Criminal justice policy in the United States is deeply dysfunctional—or so we often hear. We are told that we spend too much money imprisoning too many offenders for too long, with little, if any, benefit to public safety. For almost half a century, thousands of lawmakers across the nation have persistently set excessive punishments for crime when other means exist that could lower incarceration rates, save considerable money, and make us safer than we now are.

And what explains this dysfunction? In Prisoners of Politics: Breaking the Cycle of Mass Incarceration, NYU law professor Rachel Barkow blames the “politics of fear.” More precisely, she blames the “ill-informed” voters—moved by their fearful “gut reactions” to crime, the media’s sensational reporting, and ambitious politicians’ rhetoric (which often has “racist overtones”)—who elect officials who set “irrational” criminal policies.

To break the cycle of mass incarceration, Barkow says, we must remove from the democratic process (which is to say, from the control of the people) many of the key questions of criminal justice, especially how much to punish criminals.


To break the cycle of mass incarceration, Barkow says, we must remove from the democratic process (which is to say, from the control of the people) many of the key questions of criminal justice, especially how much to punish criminals. We should put such questions in the hands of “experts with specialized knowledge”; insulate them from political forces; and let them conduct empirical, cost-benefit analyses that could “maximiz[e] public safety and mak[e] the most of [our] limited resources.”

Start your day with Public Discourse

Sign up and get our daily essays sent straight to your inbox.

The Sentencing Reform Movement

The principal target of Barkow’s criticism is the shift in criminal justice law that began in the 1970s, especially the move to “determinate sentencing.”

Until 1976, every state used “indeterminate sentencing” for convicted felons. Under this system, when judges sentenced a felon to prison (rather than probation), it was typically for a broad range of time—say one to ten years, or ten years to life—and it was left to a parole board (staffed by experts) to determine the actual time behind bars by assessing the progress of the offender’s rehabilitation.

But starting in the mid-1970s, and into the 1980s, about a third of the states (including many large ones), as well as the federal government, began to adopt “determinate sentencing,” whereby the judge would set a specific sentence that could only be reduced by earned good-time in prison. This reform, which put parole boards out of business for most crimes in these states, was pushed by a kind of bipartisan “revolt against the experts” (my phrase, not Barkow’s). Liberals worried that the broad discretion permitted under indeterminate sentencing enabled racial and ethnic discrimination. Conservatives thought that the old system administered too little punishment and showed little success at rehabilitating criminals.

Reform began in 1976, when California required judges to choose one of three specific sentences for most felons: a mitigated sentence, a presumptive (or default) sentence, and an aggravated sentence. For example, the specified sentences for the rape of an adult were 3, 6, and 8 years. Also under the law, inmates could earn a day off for good behavior for every day that they served, therefore reducing the likely actual time served in prison for most rapists to 3 years.

Minnesota’s legislature adopted a very different kind of determinate system, which later served as the inspiration for the federal sentencing guidelines. (Note: Barkow herself served as a member of the U.S. Sentencing Commission from June 2013 to January 2019.) Minnesota’s guidelines used a two-dimensional grid that ranked crimes by seriousness down the side, and by the convict’s prior criminal history across the top. The cells specified the presumptive length of the sentence, either prison or probation, within a very narrow range. Like California, Minnesota specified a shockingly lenient punishment for rape: just two years and five months behind bars for a first-time offender who behaved himself in prison.

The shift to determinate sentencing was part of a broader “truth in sentencing” movement, which also resulted in the federal government’s and most states’ reducing good-time for violent offenders to a maximum of 15 percent.

The Reason for the Reform

By the late 1970s, many voters and their elected officials had come to believe that the experts had not protected public safety well in recent decades, and that the penal system needed to be tougher. This reaction was perhaps not so irrational.

First, as Barkow herself notes, “violent crime increased by 126 percent between 1960 and 1970, and by 64 percent between 1970 and 1980.” Similarly, property crimes more than doubled in the 1960s.

What Barkow does not mention is that, in the face of this unprecedented rise in crime, the experts who then ran our criminal justice systems oversaw a drop of 12 percent in the number of federal and state prisoners: from 226,344 in 1960 to 198,831 in 1970.


What Barkow does not mention is that, in the face of this unprecedented rise in crime, the experts who then ran our criminal justice systems oversaw a drop of 12 percent in the number of federal and state prisoners: from 226,344 in 1960 to 198,831 in 1970. In many large states the downward trend was even steeper: a drop of 18 percent in Ohio; 19 percent in Georgia and Michigan; 22 percent in New York; 23 percent in Missouri; and 43 percent in Virginia.

Let that sink in. At a time when crime more than doubled, our incarcerated population shrank. This meant either that fewer felons were going to prison, or that they were serving less time there, or both. Although our data are sketchy for this period, it appears that both were true.

According to the Bureau of Justice Statistics of the U.S. Department of Justice, by 1981 the median time in state prison for all convicts released that year was just 1 year and 5 months. Murderers served a median of 5 years and 3 months (that’s not a misprint); rapists, 2 years and 9 months; robbers, 2 years and 1 month; those convicted of aggravated assault, 1 year and 5 months; and burglars, 1 year and 1 month. Compared to 1960, time served was down 15 percent for aggravated assault, 26 percent for robbery, and 35 percent for burglary.

In response, the public asserted itself, and lawmakers modestly toughened the criminal justice system. Barkow claims that “the shift to mass incarceration is directly linked to [this] shift from leaving judgments to professionals to allowing the masses to set policies directly.”

But these reforms were hardly unreasonable.

Between 1986 and 2009, the average time served in prison for rape rose from about 3 and a half years to just under 8, and for murder (excluding all forms of manslaughter) from about 7 years to 15—a doubling of the punishment for these crimes. Most Americans would likely call this progress: our worst offenders now get something closer to what they deserve than in the days when the experts were more in charge.

Moreover, Barkow provides almost no data to justify her claim that the current system is excessive. For example, she does not mention that of all those convicted in state courts of a felony in 2006 (the most recent year of the published data), only 41 percent were sentenced to state prison (typically for a year or more). The rest were sentenced either to some months in a local jail (28 percent) or to probation or some other lesser sentence (31 percent). In addition, nearly a fourth of those convicted of violent felonies received no prison or jail time.

Retribution: More Democratic, More Just, and More Effective

In his 1986 reflection on the effectiveness of Minnesota’s guidelines, Tom Johnson, the district attorney of Hennepin County (which includes Minneapolis), argued that what the public wants to know is whether the criminal justice system is “doing justice.” And “more than anything else,” that means “focus[ing] on the sentence and on whether it seems fair and just.” As Johnson understood, the new Minnesota system was designed to produce similar sentences for offenders who committed similar crimes and had similar criminal records. To the average Minnesotan, this was the right thing to do.

But what the guidelines were not intended to do was to give offenders the amount of punishment that the citizens believed they justly deserved. The new system was built on the notion of justice as fairness, rather than retributive justice.

In one of his few writings on criminal justice, C.S. Lewis characterized “the traditional or Retributive theory” (which he favored) as that which held that the offender should be punished “because he deserves it, and as much as he deserves.” “On [this] old view,” Lewis wrote, “the problem of fixing the right sentence was a moral problem.” And because it was a moral problem, it was a question “on which every man has the right to an opinion, not because he follows this or that profession, but because he is simply a man, a rational animal enjoying the Natural Light.”

As Barkow herself suggests, perhaps the real question underlying the sentencing debate is whether we should ground our criminal law more on justice as retribution:

To be sure, expert analysis cannot definitively answer what kind of punishment someone deserves in a retributive sense. If all we are doing with criminal law is making gut-level moral determinations about the just deserts for particular crimes, then the model I propose is not as useful. The American people and their representatives may be just as qualified as any expert to decide what seems like a commensurate punishment for a given crime, based on their notions of right and wrong and morality.

Indeed, early on in the book, Barkow herself seems to embrace retribution by maintaining that no criminal should be punished beyond “their just deserts, no matter what the data say.”

But if just desert imposes a ceiling on punishment, doesn’t it also establish a floor? Are violent offenders in the United States getting more punishment than they deserve, or not as much as they deserve? And who can better answer these questions: the experts, or the American people acting through their elected representatives?

By the end of the book, Barkow seems to have rejected retribution altogether as the proper reason for punishing criminals. Though she gives lip service to the legitimacy of retribution, she drowns out this concession by her repeated attacks on the public’s allegedly fearful and irrational reactions to crime. Barkow makes no effort to distinguish the public’s reasonable desire that offenders get what they deserve from the kind of public overreaction that lawmakers should resist.

Many who write about criminal justice from a utilitarian perspective see a conflict (Barkow calls it a “battle”) between giving offenders the punishment they deserve and maximizing public safety at the lowest cost. Yet the average person (and average legislator) sees no conflict between justice and safety. To punish the criminal “because he deserves it, and as much as he deserves” treats offenders as moral beings, thus educating them in their moral responsibilities; it assures the victims of crime that they live in a just society; it announces and upholds society’s moral norms; and it promotes respect for the legal order.

The criminal justice reforms of the past half century are evidence, then, not of the dysfunction of American democracy, but of its virtue.