Fordham law professor John Pfaff argues in his counterintuitive new book, Locked In: The True Causes of Mass Incarceration—and How to Achieve Real Reform (Basic Books), that the unbridled discretion of prosecuting attorneys has been largely overlooked as a key to America’s prison population boom.
In the second part of a Q&A with TCR Contributing Editor David J. Krajicek, Pfaff, who trained as an economist at the University of Chicago, argues that the punitive approach to violent offenders favored by tough-on-crime politicians and prosecutors needs re-thinking if we want to shrink the number of Americans behind bars.
The Crime Report: Explain how prosecutors, whom you describe as “the engines driving mass incarceration,” were responsible for the firehose of prison admissions, even as crime was declining.
John Pfaff: My findings come from a study of 34 states over the years 1994-2008. Crime falls over this time period, and serious crime falls by a lot. Arrests fall too. So fewer people are entering the criminal justice system, yet the number of felony cases filed in state court rises sharply. In my data, the probability that an arrest results in a felony charge almost doubles.
That’s the only change.
Once a charge is filed, the probability that someone is admitted to prison doesn’t change, nor does the time spent in prison if admitted. Prosecutors filing more felonies is the primary engine of prison growth.
I have no clear idea why this happened, because we have almost no data on prosecutors. But I have a few theories with some scattered empirical support. Right now, I think the most significant factor is staffing. Two big changes happened. First, between 1974 and 2008, the percent of counties with a full-time prosecutor rose from 45 percent to 85 percent. Second, from 1974 to 1990, as crime rose steeply, the number of assistant prosecutors nationwide rose by 3,000, from 17,000 to 20,000. But from 1990 to 2008, as crime fell dramatically, we hired 10,000 more assistant prosecutors, bringing the total to 30,000.
So rural and suburban counties professionalized, and urban areas massively ramped up staffing as crime fell. I find no evidence that an individual prosecutor is harsher in 2008 than in 1974; it’s just that we have so many more of them, and they need to prosecute to justify their jobs. We arrest 12 million people per year. There are always people for prosecutors to prosecute.
There are other possibilities too. Perhaps elected district attorneys have grander political ambitions (attorney general, governor, senator) and see toughness as a campaign tactic. Police may be making better arrests that are easier to prosecute, either because they are better trained or we have more DNA, video, and other compelling evidence. Maybe longer sentences have given prosecutors more leverage during the plea bargain process even if people aren’t actually serving more time in the end (because they take the plea, just faster than before).
Figuring out empirically how prosecutor’s offices operate is an area crying out for more research (and for more—much more—data).
TCR: You say you had an “aha moment” while studying state court felony filings.
Pfaff: Almost every study examining where prison growth occurred looked at just four stages: changes in crime; in arrests per crime; in prison admissions per arrest; and in time served per admission. The admissions-per-arrest piece always bothered me because it implicated so many actors. Was it a change in policing, in charging, in sentencing? So I assumed there had to be data of some sort on prosecutors out there.
Eventually I came across a data-set on felony cases filed in state courts compiled by the National Center on State Courts. (It’s telling that my data on prosecutors came from the court system: Prosecutors provide almost no data to the public, but they can’t avoid the disclosure that comes once the case hits court.)
By this point I had already demonstrated that time served wasn’t driving prison growth, but I fully expected that the rise in admissions was being shaped strongly by a lot of factors—arrests, filings, and chance of admission. So I was completely surprised when I included the filing data, and the impact of everything else basically melted away. What made it particularly striking is that I wasn’t doing any fancy modeling…This was all simple algebra, and it showed that filings really were the major force. I realized right away that it had some major implications for how we think about prison growth.
TCR: Explain the concept of prison admissions vs. prison population and why you see that as a key overlooked data point.
Pfaff: My concern is this: Prison is a remarkably costly and inefficient way to fight crime, so our goal should be to scale back the number of people who come into contact with prisons altogether—but that’s the number admitted, not the number serving on any given day. And what I discovered is that several states that cut their total prison population were admitting more people. So what looked like a success—“we have fewer people in prison”—was in many ways masking a failure, since more people and their families and communities were incurring the costs of exposure to incarceration.
Think of it this way: A state admits two people to prison each year, and each serves two years in prison. Every year there are four people in prison–two serving their first year and two serving their second. Then the state changes its policy: Everyone serves only one year in prison. But prosecutors now act tougher and send three people to prison each year. Now the prison population is three, each serving a one-year term.
So the prison population falls from four to three, but the number admitted rises from two to three, so more people come in contact with prisons, which imposes all sorts of excessive costs.
I initially sketched this out as a thought experiment, but I discovered that this was actually happening in about 10 states. We talk about trying to shrink prison populations, but we should really be focused on shrinking the number of people who have been prisoners. Somewhat counterintuitively, these are not always synonymous ideas.
TCR: Explain the implications as we emphasize parole reforms rather than crimping prison intake.
Pfaff: The median time to release for someone convicted of a drug or property crime is one year. The median time for someone convicted of violence is four years. People don’t spend huge amounts of time in prison. Parole reform is a good idea, but my guess is that its impact is going to be disappointing. It would be far more effective to divert a lot of unnecessary admissions.
Almost all parole reforms are restricted to people convicted of non-violent offenses. Yet those reforms would have a potentially important impact for those serving time for violent crime—and that’s an area we have long refused to confront.
TCR: Like many others, you say that dealing with the 700,000 violent offenders in state prisons is a linchpin of mass incarceration reform. You also note the perilous politics involved. Can those two things be squared?
Pfaff: I think so. One intriguing, rarely discussed aspect of the recent prison population decline is that between 2010 and 2013 the number of people in prison for violence fell by over 30,000. Now it’s possible that much of that reduction occurred in California. But still: We cut back on incarcerating people for violence, and there was no political backlash.
There are real political challenges, which is one reason why I push so much against the “standard story” conventional wisdom of mass incarceration. By constantly overemphasizing the importance of non-violent drug offenders to prison growth, reformers have convinced Americans that they can cut the prison population substantially without having to make any really tough tradeoffs, or without having to look carefully at the way politics and punishment interact with each other. But we have to make these tradeoffs, and we need to confront them more directly and openly and honestly—and find politicians willing to take principled risks. It’s not going to be an easy process, and there’s no guarantee that we’ll succeed.
TCR: You criticize the country’s Balkanized set of city, county, state and federal criminal justice systems as “a swirling mess of somewhat antagonistic agencies.” How might organizing the Hydra help?
Pfaff: We first need to appreciate how often we divorce costs from benefits in criminal justice, and then we need to look for ways to make actors focus on both. Prosecutors are elected at the county level but send people to state-funded prisons. Cities might not adopt efficient police reforms often enough because much of the savings will come through reduced prison populations: cities bear the cost, states the benefit. Problems exist even within jurisdictions: safer, wealthier, whiter suburbs play a major role in electing prosecutors who enforce the law more often in poorer, more minority urban cores. The suburbanites feel the benefits of aggressive enforcement but few if any of the costs, so there is less political cost to being needlessly aggressive.
There are a few efforts underway to address these disconnects, but not many. The fledging Justice Reinvestment Initiative, for example, tries to encourage local innovation by providing a vehicle for the state government to share any savings from prison reductions that come from local policy improvements. I think some reforms will require us to devolve more decisions to local governments and others to centralize more at the state level, but the current arrangement is almost entirely incoherent and contributes greatly to the injustices that we see every day.
TCR: We have begun to view drug offenders not merely as lawbreakers but as flawed citizens who need help. Is this a key to broad prison reform: to regard violent offenders through that same social science prism?
Pfaff: Yes, exactly. It’s much harder in the context of violence, of course, because our desire for vengeance is much greater when there is an identifiable victim. In many ways, I think most takes I’ve seen on the conservative push for criminal justice reform focus on the wrong part of that movement. Most accounts tend to look at the efficiency/cost-cutting side (prison as an excessive expense) or the libertarian side (the state impeding personal liberty, particularly when it comes to drugs).
To me, the group with the most promise to really push for deep reforms is the evangelical branch, which favors reform because it genuinely believes in redemption. They seem like the ones best positioned to argue that even people who commit serious crimes remain, well, people.
There are reasons to be cautiously hopeful. A recent large-scale survey found that crime victims tend to be less punitive and more in favor of restorative justice approaches than the average person. This is because victims often come from the high crime areas and have a much more nuanced view of what it means to be both a victim and a victimizer.
That is why I favor increased localism: Those closest to crime are more likely to have more sophisticated takes. That doesn’t guarantee local decision-making will necessarily lead to less punitiveness. (As recent work has shown, some of the most punitive groups in the 1960s and 1970s were African-American communities that bore the brunt of the rise in crime.) But I believe it will lead to more careful approaches to crime. A prominent conservative judge, Richard Posner, just wrote in an opinion that we need to think differently about how we punish people convicted of violence. I think the conversation is starting to shift as we slowly begin to view violence in a more sophisticated way.