If Amazon is correct, there are currently more than 150 books in print with the phrase “mass incarceration” in the title or subtitle. You need no special knowledge of crime and punishment in the United States to infer correctly that the term is not one of praise for our criminal justice system. “Mass incarceration” means not simply that many Americans are behind bars, but that too many Americans are behind bars.

Currently, about 2.2 million individuals are incarcerated in the nation’s federal prisons (190,000), state prisons (1.3 million), and local jails (725,000). These are, admittedly, depressingly high numbers—and much higher, adjusted for population, than in other Western democracies. Prisons, however, are a response not to a population problem but to a crime problem. So the question is not whether the United States has too many people in prison for a country of its size, but whether it has too many people in prison for a country with its number of crimes and convictions. (Note that of the local jail inmates, about three fifths are awaiting trial and most of the rest are serving short sentences of less than a year.)

* * *

Start with crime. According to the FBI, in 2015 law enforcement authorities reported over 15,000 homicides; 124,000 rapes; 327,000 robberies; and 764,000 aggravated assaults (that is, assaults with a deadly weapon or those causing serious bodily injury). That’s 1.2 million very serious violent crimes known to the police. Authorities also reported 1.6 million burglaries, of which more than a million were of personal residences. Residential burglary is the most serious and traumatic of the property crimes and the one that comes closest to the personal violation characteristic of violent crimes. (Victimization surveys show that the true prevalence of crime in the United States, including crimes not reported to the police, is more than twice as high as official police data.)

All of these crimes are felonies that can land one in prison. And these numbers do not include such serious offenses as kidnapping, indecent liberties with a child, other sex crimes short of rape, and drug trafficking, for which we have no national incident data. Of course, not all criminals are arrested. For some of these serious crimes arrest rates are shockingly low: just 29% for robbery and 13% for burglary. Yet, altogether, police made over 10 million arrests in 2015, 1.5 million for violent crimes (including misdemeanor assaults).

What, then, of convictions for serious crimes? The most recent data from the Department of Justice’s Bureau of Justice Statistics (BJS) show more than 1.1 million felony convictions in state courts each year (2006) and 72,000 felony convictions in federal courts (2014). For the convictions in state courts, over 200,000 were for a violent crime; 100,000 for burglary; and over 200,000 for drug trafficking. Many will be surprised to learn that only two fifths of those convicted of felonies in state courts are actually sentenced to prison. Of the rest, about half receive no incarceration (mainly probation) and half are sentenced to a short term in a local jail. Indeed, at any one time there are more than twice as many convicted offenders on probation or parole—that is, not incarcerated—as there are in the nation’s prisons or jails.

With offenders each year committing well over a million very serious violent crimes and another 1.6 million burglaries, with police each year arresting 1.5 million violent offenders, and with courts each year convicting more than a million persons for a felony, it is perhaps not so surprising that state and federal prisons hold one and a half million inmates. If we have a “mass incarceration” problem it appears to be because we have a “mass crime” problem, despite the downward trend of the past two decades.

Are one and a half million felons in state and federal prisons too many? Are our prisons filled with low-level offenders, especially small-time drug users, who could be released with no harm to the public? And what of the violent offenders, burglars, and drug traffickers, who together constitute fully three fourths of all state prisoners? Would the nation be better served if hundreds of thousands of these offenders were released into our communities or never sent to prison in the first place?

* * *

It is a great virtue of John Pfaff’s Locked In: The True Causes of Mass Incarceration and How to Achieve Real Reform that it thoroughly refutes the myth, which he calls the “Standard Story,” that the “war on drugs” was responsible for “mass incarceration”—a thesis made famous by Michelle Alexander in her bestseller The New Jim Crow (2010). Reporting data from BJS, Pfaff shows that only 16% of state prisoners are serving time for a drug offense as their most serious conviction, and only 3.5% for drug possession. And he notes that even the relatively few serving time for simple possession may have committed a more serious drug offense but were allowed to plea to the lesser crime. He cites studies showing that about a quarter of drug offenders in prison had previously been convicted of a violent crime, that a fifth had used a gun in a previous crime, and that only about 1% of all state prisoners were “nonviolent, first- or second-time drug offenders.” He even titles one section of his chapter on the war on drugs “The Myth of the Low-Level, Nonviolent Drug Offender.”

A professor at Fordham Law School with a Ph.D. in economics, Pfaff goes on to challenge Alexander’s further claim that the war on drugs (which Alexander attributes to anti-black bias) explains the disproportion of blacks in America’s prisons. He shows to the contrary that whereas blacks in 2013 were 38% of all state prisoners serving time for a drug offense, they were 37% of the much larger number of those serving time for non-drug crimes. Remove all drug offenders from prison and the racial disproportion of state prison inmates would remain essentially unchanged.

How, then, to achieve the kind of massive reductions in incarceration that Pfaff believes are necessary to reduce the “hard-to-estimate ‘collateral’ costs” of mass incarceration, such as the income inmates lose while behind bars, the emotional costs on inmates’ families, the reinforcing of racial biases and inequalities, and the increased future health costs that former inmates face? No real progress will be made until we confront, as he titles his seventh chapter, “The Third Rail: Violent Offenses.” Most simply, we must send fewer violent offenders to prison and shorten the time behind bars of those we do send. States and counties must “rethink how they punish people convicted of violent crimes, where ‘rethink’ means ‘think about how to punish less.’”

* * *

Pfaff frankly acknowledges that he is talking here about “serious violent crimes.” One quarter of those in prison for a violent offense, he tells us, are serving time for murder or manslaughter (166,000 in 2012); another quarter for robbery (182,000); and 10% for aggravated assault (133,000). Add the 13% serving time for rape or other sexual assault (165,000), which he doesn’t mention here, and the total just for these four most serious violent crimes is 646,000, about half of all state prison inmates. “If we are serious about wanting to scale back incarceration, we need to start cutting back on locking up people for violent crimes.” To put it simply, we must be “less punitive.”

So, just how punitive are we now toward violent and other serious offenders? First, as noted above, most of those convicted of felonies in state courts are not sentenced to prison but to probation (supervision in the community) or a short sentence in local jail. Even for serious violent offenders surprising numbers are not sent to prison: 28% for rape; 29% for robbery; and 57% for aggravated assault. Although Pfaff does not address these specific figures, his argument requires that more of those convicted of rape, robbery, and aggravated assault get either no incarceration at all or at most a short stint in a local jail.

What about time served? Just how long do we now keep violent offenders behind bars? Pfaff calculates that time served in prison for all those convicted of murder, rape, robbery, and aggravated assault averages 3.2 years—“likely less,” he says with some understatement, “than one would expect for index violent crimes.” (The FBI tracks incidents of four violent crimes and four property crimes as an “index” of the overall crime problem.) Including non-violent crimes, the overall average is just 1.7 years. More detailed data from BJS (which Pfaff does not cite) show that time served for all major violent crimes increased from 1986 to 2009 (the last year with published data). For murder (excluding all forms of manslaughter), the average number of years behind bars before release rose from 6.8 to 14.6; for rape, from 3.8 to 7.8; for robbery, from 3 to 4.8; and for aggravated assault, from 1.9 to 2.7. And keep in mind that most of those in prison have a prior criminal conviction, about half have been to prison before, and about a third have been convicted of a crime at least three times in the past.

* * *

Many would consider these increases in time served to be real progress: violent (and often repeat) offenders getting closer to what they deserve for murdering, raping, robbing, and assaulting. But for Pfaff, what our worst offenders deserve has nothing to do with it. His is a purely utilitarian approach; for he tells us that there are really only two reasons to send someone to prison: deterrence and incapacitation. Yet deterrence, he argues, is not increased by sentence length, only by swift and certain punishment; and incapacitation has diminishing returns because offenders “age out of violence”: “They are not violent people; they are simply going through a violent phase.” But if you look at our best recidivism data (a BJS study that tracked for five years 400,000 offenders released from prisons in 30 states in 2005), you discover that fully 69% of those at least 40 years old at the time of release were rearrested for a felony or serious misdemeanor within five years—not much lower than the 77% for all offenders. The aging effect, apparently, does not turn most hardened criminals into law-abiding citizens, though it may slow them down a bit. (BJS does not show what percentage of those 40 and older were rearrested specifically for a violent crime. For all those released, it was 29%.)

Pfaff admits that there are “moral arguments for severity, but I won’t address those here.” Although some people believe that prisons serve “moralistic, ‘retributive’ goals,” “retribution…isn’t really focused on what prison accomplishes.” This seems an odd way to put the point. For the retributivist, prison accomplishes something quite important indeed: retribution. And retribution, in turn, educates the offender in his moral responsibility, assures victims (and the loved ones of murder victims) that they live in a just society, announces and upholds society’s moral norms, and promotes respect for the legal order. A rapist may not be deterred by the knowledge that some other rapists go to prison, and you may well be able to incapacitate the convicted rapist through chemical means; but something good is accomplished when rapists receive their just deserts quite apart from measurable utilitarian effects. Alexis de Tocqueville captured the point in Democracy in America when he contrasted the punishment for rape in the United States and France in the 1830s. A capital crime in the United States, rape in France was “visited with far milder penalties,” if one could get a conviction at all. In this the French showed both “contempt of decency” and “contempt of women.” But for Americans “no crime is visited with more inexorable severity by public opinion.” Our criminal punishments both reflect and reinforce our moral judgments. Yet for Pfaff the whole issue of desert can be effectively ignored in a 300-page, heavily documented, data-filled book that would show us how to achieve the goal of “real decarceration.”

* * *

Of course, today’s average American, like those in Tocqueville’s time, still thinks in terms of desert. Ask him or her to approve a plan to send fewer murderers, rapists, or robbers to prison or to reduce time served to less than 14 years for murder or 7.8 years for rape, and you are not likely to get very far. Pfaff understands that open and public debates about reducing punishment for violent offenders will not give him the results he wants, at least not until he and others can effect “a significant cultural change” in attitudes about violent offenders. But that will take some time. In the meantime steps must be taken to “better insulate the criminal justice system from political passions.” “Political accountability,” it turns out, is bad for criminal justice policy. As a “check on criminal justice actors” it is “systematically flawed.”

So what can be done to counter “tough-on-crime voices” in the electorate pending the future cultural transformation? First, he proposes to remove popular control over county prosecutors by replacing elections with appointment. And, second, by all means remove “(some) direct criminal justice policymaking from the hands of legislators, who are immediately accountable to voters and thus prone to overact” by establishing within the states at least quasi-independent sentencing commissions to set the key punishment policies. The goal in all this is to give “experts a bigger voice in policymaking,” thus severing punishment policies and practices from public moral judgments and deeper matters of justice. Once all this is accomplished, we will need a new name for the policy area; for what we used to call “criminal justice” will have ceased to be about substantive justice at all.

* * *

Prison Break: Why Conservatives Turned Against Mass Incarceration by Ph.D. candidate David Dagan and political scientist Steven Teles, both of Johns Hopkins University, is a very different kind of book. It tells the story of how some prominent conservatives shifted over time from advocating tough punishment for criminals to opposing mass incarceration. The authors document in some detail the influence of such key players as Newt Gingrich, Grover Norquist, Pat Nolan of California, and Charles Colson’s Prison Fellowship; of alliances between conservatives and center-left and even far-left think tanks; of criminal justice changes in the “deep red” states of Texas and Georgia; and of efforts to change federal law. “Conservatives,” they write, “are changing their minds on criminal justice. They are not doing so because new information has suddenly emerged. In fact, compelling evidence on cost, recidivism, and the racial disparities of mass incarceration has been available for well over two decades, but only recently have conservatives paid attention to it.”

It turns out that the conservative case for combatting the crime epidemic of the 1960s, ’70s, and ’80s had little to do with facts, reason, and argument, and everything to do with politics. “[R]apidly spiking crime gave Republicans plenty of kindling with which to enflame the passions of conservative white voters”; “[c]rime rhetoric was thus loaded with racial appeals”; “frenzied rhetoric generated even greater public fear”; “[d]emagogu-ery” in Texas “combined with the state’s deep conservatism…[had] produce[d] an overheating machinery of punishment without an ‘off’ switch”; and Georgia conservatives used “blustery rhetoric” and “fiery rhetoric” to defend the imprisonment of violent offenders. All this “frenzied,” “blustery,” “fiery,” racialized, and demagogic rhetoric was marvelously effective in getting Republicans into office, which, of course, was the whole purpose.

* * *

Lest the reader miss the point, here is the authors’ summary statement in their conclusion:

In the punitive era policymakers were encouraged to continue piling up longer sentences because they learned that it paid serious political returns. Like the mouse who gets rewarded with a piece of cheese each time he presses on the right lever, policymakers who pulled the incarceration lever continually received pleasing benefits from voters and supportive interest groups. So they kept pulling, ignoring the negative consequences of doing so.

All those mayors, county prosecutors, state legislators, governors, and congressmen who sought to reverse the major drop in punishment in the 1960s and early ’70s while crime was skyrocketing; who worked to establish “truth in sentencing” by reducing massive “good-time” allotments to imprisoned violent offenders; who fought to see to it that murderers served more than six to seven years and rapists more than three to four; who pushed laws to escalate punishments for those who repeatedly victimized others; and who combatted the devastation to inner-city communities by the spread of crack cocaine and other drugs—all these, just mindless mice, pulling levers.

Though you would hardly know it from the account here, these public officials actually succeeded in turning the corner on the crime problem: large reductions in both violent and property crime between the early 1990s and a year or two ago. Had they not moved aggressively against violent criminals, our communities today, and especially minority communities in our major cities, would be much more dangerous places in which to live and work. We owe these public officials not our disdain but our gratitude. (Full disclosure: I worked for one—Richard M. Daley, the elected state’s attorney of Cook County, Illinois—from 1981 to 1984.)

One of the frustrations that Dagan and Teles share with Pfaff is the tendency of “reformers” to push for leniency for drug and property offenders in order to ensure sufficient prison space for violent offenders. As an important lawmaker in Georgia told the authors, “these potentially non-violent addicts…were taking up bed space that should be occupied by violent criminals.” Over and over, conservative opponents of “mass incarceration” make the same point. But like Pfaff, Dagan and Teles well know that they will never achieve the prison reductions they seek without reducing punishment for violent offenders. After spending a chapter praising reforms in Georgia, they concede that the changes of the type enacted there would “make only modest dents in a huge problem of over-incarceration,” for the leaders of justice reform in Georgia “largely shied away from tackling one of the key drivers of prison growth: unnecessarily long sentences for violent criminals.”

So how long do violent criminals now serve in Georgia? We are not told. How long should they serve? We are not told. Are Georgia murderers serving “unnecessarily long sentences”? What about rapists, robbers, and those who seriously injure their victims? The reader is in no position to judge because the authors simply assume that with so many people behind bars, even serious violent offenders must be serving too long.

* * *

If we learned anything from the past half century, it is that punishment works. Reduce the likelihood that criminals will go to prison or spend much time there, and you will get more crime. Increase the costs that criminals will pay for their deviant behavior, and crime will go down. Some of this is deterrence; some is incapacitation; and some results from the ways in which criminal laws reinforce moral norms. Of course, recognizing the value of prisons and punishment is not inconsistent with doing everything we can to keep young people from launching criminal careers in the first place, or with providing adequate funding within our prisons and jails for educational, vocational, mental health, and substance abuse programs.

One has to admire the clear-eyed approach of the authors of these two new books: we have too many people in prison; most inmates are violent offenders; therefore no serious decarceration will be achieved without reducing punishment for violent crime. Though the “experts” will continue to push in this direction, sober public opinion will resist, and rightly so. The crime drop of the past two decades happened for a reason, and the recent increase in violent crime throughout the United States reminds us of the stakes. Americans expect their government to protect them from those who murder, rape, rob, and assault. It is, after all, why we have government. Sending fewer serious violent offenders to prison or reducing their sentences breaks faith with the American people and puts us all at greater risk of a tragic encounter with the lawless.