The unbridled discretion of prosecuting attorneys—and not the war on drugs or draconian sentencing—is largely responsible for America’s mass incarceration mess.
So says John Pfaff, whose contrarian book, Locked In: The True Causes of Mass Incarceration—and How to Achieve Real Reform (Basic Books), has positioned the Fordham law professor as an ascendant criminal justice intellectual. The book serves as a counterweight to Michelle Alexander’s widely lauded The New Jim Crow, which argued that the institutional racism of the drug war packed our prisons. (Charles Lane of the Washington Post recently called Pfaff’s book a “corrective” of Alexander’s assertions.)
Building on a series of scholarly papers he published over the past decade, Pfaff, who holds both a law degree and a Ph.D. in economics from the University of Chicago, uses his book to challenge the “standard story” of mass incarceration, making his case through data analysis. In a two-part Q&A with The Crime Report’s David J. Krajicek, Pfaff describes how pursuit of reforms fitted to the flawed “standard story” narrative has distracted from a more appropriate redesign of punishment.
The Crime Report: You criticize the three-legged narrative about mass incarceration: that government-orchestrated drug convictions, longer sentences, and unscrupulous private prison firms were largely to blame for overfilling penitentiaries. What’s wrong with it?
John Pfaff: It distracts us from confronting the tougher, but more central, issues. In a recent poll, a majority of Americans—from liberals to conservatives—said they opposed being less severe towards those convicted of violent crimes, even if such people pose little to no risk of recidivism. Yet over half of all people in prison are there for a violent crime; we can’t cut prisons in any meaningful way without changing how we punish violence.
The constant drumbeat of “prisons are full of low-level non-violent drug offenders” has produced a public still unwilling to tackle what has to be tackled. Similarly, by continuously paying attention to private prisons, reformers and the public end up ignoring the far more powerful public correctional officer unions. And the insistence on cutting excessively long sentences at the legislative level has produced a reform effort that has taken almost no steps to meaningfully reduce the charging power of prosecutors, even though that power is the most important engine of prison growth.
I wouldn’t call the narrative “false,” though. Each of the problems it identifies are real problems: too many drug offenders are in prison, we lock people up for too long, private prisons can wield problematic power. But each takes our attention away from something related that is far bigger and far more important. It leads us to tweak things on the margin rather than attacking the heart of the problem.
TCR: Michelle Alexander has replied that stepped-up drug enforcement was the leading edge of a punitive wave that swept across criminal justice. Agree?
Pfaff: I don’t, or at least not entirely. Take New York State. It passed the Rockefeller Drug Laws, some of the harshest drug laws in the country, in 1973. Yet by 1984 there were fewer people in New York prisons for drugs than in 1973. The spike in drug prisoners starts in 1984—when crack-related violence swept across the state. And then the number in prison for drugs starts to steadily fall in 1999, which was years before any reforms were passed…I don’t think it is so easy to disentangle the appeal of war-on-drugs rhetoric from the rising violence that was occurring at the same time.
No doubt there’s a feedback loop: rising violence makes drug-war rhetoric more appealing, and drug-war rhetoric perhaps pushes us to take a tougher approach towards crime more generally, and a less public-health focused view of drug abuse more specifically. But I think violence has played the dominant role here.
TCR: The mining and analysis of criminal justice data is a mini-industry. Yet you write you were surprised to discover “the magnitude of what we don’t know.” Have we been asking the wrong questions?
Pfaff: Not the wrong questions, but I think we end up focusing on questions where the data are easiest to find and thus keep asking the same ones over and over. We have data on crimes, police and prisons, so most of the papers on prison growth—including the big National Research Council report—just keep looking at trends in crime, arrests per crime, and admissions per arrest.
In doing so, they completely ignore prosecutors, in no small part because we simply have so little data on what they do. And that creates a dangerous feedback loop: We study the things we have data on, and so we insist on more data about those things. I think one reason we lack data on prosecutors is because we don’t study them and thus don’t push for it—which is partly because we had no data on them to start with.
TCR: You urge more oversight of “profoundly powerful” prosecutors, the antagonists of your book. What data would you seek?
Pfaff: First—and this might surprise some people—if I weren’t a law professor, I would be a prosecutor (although by this point it may be hard to find an office that would hire me). That’s my bias. They aren’t really my antagonists. But they have come to use their unregulated discretion in such a way that even someone with my political bent feels compelled to push back against how they operate today.
To better understand prosecutorial decisions, I would like to see how they choose which cases to pursue and which ones to dismiss, how they choose the charges they file in individual cases, and how they operate at plea bargaining.
To do this, I’d like prosecutors to provide data on each case that the police deliver them, with information about the arrest charge, the initial charge filed in court (if different), and the final charge either tried or pled to, as well as any other charges threatened during plea bargaining. Combine that with demographic and criminal-history data on the defendant, plus demographic data on the prosecutor and (ideally) the defense attorney, and I think we could figure out a lot.
This isn’t entirely a pipe dream. I’ve seen a few papers with office-specific data on many of these things. But for now we have nothing systematic along these lines.
TCR: The journalist David Cay Johnston says criminal justice should produce the same sorts of monthly or quarterly data releases that we expect from the economic sector. Would that help?
Pfaff: I think it would, although it would be a much different enterprise. The federal Bureau of Justice Statistics has a budget of $40 million, and the FBI spends less than $2 million on Uniform Crime Reporting. The Bureau of Labor Statistics has a budget about 10 times larger, $650 million.
So while our criminal justice agencies rely on self-reports from state, county and city agencies, the BLS is able to run regular, well-designed surveys to keep track of economic trends in the US.
There’s no reason we couldn’t run similar real-time surveys for criminal justice, but it would require a massive redesign of the data-gathering infrastructure. And there would be complications, like the fact that the populations most vulnerable to crime and victimization are among the hardest to accurately survey and sample. The merging of data is also complicated by the fact that states and locales define crimes and classify prisoners in different ways, and so on.
TCR: You note (A) that Americans are a “low-information, high-salience electorate,” treating crime anomalies as trends. Meanwhile (B), we have a president and attorney general who play on that by painting the U.S. as a crime “nightmare.” You also write (C), “More important than any legislative reform effort will be attempting to change people’s attitudes toward crime.” How is C possible, given A and B?
Pfaff: We often overstate the importance of presidential rhetoric when it comes to crime policy. The 2016 election provides an interesting case study. At the same time Trump was using dated tough-on-crime rhetoric to eke out an Electoral College victory, over a dozen smart-on-crime candidates unseated tougher-on-crime incumbent prosecutors. And Oklahoma, where Trump won by a wide margin, passed two state-wide referenda decriminalizing all sorts of drug offenses and funding treatment.
I think people’s views on crime are influenced much more by local conditions than by what people in D.C. are saying.
That said, I agree that changing attitudes will be tough. Part of what we need is just a generational shift. It doesn’t surprise me that the people in Congress pushing for reforms are younger senators like Cory Booker (D-New Jersey), and those resisting are older senators like Chuck Grassley (R-Iowa). If you’re my age, 41, crime has been dropping for almost all of your politically aware life; Boomers still remember much more acutely the rise in crime that happened in the 1960s, ‘70s, and ‘80s.
As a result, I think younger cohorts will generally be less punitive. JK Galbraith once joked that “funeral by funeral, economic theory advances.” Sadly, I think a lot of things advance that way…
I think the challenges posed in your question point to the need to shield criminal justice actors from the worst of immediate democratic accountability. Perhaps we should appoint rather than elect local prosecutors. And we should definitely stop electing judges. Perhaps we should force new criminal laws to go through a sentencing commission that can slow down passage so the “crime of the week” has less of an impact on policy. So far, however, there has been almost no talk about how to make sure criminal justice actors don’t have to (over) react to anger produced by rare but high-salience cases.
TCR: The FBI still groups crime statistics into headings that date from the early years of J. Edgar Hoover’s reign. Should we rethink our concepts of crime itself?
Pfaff: The Uniform Crime Reporting headings still seem to make sense: most index crimes really do seem like the most serious (murder, rape, robbery, aggravated assault, burglary, theft, auto theft, and arson). I think what people need to appreciate is what a rough estimate the UCR is.
It rests on a lot of assumptions, like the fact that it measures only reported crimes—and only a handful of crimes at that. Moreover, it isn’t our only measure of crime: We have the annual National Crime Victimization Survey; we have new data like ShotSpotter, which suggests very few gunshots are reported to police; we have victimization and drug overdose data from ERs and hospitals, and so on.
None of these tells the same story. From the 1960s to the 1980s, for example, the UCR was rising while the NCVS was falling. And it wasn’t necessarily that one was “right” and the other was “wrong,” but that they were measuring slightly different things which, when combined, tell a slightly different story than either one alone.
Across all these various data sources, we have a rough idea of the general trends in crime, but rarely if ever do people really try to put together this comprehensive picture. Instead, they tend to take the UCR as the objective measure of crime, which it isn’t. It’s quite useful, but it is just one snapshot of a complex phenomenon.
Tomorrow: In Part Two, Pfaff describes the revelatory moment—while reading datasets from the National Center on State Courts—that convinced him unchecked prosecutorial discretion had been overlooked as a fundamental cause of America’s prison population boom.
David J. Krajicek (@djkrajicek) is a contributing editor of The Crime Report. He welcomes comments from readers.