, the population had become persuaded that using or dealing drugs was just about the most terrible thing that you could do and somehow people should just be locked up forever for doing that. That’s one reason. Another reason is punitiveness. Some crimes when people commit a crime, people feel like they need to be punished somehow. You don’t care whether they’re rehabilitated or not, you just want them to be punished. Those are some of the arguments, there’s also, in the era of mass incarceration there was a big argument about incapacitation, just the idea that criminals really can not be reformed. If you’re bad, you’re bad, so the only thing to do is just to lock people up so that they’re stuck in prison. And I would say the incapacitation arguments became ideologically dominant especially in the 1990s and early 2000s as incarceration was just taking off.

Chancellor: What does the research say about the effectiveness of incarceration. Or when does it work and when doesn’t it? What should we know as a background?

Oliver: So that’s a highly contentious area. Many of the reformers who are against incarceration point out that the supporters of mass incarceration are rarely asked to justify how good incarceration is. There’s a lot of research on what are called the criminogenic effects of incarceration. One question, one is the incapacitation argument is ‘you’re not able to commit a crime while you’re locked up in prison.’ And the reformers will point out more correctly that you’re not able to commit a crime outside prison if you’re locked up in prison, there’s actually a lot of crime that is committed inside prison. But there’s a big debate about, lets say you do something wrong. What is the effect of sending you to prison rather than putting you on, say, probation for the same crime? And it’s hard to do the research on that because the groups of people who get probation versus get sent to prison are rarely the same people. But the research is very mixed. And the question would be ‘what is the likelihood of committing another crime if you’re not sent to prison, what is your likelihood of committing another crime if you are sent to prison after you get out?’ And early research, the results are mixed but quite a few studies said it was not very easy to predict who was going to commit a second crime from anything about the first crime. There were studies about what were called the criminogenic effects of going to prison. That is, things like inside prison you associate with other criminals. You also lose your chances at getting jobs and stable family life which might pull you out of a life of crime. So, there are those arguments. There’s also the old ‘nothing works’ discussions which argue that rehabilitation programs did not work. And interestingly enough, the very first nothing works article actually was saying therefore incarcerating people on the grounds that they’re going to be rehabilitated in prison won’t work so that’s not a basis for incarceration, but then later the nothing works argument, that rehabilitation doesn’t work, was actually flipped and used as an argument for incarcerating people on the grounds that rehabilitation was impossible. But the problem is that many people that commit crimes desist, they just stop committing crimes even without being punished or without being sent to prison. It’s a very mixed area. In terms of people with addiction issues and drug offenders, that’s the place where most of the research says that incarceration is not particularly helpful, that rehabilitative programs that focus on the addiction issues work better than incarceration. But, in general, the research in that area as I understand it is very mixed and somewhat contentious. But there’s certainly no clear evidence that sending people to prison makes them stop having a life of crime or the deterrence effect of going to prison is pretty low.

Chancellor: I’m really grateful to Professor Oliver for sharing this work with us. We’ll release the second part of the interview where she talks about actual approaches to reducing incarceration next week. And if you’re interested in this work, you should definitely check out her paper and others in Volume 103, Issue 3 of the Marquette Law Review published in June of 2020.

This podcast was supported as part of a grant from the U.S. Department of Health and Human Services, Office of the Assistant Secretary for Planning and Evaluation but its contents don’t necessarily represent the opinions or policies of that Office, any other agency of the Federal government, or the Institute for Research on Poverty. To catch new episodes of the Poverty Research and Policy Podcast, you can subscribe on Apple, Stitcher, or Google Play Podcasts. You can also find all of our past episodes on the Institute for Research on Poverty website. Our theme music for this episode is “Staring Straight” by Maarten De Boer. Thanks for listening.

View Transcript - Part 2

Chancellor: Hello and thanks for joining us for the Poverty Research and Policy Podcast from the Institute for Research on Poverty at the University of Wisconsin–Madison. I’m Dave Chancellor. This is the second of a two-part interview with University of Wisconsin–Madison Sociologist Pamela Oliver about what the numbers say about how to reduce imprisonment in the United States. In part one which we released last week, she talked about how she became interested in this research project and how we got to where we are today when it comes to U.S. incarceration. If you haven’t listened to that episode yet, you might want to check that one out first. For this part of the interview, Oliver discusses approaches to reducing incarceration and how much of a difference those approaches would actually make if we wanted to lower imprisonment to levels that the U.S. had prior to mass incarceration. So, let’s get to the interview.

Chancellor: I want to move on to the second part of your paper where you’re actually looking at the numbers behind what it would take to reduce incarceration in the U.S. And the benchmark number that you use in the paper is that to return to a 1970s level of incarceration, we would need to reduce incarceration by 75 percent. Could you talk to us about why you chose that endpoint or that particular goal?

Oliver: I actually picked that point because that’s the last point before mass incarceration started. 1970 was the end of the fairly stable steady state incarceration rates. I just said, well, mass incarceration started after that, so what would it take to get back to the pre mass incarceration level that would fit what had been going on in the 20th century before we got to 1970.

Chancellor: What are then some of the changes or reforms to the current practices that we have when it comes to incarceration that you look at in your paper and what sort of impact would we see if we changed the way those things were done?

Oliver: One big thing that people have been talking about is reducing drug offenses. There are still a significant fraction of people in prison who are actually drug offenders. That’s one big category of reform. There are many people who feel like the entire area of drug offenses should just be decriminalized and just not send people to prison for it, done. And that would be a substantial fraction of people going and coming from prison. My numbers indicate that about 15 percent of the people in prison at the snapshot I took were drug offenders and it’s more like a quarter of people who are going and coming from prison are drug offenders. The logic of the conference was that it’s not enough just to focus on drug offenders, that you also have to pay attention to violent offenders because the majority of people in prison are in a violent offender category. As I argue in the paper, that’s because the violent offenders spend more time in prison so they’re a higher percentage of the snapshot whereas violent offenders are only a little over a quarter of the people coming and going from prison. So, another thing I looked at was really long sentences because there were people talking about how—even in the past, even for murder, people would spend about 20 years in prison and then get out. In fact, it was known that the recidivism of murderers is really low. Most murderers who went to prison and then got out again tended not to do it again and that’s also true now. Most murderers stay in prison for life in the current regime, but the murderers who get out have pretty low recidivism rates. There’s a discussion about taking people who’ve been in prison a long time and giving them a chance to get out. I found that about 17 percent of the people who were in prison have been in prison at least 10 years, so that’s another thing you could consider is letting out the people who’ve already been in a long time. And the logic for that is twofold. One, almost all, not all, but almost people reduce their tendency toward crime the older they get. Crime is primarily concentrated among younger people and so if you wait until people are just older, they’re less likely to commit crimes when they get out so if they’ve already been in a long time, they’re probably less likely to commit crimes when they get out. So that’s one argument and then the other argument is just the fairness argument, that they’ve already done their time. There are some people who are in prison for really long sentences that the sentencing seems to have been really unfair, actually. But it seems to me when I looked at the data, there was some play about these people. There are people in prison for life sentences or such long sentences they might as well be life sentences who never killed anybody. But they’re a tiny proportion of the total population. This idea of looking at people who’ve been in a long time was another thing I looked at. But it was about 17 percent, again it could make a difference but not a huge difference. And then what I looked at was I realized from various data sources, how long are the people who are in prison, when are they expected to get out? And my best estimate, the variables that you have to use in the dataset for this are kind of cobbled together, but it seems likely that about 45 percent of the people in prison in 2016, which is the last date for which I had data, were projected be released within three years, so that’s nearly half. And 59 percent are projected to be released within 5 years. What that means is that not sending people into prison would also very rapidly reduce the prison population. As people get out, if you don’t replace them, that’s another way to reduce the prison population. That implies looking at the folks who are being sent to prison on these lesser offenses that have shorter sentences and maybe reconsidering and not sending them to prison at all so that’s another group. Then the other set of policies is to look at parole, letting people out earlier, shortening sentences, and also then blocking or cutting way back on returns to prison. It was a smaller number than I had expected that were involved in being returned to prison in these data. I think that’s perhaps a consequence of the policy changes that have happened in the last 10 years in some states. But in general I have found that about 60% of the people were in their first spell in prison. My best estimate is that about 10% were in prison because they’d been returned on what is called a technical violation with no new crime and about 30 percent were in prison where they’d been returned with a new crime in the record, and in the paper I spend a lot of time explaining how I actually looked not just at the categorization but at the actual sentences listed in the records to try to determine if there was some new crime in the record. That was another pattern to at, stopping the returns. Because about a third of the people, 30–40 percent of the people, who were in prison, it was not their first spell in prison, they’d been returned to prison.

Chancellor: One of the things I found really striking in your paper were the breakdowns that you did by state, especially when it came to sentencing or how long people stayed in prison and I’m wondering if you can give us a snapshot of that, because there are vast differences based on where someone committed a crime or where they happened to be. Can you give a picture of that?

Oliver: That’s much harder to do talking than with all of those graphs I put in the paper, but you are correct that what I have in the paper is a bunch of scatterplots that are basically just trying to convey the visual impression that states are really different. The states are really different in the mix of what sentences are in the prisons and they’re really different from each other in the proportion of people who are let out of prison who end up going back to prison. And they’re, in some cases, for most offenses, quite different in the median or average sentence length. The average sentence length range quite markedly for some offenses, so for example, sexual assault, the median sentence length ranges from under 5,—3 years to about 25 years. You have these huge variations in sentence length and these huge variations in the likelihood of going back to prison after you’ve been let out, as well as huge variations in just what the mix of offenses are in a given prison system. So, I think one of the lessons is that it’s really wrong to just look at the national patterns and draw conclusions about therefore this is what’s happening in each state. In general, the national patterns tend to be hiding the variations between states. And you really need, I think, information within a state to find out what is going on in your particular state and which reforms are most likely to reduce the prison population in your state if that’s your goal.

Chancellor: We talked about variation in sentencing, but what about variation in crime?

Oliver: Well, in terms of this paper, the other thing that’s in here that’s potentially interesting is the section that just talks people through the different kinds of crimes. I do feel like you need to talk back to the homogeneity of crime. A lot of people, if you say “crime” they just think murder and only 13 percent of people in prison nationally are murderers. And only 25 percent of people in prison are violent offenders or murderers. And murderers are a tiny fraction of the people entering prison in any given year. Even those people who are charged with murder for example, there’s even some people charged with murder who didn’t actually kill anybody. They were charged under what are called felony murder statutes, charged with murder, sometimes first degree murder if they were involved in a felony where somebody else killed somebody. And then I work through the different offenses like, in particular I make that point that assault is a highly variable category where some assaults are just brutal one-sided attacks, where other people accused of assault have actually been involved in what is essentially a mutual fight. There’s a lot of male subcultures in our society in which getting into a fight is just something men do. Similarly, the sexual assault category ranges from forcible rape, often brutal forcible rape, but also includes the criminalization of some categories of consensual sex, in particular between teenagers who are not particularly different in age and actually even in some states, two juveniles who have sex with each other can both be charged with having sex with a minor. And I try to get people to think about the fact that not all crime is the same and that when you’re trying to think about alternatives to incarceration, there would be some crimes for which a reasonable response is decriminalizing. There are other situations where restorative approaches are clearly preferable to a punitive approach that locks somebody up. There are other situations where it’s clear where the person has serious treatment needs or some sort of therapeutic response would be better. And then there are situations where the person did something bad and egregious where many people feel like some sort of retributive response is needed just to make people feel that justice has been served. But I do feel that one thing that people should do is just unpack their idea of what crime is and the different kinds of crimes and within these official categories the quite wide range of actual circumstances that could have led to that particular charge. And I stress in the paper that there’s no way of knowing from the statistics within each of these categories what’s the mix of the people who did serious things within that category or the people who you might feel have been overcharged or inappropriately punished within that category, you can’t tell that from the statistics. But just to remind people that you don’t know for sure exactly what the circumstances are just from reading the offense category.

Chancellor: I do want to give you an opportunity to give us some final thoughts here. What was your takeaway after doing this research and writing this paper, where did it leave you, and where would you want someone who has read this paper or heard about it, what do you want to leave them with?

Oliver: That’s a good question. One of the things I want to leave them with is actually frustrating for everybody, which is it’s not as simple as you think it is. The causes of mass incarceration and the ways in which you’re going to undo mass incarceration are not going to fit in a ten word slogan, but really involve many, many different policies and practices. A lot of different things went into mass incarceration and a lot of different things are going to have to go into undoing mass incarceration. And I also think that for people who care about these issues, it’s important to realize these huge differences between states and basically the need to essentially do this kind of an analysis separately for every state so you can figure out what it actually takes to undo in your particular state and no just assume that some reform that happened in New Jersey or California or Connecticut or wherever will have the same consequences in your particular state if the configuration of what was going on in your state is different.

Chancellor: Again, a big thank you to Professor Oliver for taking the time to talk to us. If you’re interested in this work, you should check out her paper and others in Volume 103, Issue 3 of the Marquette Law Review published in June of 2020. This podcast was supported as part of a grant from the U.S. Department of Health and Human Services, Office of the Assistant Secretary for Planning and Evaluation but its contents don’t necessarily represent the opinions or policies of that Office, any other agency of the Federal government, or the Institute for Research on Poverty. To catch new episodes of the Poverty Research and Policy Podcast, you can subscribe on Apple, Stitcher, or Google Play Podcasts. You can also find all of our past episodes on the Institute for Research on Poverty website. Our theme music for this episode is “Staring Straight” by Maarten De Boer. Thanks for listening.


Incarceration, Justice System, Policing, Prisoner Reentry