A Debate About:

Locked In: The True Causes of Mass Incarceration—and How to Achieve Real Reform, by John Pfaff, https://www.amazon.com/Locked-Causes-Incarceration-Achieve-Reform/dp/0465096913

Note from the Editor:

The Federalist Society takes no positions on particular legal and public policy matters. Any expressions of opinion are those of the authors. When we publish an article that advocates for a particular position, we usually offer links to other perspectives on the issue, including ones opposed to the position taken in the article. Occasionally, as here, we publish two articles featuring opposing views. We also invite responses from our readers. To join the debate, please email us at [email protected].

 

 

An Interview with Professor John Pfaff of Fordham University School of Law

By Vikrant Reddy

In December, Congress passed and the president signed the First Step Act, the most important federal criminal justice reform legislation in a generation. The legislation made an immediate impact on the existing federal prison population by increasing the amount of good time credits that offenders can earn off their sentences (to a maximum of 54 days per year, up from 47), making retroactive the 2010 Fair Sentencing Act that reduced the disparities between federal crack and cocaine sentences, and even by including a provision to ban the shackling of pregnant women giving birth while incarcerated. The legislation is also expected to have long term effects through the allocation of $75 million over five years for expanded rehabilitative programming, a requirement that inmates be housed closer to their families to make visitation easier, and reducing the impact of federal sentencing enhancements such as the infamous § 851 (prior convictions) and § 924(c) (carrying a firearm).

Even after all of this, more legislation—a second step—seems inevitable. In April, President Trump announced that his administration was exploring ways to push even further on criminal justice reform. Moreover, new criminal justice legislation is one of the most frequently discussed topics among the candidates for the 2020 Democratic presidential nomination.

I sat down for a conversation about criminal justice reform with Professor John Pfaff of Fordham Law, the author of one of the most highly regarded criminal justice books of the decade: Locked In: The True Causes of Mass Incarceration—and How to Achieve Real Reform.

1. Your book, Locked In, is a critique of what you call the “Standard Story.” What is the Standard Story?

The “Standard Story” is a term I use to refer to three popular theories about the forces driving mass incarceration: increasingly longer sentences, the War on Drugs, and the power of private prisons. None of these theories is wrong—our sentences are long, the War on Drugs has sent a lot of people to prison, and private prisons generally are not advocates of reform—but each is far less important than most people think. The problem with the Standard Story is that focusing on these theories often blinds us to things that matter far more.

Take long sentences. Compared to Europe, our sentences are long, and they did get slightly longer over the 1990s and 2000s. But the median time to release for someone sent to prison for a drug or property crime is just one year, and it’s barely four years for a crime of violence (with most very long terms being for homicide). Sentence length matters, but it turns out that the number of admissions—and the role prosecutors play in setting those numbers—matters far more. Yet we still emphasize legislative reforms targeting sentence length while paying little (though increasing) attention to prosecutors.

Similarly, we focus a lot on drug crimes, even though only 15% of state prisoners are in for drugs, compared to 54% in for violence (and 90% of prisoners are in state prison, so they are far more significant than federal prisons in our criminal justice system). Real reform requires us to change how we punish violence, particularly serious violence like homicide, but our emphasis on drug offenses has led the public to refuse to accept this. And while we talk a lot about private prisons, they hold only about 8% of all inmates, with 92% in public facilities. About two-thirds of the $50 billion we spend on corrections goes to wages and benefits for the public employees in those state-run prisons, yet by talking so much at the private prisons, we effectively give far more powerful organizations like correctional officer unions an undeserved pass.

2. Do current federal criminal justice reform proposals address your deeper critique of the criminal justice system, or do they misguidedly address the Standard Story? (Or neither?)

Things are a little different for the Feds. Due to some constitutional issues involving federalism, the federal system can only prosecute a limited number of cases—it’s really hard to face a federal arson charge unless you, say, burn down the White House. As a result, about 50% of all federal inmates are in for drugs (one of the few kinds of offenses the Feds can easily target), compared to 15% in the states. So the emphasis on drug sentencing in federal reform makes a lot more sense and is a lot more important.

Moreover, the power of prosecutors in the state system differs significantly from that in the federal system. Local prosecutors are far more autonomous, raising all sorts of complicated issues about how to regulate their decisions. In the state system, the prosecutors are almost always independently elected at the county level, while U.S. Attorneys are at-will employees of the Attorney General, who is the at-will employee of the President, creating a (possibly) more effective chain of command. That said, AGs have struggled to ensure consistency across the various U.S. Attorney offices, so independence and discretion still permeate the federal system.

Moreover, unlike the states, the Feds tend to have much longer sentences available, and are much more likely to impose those sentences on people convicted of drug offenses. So the Standard Story holds a bit more in the federal system.

Which actually points to one possible risk with federal sentencing reform. Even though the Feds hold about 10% of the nation’s prisoners and vanishingly smaller shares of our jail detainees, parolees, and probationers, the federal system, and thus federal reform efforts, seem to receive about 95% of our media attention. This is one reason why the Standard Story has such traction: it fits the (distinctly idiosyncratic) Feds much more closely than the states. So while federal reform is worthwhile, the attention it receives runs the collateral risk of further convincing people that Standard Story-like reforms are appropriate for the states as well.

3. Many people think we could dramatically reduce incarceration and solve a host of other criminal justice problems simply by ending the War on Drugs—particularly because the trade in illegal drugs is intrinsically connected to violence. You think that’s too simplistic. Why?

To start, if we focus on just those who are in prison for drugs, it’s only 15% in the state systems. Were we to free every person in prison on a drug charge tomorrow, we’d still have 1.2 million people in prison and something close to, if not actually, the highest incarceration rate in the world.

Of course, that’s a narrow definition of the War on Drugs. Someone who commits murder in a drug deal gone bad is classified as serving time for “violence,” not “drugs,” even though the root cause of that murder was prohibition. So legalization would likely reduce some of those crimes, and thus some of those incarcerations. Sort of. Maybe. To some degree.

There are a few complications. To start, some, and perhaps many (though surely not all), of the homicides that arise from prohibition may occur in a world with legalization too, just for different reasons. Drug-related violence tends to occur in neighborhoods that are already under immense economic, and thus social and emotional, strain, which are all causes of violence; ending criminal prohibition may help, but if the underlying stress remains, so too will much of the violence. The nominal cause of the violence may shift (from a drug deal gone bad to some other dispute), but the levels may not change as much as many hope.

Moreover, legalization will lead to lower prices and less stigmatization, and thus more people using drugs (at least for some types drugs). For some, the cheaper drugs will mean they are more affordable, and will lead to less property crime to fund consumption. For others, the cheaper drugs will lead to increased use that further destabilizes the person’s life and may lead to more property crime. The effects are complicated and confusing and hard to fully predict in advance.

Will a shift to a legalized regime of some sort make things better? My guess, and I think that of most other people, is yes: that the net effect of legalization will be a net reduction in violent and property crime. But the dynamics will be complicated, and if we legalize without taking steps to address the public health issues of drug use and abuse as well as the other social pressures that lead to violence, the gains will likely be less than many people expect.

4. You are critical of the argument that private prisons drive increases in incarceration. Why? Incentives matter in political institutions. Isn’t it obvious that those who profit off incarceration would favor higher levels of incarceration? On a related note, didn’t an Obama Administration report conclude that private prisons are less well-run than public facilities?

To start, the data on the relative quality of public v. private prisons is quite thin, and people who study the issue rigorously note that what little data we have is wholly inconclusive. There are terrible private prisons and terrible public ones, and there are better private prisons and better public ones.

Here’s a confounding example. Until recently, the Florida Department of Corrections website proudly bragged that as a general matter its prisons did not have air conditioning—a feature that has led to inmate deaths from heat exposure in Florida and elsewhere. But that same webpage grudgingly acknowledged that all private prisons in Florida were fully air conditioned. It’s complicated.

More important, only 8% of all prisoners are held in private prisons, most of those in just five states, and there’s no evidence that trends in those five states differ from those elsewhere.

You’re right that incentives matter, but (1) public prisons often profit just like private ones do, and (2) public prisons create “profits” in all sorts of other ways besides rewarding shareholders. To start, as I mentioned above, we spend about $30 billion per year on wages for correctional officers in state-run prisons. That’s a strong incentive for correctional officer unions to fight reforms. For comparison, private prisons only make something around $400 million in profit.

But there is also political profit: prisons bring good jobs to depressed areas, for example, which provides votes for politicians who support the prisons and keep them full (and thus fully staffed). And in most states, politicians with prisons in their districts gain because those prisoners count as living in that prison for legislative districting, even though they cannot vote (which, because prisoners are disproportionately people of color from urban places and prisons are mostly located in more rural areas, provides a clear electoral benefit to the Republican Party). In the end, it seems likely that the availability of these sorts of public-sector benefits create incentives for public-sector resistance to reform that is far more significant and effective than private-sector resistance.

5. Your key argument has always been that the rise in incarceration has been caused by charging decisions made by prosecutors in recent decades. What kinds of reforms could address that problem?

We’ve started to see efforts, at least in urban counties, to elect reform-minded prosecutors, which I think is a good first step. But it is also important to note the limitations of elections-driven reforms. To start, at least in large urban counties, the offices are quite large, so reformist district attorneys may still struggle to get their message down to the rank-and-file prosecutors who make day-to-day decisions. Perhaps more concerningly, it is easy now to sound like a reformer without having to necessarily be one: there’s a set of terms that are easily bandied about—“bail reform!” “low-level drug offenses!”—but which are generic enough that they may not reflect much of a real desire for change, and that allow reform-sounding candidates to easily carve out exceptions that drown the rule once in office.

Given that it is increasingly easy to sound like a reformer without necessarily being one, I am glad to also see the rise of groups dedicated to making sure district attorneys’ actions match their words. In particular, we are starting to see more “court watching” groups. In New York City, for example, there’s a group called CourtwatchNYC, whose members sit in court daily all across the city publicly reporting on whether line prosecutors are upholding their bosses’ campaign promises. Twitter has also given public defenders a platform to report on whether actions track rhetoric as well.

I also favor the development of charging and plea bargain guidelines. This could curtail some of the abuses of discretion that prosecutors sometimes engage in, but more broadly it could address a problem with experience. We often call on newly-hired prosecutors fresh out of law school to make truly life-changing (for the defendant) decisions, like whether to set a reasonable bail. Guidelines could not only help minimize abuses, but they could simply provide prosecutors with a sense of what the optimal decision should be. So much of a prosecutor’s job is making complex risk assessments that law school did not train them to make; some sort of assistance is clearly necessary.

6. Didn’t higher levels of incarceration over the past twenty years correlate with a crime decline? This is the argument made by criminal justice reform skeptics like former Attorney General Sessions.

Sessions’ view is wrong, but it is important to understand precisely why. During the 1970s and 1980s, when crime was high and rising and prison populations were low and rising, prison growth probably did reduce crime; you can’t add 1.1 million people to prison and have no impact.

But that does not mean that prison was the right response to rising crime (although it may have been politically necessary, and it was certainly politically expedient). Prison worked, but other options—at the very least, increased policing—would have worked just as well at far lower social cost. If you get an infection in your finger and cut off your arm, you stop the spread of the infection. The amputation worked. But . . . maybe try an antibiotic first? Prison growth was a blunt tool with massive—and avoidable—collateral consequences.

And today? Sessions’ approach is even more invalid. It’s clear that in a time like now—with low crime and high levels of imprisonment—increased incarceration has close to no additional impact on crime. Evidence indicates that its deterrent effect is weak to negligible, and one impressive recent paper suggests that the more time someone spends in prison—the longer he is incapacitated and thus not committing crimes outside—the greater the risk of his reoffending upon release, to the point that the increased risk of reoffending almost wholly offsets the incapacitation effect.

7. You think prosecutors should be charging fewer people in ways that trigger prison sentences. You also argue, however, that most people in prison are there for violent acts. This seems like a contradiction. After all, isn’t it reasonable to expect that all people who have committed an act of violence serve at least some prison time?

Not necessarily. To start, it’s worth noting that lots of people who commit violent crimes don’t go to prison already: we arrest 500,000 people for serious violent crimes every year, but we admit fewer than 200,000 to prison annually for such offenses. My guess is that even more could be better served—solely from a public safety perspective—by options outside of prison.

Moreover, resources are limited, so a dollar spent on prisons is a dollar not spent on policing, or on non-police interventions like Cure Violence, or on drug treatment. And as I just mentioned, prison is a relatively ineffective way to combat crime. So we could achieve the same reduction in violence at far less social cost by relying on non-prison approaches.

Moreover, any analysis of the gains from prison has to wrestle with its costs—not just the $50 billion we spend to run the prisons, but the surely far vaster (but completely unmeasured) social costs: the risk of physical and sexual assault in prison, the loss of income upon release, the financial and emotional and social costs to the inmate’s family and loved ones, the way in which prisons spread diseases and increase the risk of drug overdose deaths, and so on. Trying to get a rough handle on these costs is my next big project, and my guess is that the number is going to be staggering.

8. Do you believe, like many criminal justice reformers, that there is a “crisis” in indigent defense? If so, what policies would fix it?

The crisis is this: about 80% of all defendants facing prison or jail time count as indigent and qualify for a state-provided defense lawyer, and we grossly underfund their defense. At least as of 2006-2007—the last years for which the Bureau of Justice Statistics and the American Bar Association have comprehensive data—we spent about $6 billion on prosecution and about $4.5 billion on indigent defense. But that significantly understates the difference. Prosecutors’ offices have access to all sorts of free services that defense lawyers have to pay for: police departments to do the investigation and forensics, state labs to test DNA and alleged drugs. A study in North Carolina, for example, found that while the nominal budgets for public defense and prosecution were roughly identical, the prosecutors’ budgets were essentially triple the defenders’ once all the free services that prosecutors (alone) received were accounted for.

The easiest solution would be some sort of federal assistance for indigent defense. $4.5 billion is a rounding error in the federal government’s $3.5 trillion budget—federal grants could easily double or triple what we spend on defense for the poor. And this is an issue that is increasingly getting bipartisan support. The support from the left has always been there (at least nominally), but a recent article in the Marshall Project pointed to growing conservative support motivated by Second Amendment concerns: poor people are losing gun rights due to inadequate representation. More broadly, though, I would think that indigent defense should be a core conservative value: what is more conservative than taking steps to ensure that the weakest among us are protected from that state at its most powerful?

And if Congress won’t act, then perhaps the Supreme Court can. The thing about Gideon v. Wainwright—the case that requires states to provide the poor with lawyers—is that it is the classic example of a Supreme Court unfunded mandate: the Court tells states to provide lawyers to the poor, but says nothing about how. A more robust, substantive take on what counts as adequate defense—not at the individual case level, but at the more systemic level—could push states in better directions as well.

9. What is the moral hazard problem in our criminal justice system? How was that problem solved by California’s Realignment? On a related note, has Realignment caused crime to increase as some people worried?

The moral hazard problem that concerns me (and others) arises from the baffling and poorly-reasoned (if reasoned at all) way that we fracture political and financial responsibility for crime control across city, county, and state (and, to a much lesser degree, federal) jurisdictions. One such example involves the political and financial costs that prosecutors face. In almost all states, prosecutors are elected by county voters and mostly funded by county budgets. Jails—where we detain people pre-trial and for low-level misdemeanor offenses—are county-funded as well, as are many probation services. But prison, the punishment for serious felony offenses? That’s funded at the state level.

Look at what that does. If the prosecutor is more lenient and seeks a misdemeanor charge, his home county has to pay for the jail term. But if he’s harsher, not only does he look tougher on crime—a move that, more often than not, is politically safer—but it is fiscally cheaper for him, since a different government picks up the tab. Even if prosecutors aren’t cynically taking conscious advantage of “free punishment,” the moral hazard problem means they will ignore the costs of sending people to prison, which is the classic definition of an externality or moral hazard problem.

In response to a judicial finding that its prison system was unconstitutionally dangerous and overcrowded, California adopted a sprawling, complex reform law colloquially called “Realignment,” which has been quite successful in scaling back California’s reliance on prisons—about 45% of the national decline in prison populations since 2010 is just California’s decline. One major reason for this success is that Realignment insisted that county jails, not state prisons, had to house people convicted of a wide range of felonies, not just misdemeanors. In other words, it pushed the costs back onto the counties, and the counties balked. This is not the only reason for California’s decline—like I said, Realignment is complex—but it certainly appears to have mattered.

As for the rising crime some feared (and which some still tout), there is simply no evidence that it has actually happened. One study found a slight increase in auto theft, but a later study identified the same increase . . . and then found that it had subsequently disappeared.

The people pushing the “rising crime” story are those most politically opposed to reform: the police, sheriffs, and prosecutors. One of my biggest hopes about the current reform push is that we come to view these criminal justice actors as the political actors that they are—that we stop accepting their claims at face value and subject them to the same scrutiny and skepticism that we apply to those on the reform side.

10. Why does Chicago struggle with violent crime in a way that New York City does not? New York has done several things which reform opponents have worried will lead to higher crime—incarcerating fewer people, eliminating stop-and-frisk, etc.—but it continues to be one of the safest big cities in the country. Why can’t Chicago just be New York?

I think that’s a tough question for which no one yet has a good answer. I’d point out just three things, though. First, the focus on Chicago is peculiar. While its homicide rate is still higher than it was in the past, even at its recent peak its rate was lower than in other cities that receive no attention at all. And over the past two years, its rate has dropped significantly, yet it still remains our go-to example of “rising homicide.” Chicago is being singled out for reasons that are, sadly if predictably, unrelated to crime trends.

Second, it is worth noting that the year homicides spiked, about half the citywide increase in homicides occurred in just five neighborhoods that were home to less than 10% of the city’s population. So the story in Chicago—and the story of crime in cities nationwide—is significantly local, even within the city, and thus defies any sort of easy “problem in Chicago” answer.

And third, while the causes of Chicago’s homicide spike, and broader gun violence spike, are complicated, one factor that certainly played some role was the state’s decision to defund Chicago’s Cure Violence program (called CeaseFire in Chicago), a street-level violence intervention program that relies on local community members to try to intervene to prevent retaliations after a shooting. Shootings and homicides in Chicago had been declining prior to the defunding and rose shortly after the state cut funds. More interestingly, one neighborhood’s program was spared the cuts, and that neighborhood did not see shootings or homicides rise, even as they rose in other areas whose programs had been cut.

It’s important not to oversell this story: the timing isn’t a perfect fit, and trends in complex social problems like homicide rarely if ever have a single neat explanation. But it seems likely that CeaseFire was playing a real role in Chicago, which suggests that we need not unleash on Chicago a heavy-handed police response.

 

 

Refreshing Candor, Useful Data, and a Dog’s Breakfast of Proposals: A Review of Locked In by John Pfaff

by Kent Scheidegger

John Pfaff gives us two books under one cover in Locked In: The True Causes of Mass Incarceration and How to Achieve Real Reform.[1] In the first book, he tells us that nearly everything we have been told about so-called mass incarceration by his fellow “reform” advocates is false. His candor is a breath of fresh air. He convincingly makes the case with a mound of useful data.

The second book, in contrast, is thinly supported and heavily influenced by Pfaff’s predispositions. He tells us that high incarceration rates are caused primarily by overcharging prosecutors, though his data do not rule out alternative hypotheses. He claims that the election of tough prosecutors is caused by the “low-information, high salience electorate,” not by informed people who genuinely and justifiably disagree with him on priorities. The primary ingredients in his stew of solutions are tools to save the ignorant masses from themselves by making our society less democratic and our criminal justice decision-makers less responsible to the people. Other intriguing possibilities raised by his data go unexplored.

Pfaff does not define what he means by “reform,” but he appears to use that term for policies that have the single-minded purpose of reducing the number of people incarcerated. Obviously, that is not the sole or universally accepted meaning of the term in criminal justice. The Sentencing Reform Act of 1984[2] definitely did not have that purpose. In this review, I will put the word “reform” in quotation marks when used in Pfaff’s sense.

I. The (False) Standard Story

In the first half of the book, Pfaff describes what he calls the Standard Story and proceeds to demolish it. The Standard Story is the one that nearly all advocates of “reform” use to sell their proposals to legislatures, courts, and the public. The Standard Story’s central premise is that America’s high incarceration rate is caused by sentencing harmless, low-level, nonviolent offenders—especially those whose only crime was possession of illegal drugs—to long prison terms.

A. Causes of High Incarceration Rates

In a well-publicized book provocatively titled The New Jim Crow: Mass Incarceration in the Age of Colorblindness, Michelle Alexander claims that drug convictions account for the majority of the increase in prison population. President Obama picked up the theme in a speech to the NAACP, declaring that locking up nonviolent drug offenders is “the real reason our prison population is so high.”[3]

But it’s not true, as Pfaff demonstrates. Most of America’s prison population is in state prison, not federal. While the number of state prisoners convicted of drug offenses did indeed increase sharply from 1980 to 2010, the number is still not enough to be the primary cause of the increase. Far more people are in prison for violent crimes than are in for drug crimes. Of those who are in prison on drug charges, only a small portion are genuinely low-level offenders. In addition, people who have committed both drug and violent crimes are often convicted only of the drug crimes because those are easier to prove. The notion that our prisons are full of low-level, nonviolent drug offenders is a myth. Decriminalizing or even legalizing drugs would produce only a modest drop in prison population.

So if we are imprisoning mainly violent offenders, and the prison population is exploding, then one might think excessively long sentences are the problem. Many people do. But this, too, is wrong, Pfaff says. Despite some well-known anecdotal examples, actual time in prison for the typical offender is much shorter than most people think. For example, the median armed robber convicted in 2010 was released in less than three years.[4] Even nominal life sentences do not generally result in offenders actually spending their whole lives in prison.

The real reason the prison population has grown and remained high despite falling crime rates, Pfaff contends, is that a growing percentage of those arrested are being charged with felonies rather than misdemeanors. Crime rates, clearance rates (percentage of crimes solved), and arrest rates all fell between 1994 and 2008, but the number of felony cases filed rose substantially. The chance of a filed case resulting in prison time has remained stable.

The percentage of arrests resulting in felony charges is therefore the only factor that increased in the system. Breaking the numbers down by offense, it appears that this increase is mostly for violent crimes. The prison admission/arrest ratio rose substantially between 1991 and 2011 for murder/manslaughter, robbery, and aggravated assault. It rose less for burglary, and not at all for theft and drug possession. The ratio for drug trafficking rose somewhat, but then declined slightly.[5]

Pfaff notes, correctly, that the data do not tell us the reason for the increase in admissions per arrest for violent felonies. “Maybe police are doing a better job investigating certain types of crimes, maybe prosecutors are being more aggressive in charging them, maybe judges are more willing (or compelled) to send defendants convicted of them to prison.”[6] Yet having set out three plausible hypotheses, Pfaff spends most of the rest of book focused on only one of the three: aggressive prosecutors. The police and judges get only passing mention.

The fact that the admission/arrest ratio for murder and manslaughter increased alongside the ratios for other violent offenses suggests that the quality of arrests coming to the prosecutors from the police may indeed be a major factor. These numbers are primarily for murder and voluntary manslaughter; involuntary manslaughter cases constitute only a small fraction.[7] It would be a rare case of intentional homicide where prison was not the appropriate disposition. If only half of homicide arrests in 1991 resulted in a prison admission, the most likely reason is that a large portion of the cases had shaky evidence, resulting in a dismissal or a generous plea bargain; more arrests today might indicate that police are doing a better job of collecting evidence to support their arrests. Yet Pfaff spends little time on this intriguing hypothesis.

Pfaff spends half a chapter debunking the notion that lobbying by private prison interests is the cause of high levels of incarceration. He notes that the power of the commercial corrections lobby “is overhyped at every turn.”[8] There is no need to go into that in any depth here. Private prisons are such a small portion of the total—about 8%—that it would take anti-capitalist tunnel vision to consider such an argument remotely plausible.

B. Costs and Benefits

The discussion of public finances is more salient. The period of expanding prison population was a period of economic growth and ballooning government budgets. While corrections spending did rise in overall dollars, its rise as a percentage of state and local budgets was modest during the rising crime years. With a lag of a few years, the percentage fell after crime rates fell.

The notion that we are spending huge amounts on prison and that this spending is crowding out other priorities such as education, health care, and transportation is simply wrong. Pfaff notes, “we don’t really spend that much on corrections,” about 3% of total state budgets.[9] This is heresy among many self-styled reformers. It is also true.[10]

The main purpose of this line of argument is political. It serves to win over fiscally conservative voters who would otherwise be inclined to support tough sentencing, and it provides cover for conservative politicians who might actually want to support softer policies for other reasons.[11] Politically, then, just as the prosperity of the 1990s and most of the 2000s enabled growth of prison populations, so the fiscal crisis of 2008 and the pinch on government budgets provided a window for a facially plausible argument that we could save money and relieve the pinch by incarcerating fewer criminals. This argument has struck a chord with many moderate and conservative voters.

Yet Pfaff worries that the window may be closing with renewed prosperity. Although he doesn’t say so, the window might also close when the false claims of the Standard Story come to light. Pfaff warns his fellow “reformers” that focusing on factors that are not really major contributors to the prison population risks postponing the “reforms” that actually would produce a major reduction in prison population (i.e., releasing more violent criminals) until after the window has closed.

The “reforms” undertaken so far—mostly focused on nonviolent criminals and largely excluding those convicted of violent crimes—have yielded far less fruit than is commonly believed, Pfaff contends. The much-heralded legislation in Mississippi, for example, merely brought that state’s incarceration rate down from a level far higher than the national average to a level substantially higher than the national average. After the initial drop, it leveled off. Pfaff says “short-run declines followed by stasis” is the typical result of reform legislation.[12]

The national prison population fell four percent from 2010 to 2014, but two-thirds of that drop was in California alone due to developments there that Pfaff calls “highly idiosyncratic.” For the rest of the country, the drop has been modest. Moreover, the drop has not been solely due to “reforms,” perhaps not even primarily so. Pfaff provides a graph of annual changes in prison population growth,[13] and it shows a point of inflection in the mid-1990s, shortly after the peak in the crime rate and long before any significant “reform” legislation. Prison population grows more slowly after that, with the rate of change finally going negative in 2010. The reductions since 2010, Pfaff says, are “not entirely attributable to reforms, but are partially the continuation of a preexisting trend tied to falling crime.”[14]

The Standard Story has focused on nonviolent criminals, Pfaff notes, because “reform” advocates believe that is the limit of what is presently politically feasible. He acknowledges that is likely true, but he has a blind spot that lets him see only half of why it is true. He notes, correctly, that “reforms” cutting imprisonment for violent criminals would risk increasing violent crime, which would undercut political support for the whole effort. This is true, but not the whole truth. The other half is simple justice, a concept that is nearly absent from this book.

Imagine you are the victim of a brutal rape, and a year later you sit down at a restaurant and see the rapist at the next table, happily enjoying his meal and laughing with his friends. Even if we had a magic pill that 100% guaranteed he would never commit another sexual assault (we don’t), it would still be horribly wrong for him to get off that lightly for such an evil act. As discussed earlier, Pfaff’s own numbers show that sentences for violent crimes are generally not greater than the perpetrators deserve. People value justice for its own sake, irrespective of the practical consequences of punishment.[15] That is an independent reason for punishing serious crime, and it is an independent reason why any effort to reduce punishment for violent crimes faces stiff headwinds. Pfaff seems to be nearly oblivious to it. In a footnote near the end of the book, he finally acknowledges the issue, but he merely says it is “beyond the scope of this book.”[16] That is a strange statement. The book is all about the punishment of crime, and the reasons why people vote for “tough on crime” prosecutors and legislators is a major theme. Yet one of the primary reasons they do so—because they think violent criminals deserve tough punishments[17]— is “beyond the scope”?

On the purely utilitarian side, the costs and benefits of incarceration are complex, difficult to measure, and partly intangible. The most obvious benefit, of course, is reduced crime. Here, Pfaff is again more candid than many “reformers.” He acknowledges that statistical problems cause many studies to underestimate the utility of imprisonment in reducing crime. The stronger studies tend to show a greater effect.[18] Steven Levitt’s estimate that high incarceration rates produced a quarter of the great crime drop of the 1990s is consistent with other work.[19] That is a massive reduction in victimization and suffering. Pfaff asserts that the same decline could have been produced “by investing in other, less costly, less brute-force solutions.” Later in the chapter he suggests that hiring more police officers is one such solution.

Pfaff offers empirical support for the thesis that there are diminishing returns to incarceration. If that increased incarceration is caused by harsher sentencing of less serious offenders, it does make sense that there would be diminishing returns. It does follow, as he says, that when incarceration rates are high due to harsh sentencing, the rate can be brought down without a large increase in crime. Yet that is the “low-hanging fruit” that has already been picked, i.e., the “low-level offenders” who never did make up a large portion of the prison population.

Pfaff acknowledges that further cuts may reach a point where they cause increased crime, undercutting political support for further cuts. But because crime rates are affected by a great many factors and many of the other factors point to continued declines, Pfaff believes that we have a lot further to go before sentencing strictness drops below politically acceptable levels. However, California may have already reached that point. The 2020 ballot will test that possibility with an initiative to partially roll back some of the state’s recent “reform” measures.[20]

As difficult as the crime-prison connection is to gauge, other elements of cost-benefit analysis related to possible changes in policy are even more nebulous, with little or no hard data. An absence of empirical evidence provides greater leeway to indulge one’s preconceived notions. Pfaff speculates about the benefit to 100,000 children who presently have one parent in prison instead having both parents at home. Wouldn’t releasing their parents be worth a five percent increase in aggravated assaults, and even some increase in the number of murders?[21] Isn’t this a debate worth having?, he asks.

We can debate it, but I think the answer is quite obviously no. On the benefit side, Pfaff simply assumes that the incarcerated parents would remain in their children’s lives and that their influence would be positive. Yet he established earlier in the book that the notion of large numbers of low-level, nonviolent offenders is a myth, so we are talking mostly about parents who have committed serious crimes of violence. Pfaff offers no support at all for his assumption that these violent men (mostly) would be good fathers. I think of Huckleberry Finn. Huck suffered from the lack of a caring, nurturing, supportive parent, but he suffered a lot more when the drunken, abusive father he actually had showed up. For the dubious benefit of 100,000 criminal parents returning to homes where they may do more harm than good, Pfaff would accept about 40,000 people being attacked and severely injured, and some unstated number being killed.[22] In my view, it is not a close question.

Pfaff notes as costs of incarceration the violence within prison, the loss of employment skills, and barriers to employment after release. These are inevitable costs to some extent, but not to the extent that they exist today. More effective prison discipline, better smuggling control, and greater employment of inmates (including unrestricted sale of prison-made goods in market segments with no substantial domestic production) are all matters that warrant action but will require overcoming political opposition. Employer reluctance to hire is more a consequence of the conviction than the sentence, but critical examination of government licensing requirements and encouragement of voluntary hiring of released prisoners in appropriate jobs are also worthwhile endeavors.

Pfaff ends the first half of the book with the conclusion that current “reform” policies are not reducing prison population as quickly as the “reformers” had hoped they would. He considers “faith in the Standard Story” to be one reason for this. Was it really faith, or was the Standard Story a deliberate fraud from the beginning? On the other side, he claims it is “fear”—not genuine, well-founded concern for the innocent victims of crime—that motivates opposition to softer sentencing practices. The second half of the book is devoted to ways to sharply reduce incarceration.

II. A (Dubious) New Narrative

The second half of the book is markedly more political and ideological and less data-driven than the first half. The overall thrust is that American criminal justice is suffering from too much democracy, and that the path to deeper “reform” is to take justice decisions away from the Great Unwashed and assign them to wise Philosopher Kings.

A. Supposedly Harsh Prosecutors

Pfaff argues that prosecutors have much greater power to determine a defendant’s sentence than is generally understood, and that their power is excessive. The argument is thinly supported. Pfaff points to mandatory minimums and uses the federal system as an example.[23] While debunking the Standard Story, he correctly noted that most prisoners are in state, not federal, prisons, and that the federal system is not typical of the states. But in his indictment of prosecutors, he commits the very error he had critiqued. To support the proposition that in “many states” the prosecutor’s charge can restrict the judge to a narrow sentencing range,[24] he gives us no hard data but only a general observation that “[s]everal states use sentencing guidelines . . . .” How many? How rigid are the guidelines? Do they allow departures? He doesn’t say.

Pfaff concedes that in his home state of New York the prosecutor’s charge does not set a sentencing floor.[25] That is also generally true in California, where the judge has broad power to dismiss allegations that would otherwise set a minimum punishment, except in the few instances where that power has been revoked by statute for a particular charge.[26] Even the much-criticized Three Strikes law permits judges to “strike strikes” to avoid the law’s mandatory sentence, including over the prosecutor’s objection.[27] If prosecutors’ use of their charging discretion to lock judges into unjustly harsh sentences is really a major national problem, it should not be difficult to demonstrate that sentencing laws in states comprising the bulk of the population give them that power, but no such support is offered.

Direct evidence of prosecutorial harshness, Pfaff notes, is unavailable due to a lack of data focused specifically on charging decisions.[28] He goes on to speculate on reasons, yet some of the reasons he cites are not harshness at all, as most people would understand the term. One possible reason for prosecutors seeking prison terms for a greater percentage of felons is that more defendants today have long criminal histories than in the past. The peak crime years of the 1980s and 1990s produced a lot of long records. “America is the land of the second chance,” President Bush famously proclaimed in his 2004 State of the Union Address.[29] He did not say that America is the land of the seventh chance. To the extent that prosecutors are seeking more severe punishments for habitual criminals than they do for first-time offenders, they are not acting harshly. They are doing their jobs exactly right. Yet Pfaff calls it “harsher.”[30]

Another possibility, noted in the first part of the book, is that improved policing may be creating stronger cases against felons.[31] Perhaps criminals were previously getting off with unjustly lenient sentences or walking altogether due to inadequate evidence, and with better police work and technology such as DNA and widespread video cameras they are now getting the punishment they deserve. If so, and if we care about justice, the change should be applauded, not lamented. That second “if,” however, seems to be one of Pfaff’s ideological blind spots. If he does care about sentences being just punishment for the crimes, he doesn’t show it in this book.

B. Too Much Democracy

In most states, prosecutors are elected locally, either by county or, in rural areas, by judicial districts with more than one county. Pfaff says the record on election of prosecutors is “bleak,” because nearly all are reelected and most are unopposed.[32]

It is true that voters often know little about their local prosecutor, but that is not necessarily a problem. Absence of local news about the prosecutor’s office, except in cases of exceptional crimes, indicates a lack of controversy. Lack of controversy, in turn, indicates that the district attorney is conducting the office in a way that people would approve of. There is no shortage of lawyers, and even some judges, who would like to take that office for themselves if they thought they had a shot. The fact that incumbents are rarely challenged indicates that potential challengers do not think voters would find the incumbent’s policies objectionable if they were brought to their attention in a campaign. The fact that most incumbents are reelected when they are challenged tends to confirm that judgment.

What is bleak about that? Elected officials who pursue policies consistent with the people’s view of justice is exactly how democracy is supposed to work.

In the federal system, the people elect only the President, and all other executive policy-making officials, including U.S. Attorneys, are appointed by and can be fired by him. Is this unitary executive of the federal government better than the distributed executive of separately elected officials, including local prosecutors, that we have in most states? Pfaff thinks so,[33] even though federal prosecutors have been widely criticized by others for being overly aggressive.[34] The people of the states, as they have written and rewritten constitutions in the last 240 years, have chosen not to follow the federal model. They have instead made more officials directly answerable to the people. Should we cast that collective wisdom aside because local prosecutors nationwide are exercising discretion in a way that “reformers” disagree with?

Pfaff laments the fact that locally elected prosecutors are not constrained by the cost of imprisonment when they seek prison sentences because prison costs are paid at the state level. He considers this to be a “moral hazard.”[35] Most people, I suspect, would consider determining punishments according to justice and not dollars to be a valuable feature of the system, not a problem to be fixed. Do we actually want a world where a prosecutor tells the victim, “I am only going to ask for two years in prison for the man who raped you because we can’t afford any more”? Do we want a person who was caught burglarizing a home and found in possession of stolen property from fifty unsolved home burglaries to get probation because the jurisdiction wants to cut prison costs? How many more people’s homes would be invaded as a result?

Pfaff notes with approval California’s “realignment” experiment, which places lower-tier felons in county jail rather than state prison.[36] However, this change was not aimed at changing prosecutors’ charging decisions. It has produced a one-time drop in overall incarceration,[37] to be sure, but Pfaff cites no evidence that changes in charging are the cause. He does cite a study by a left-leaning think tank claiming that realignment produced only a small increase in property crime “compared to what it otherwise would have been.”[38] However, the study says, “we find robust evidence that realignment is related to increased property crime”; considering the source, that is quite an admission. Overall, California’s property crime rate ran about 9% below the national average in the years leading up to realignment and has run about 5% above the national average since, a 14% jump in relative rates.[39] California is hardly a model for the rest of the country to emulate.

The root problem, in Pfaff’s view, is voters he regards as ignorant and inconvenient. He identifies several “defects” in the politics of crime. First, voters are more likely to punish a prosecutor or a judge at the polls for a “false negative”—a failure to sufficiently punish a person who should have been confined and subsequently commits a major crime—than for a “false positive”—an excessive sentence of a person who would not have committed a new crime even if he had not been incarcerated.

Pfaff supports his argument about the “false positive defect” with an atrocious misuse of the old maxim that it is better for ten guilty men to go free than for one innocent one to be convicted.[40] He blithely assumes that the calibration of punishment for a person who is truly guilty belongs in the same tier of justice as the determination of whether he is guilty or innocent. While one might conceive of hypotheticals where overpunishment amounts to an injustice of the same magnitude as sending an innocent man to prison, there are no such examples in American law today.[41]

A more realistic example illustrates that the 10-to-1 maxim may work the other way around when it comes to sentencing. Suppose ten people are convicted of rape, and the question is whether to sentence them for three or six years.[42] Let us further assume—generously—that nine of the ten would be law-abiding in years four to six, but one would commit another rape. Is it better that ten rapists serve an extra three years in prison for evil acts they chose to commit than that one innocent person be raped? It is in my book.

Incarcerating a convict for a period in which he would not have committed a crime is not a false positive because incapacitation is not the sole reason for imprisonment. Justice is a sufficient reason by itself. In my example, we should not wring our hands for the rapists who do six years because that is not greater than their just deserts for their crime, whether there is any utilitarian benefit to the additional time or not. The sentences Pfaff regards as false positives may well be sentences that the typical voter regards as just. Perhaps district attorneys who seek sentences for burglars that result in their serving a year and two months, on average,[43] usually have no opponent, not because the people are ignorant of false positives, but because anyone who ran on a platform that such sentences are unjustly long would be greeted with derision and lose in a landslide in most jurisdictions.

The complement to the false positive problem is the idea that voters pay excessive attention to rare but extreme cases. The problem here, according to Pfaff, is “low-information, high-salience” voters who only know about the headline-making cases, not how the system normally works. Pfaff’s example of this phenomenon is revealing: the Willie Horton case. Pfaff recites the version of the Willie Horton fable that casts former governor and presidential candidate Michael Dukakis as the victim of a misleading smear. In this telling, Dukakis had simply inherited a prison furlough program from his predecessor. It was largely successful, but one isolated failure caused a public relations disaster. Horton absconded from the program and committed a brutal home invasion, aggravated assault, and repeated rapes. The Bush campaign, Pfaff says, “released a powerful attack ad with racist overtones . . . using the Horton case to argue that Dukakis was soft on crime.”[44] Conspicuously absent in this telling is any mention of what crime Horton was in for when he was furloughed.

Horton was in for murder and ineligible for parole; a rational furlough program would not have included people with such sentences.[45] True, Dukakis had inherited the program from his predecessor, but it was a court decision construing the statute that made murderers eligible, and Dukakis vetoed a bill to tighten up the system. He continued to resist even after Horton’s crime. Far from “racist overtones,” the Bush campaign’s ad refrained from using Horton’s picture or mentioning his race.[46] The ad with the picture was produced by independent operatives, and the contemporary criticism of Mr. Bush was merely that it took his campaign too long to express disapproval.[47]

In short, the Bush campaign’s attack on Dukakis’s furlough program was a valid campaign challenge to an astonishingly ill-considered policy, one which had tragic consequences for innocent people. The fact that Pfaff uses the stock story without fact-checking is typical of the mindset that pervades this book. All who support “reform” get a pass, even when spreading or accepting disinformation, and all who oppose it get disparaged on thin or false evidence. Pfaff uses the Horton ad myth to brand voters who disagree with him as “low-information, high-salience” (LIHS) voters. What about the voters who agree with Pfaff on “reform” because they bought the Standard Story after hearing about rare, atypical long sentences for relatively minor crimes? The LIHS label would apply just as much to them, but they do not receive it in this book.

Public reaction to atypical crimes (and sentences) is a problem to be sure, but Pfaff fails to make the case that this is a major reason why supposedly harsh prosecutors are so routinely reelected. The alternative hypothesis—that people are generally satisfied with their elected prosecutors’ overall charging practices—is not explored.

Comparing the United States and Europe, Pfaff notes that politicians generally are more sensitive to voters’ wishes in this country because our government is much more decentralized. Indeed, our federal system, separately elected legislative and executive branches, and variety of independently elected officers in state and local government produce a degree of accountability that is unique in the world. Pfaff recounts this aspect of American exceptionalism as if it were a bad thing. It produces a result he disagrees with, to be sure. Yet this extensive separation of powers was deliberately designed to preserve freedom from the dangers of excessively concentrated power. We should be very careful about tinkering with it just to change a result on one policy question.

C. The Third Rail: Releasing Violent Criminals

With the foundation built to this point, Pfaff advances his main theme. “If we are serious about wanting to scale back incarceration, we need to start cutting back on locking up people for violent crimes.”[48] He appropriately calls this the “third rail.”

The third rail of an electric train system is the one that carries the electric power. It delivers high currents at high voltage, and therefore it would be instant death to touch it, at least if one does so while connected to ground. It has become a widespread metaphor for a political issue so dangerous to one’s reelection chances that no politician wants to touch it.[49] In an attempt to lower the voltage, Pfaff argues that we do not need to incarcerate violent criminals to the extent we do for the purposes of incapacitation or deterrence; he simply refuses to discuss justice.[50] Indeed, he goes so far as to assert that less punishment of violent crimes will actually make us safer.[51] Unlike his thorough, data-grounded demolition of the Standard Story, however, the evidence cited here is off-target.

Pfaff notes that most offenders follow a life course in which they are more likely to commit crimes while relatively young and less likely to do so in their older years. He acknowledges that some people do not follow this course and do not desist.[52] Therefore, he claims, “long sentences frequently over-incapacitate. We don’t need to lock up most violent twenty-year-olds for thirty years to keep ourselves safe, since most of them would naturally desist from offending much sooner than that.”[53] Yet Pfaff’s own demolition of the Standard Story shows that this argument is off-target. We don’t need to lock up most twenty-something violent offenders for that long, and indeed we are not doing so at present. Just four pages earlier, Pfaff noted that violent offenders admitted in 2003 served an average of only 3.2 years.[54] Pfaff points to data on people released after California revised its Three Strikes law for the proposition that this group committed fewer crimes after release “because they are simply older.”[55] Indeed, the one treatment that we know works to decrease violent propensities is aging.[56] So while we may not need to lock up a violent 24-year-old until he is 54, locking him up until he is 34 may indeed be quite effective in preventing a number of crimes of violence and sparing multiple innocent people from victimization.

Pfaff’s extended attack on long sentences is very odd in light of the fact that he so thoroughly documented in Chapter 2 that long sentences are not the cause of mass incarceration. Although very long terms make headlines, there are too few of them to be a major factor in the incarceration rate.[57] A term of years in single digits is not an unjustly long one for a violent felony, and Pfaff fails to make the case that such terms are ineffective as incapacitation.

Pfaff also attacks the deterrence rationale for long sentences,[58] and this attack is similar in that Pfaff again misses his own target—the rate of prison admissions—and instead hits the one he has shown is not important in the large scheme of things: very long sentences. He begins by noting that certainty of punishment is more important for deterrence than severity of punishment. That is true, but no one, to my knowledge, is arguing to the contrary, or has made such an argument for a very long time.[59] The question is not whether severity has a greater effect than certainty, but simply whether severity has an effect that makes a significant contribution that is worth its cost. “Empirically,” Pfaff notes correctly, “it is hard to separate out the deterrent impact of longer sentences from their incapacitative effect . . . .”[60] Yet he curiously fails to mention a well-known study by a well-known economist whom he cites elsewhere in the book, Steven Levitt.[61] Levitt and Daniel Kessler used the changes in crime in the wake of the enactment of sentence enhancements in California’s 1982 Proposition 8 to make the separation between deterrent and incapacitative effects. Immediately after the law’s enactment, potential criminals would be facing increased time but would not have actually served any of the additional time. Kessler and Levitt found that there was indeed a significant drop in crime that cannot be attributed to incapacitation.[62]

This is a critically important point of the argument: whether reducing sentences for violent crime would cause increased victimization of innocent people through loss of deterrence. Yet Pfaff cites only a single article by Daniel Nagin for the proposition that there is “little or no evidence that long sentences have any real deterrent effect.”[63] What Nagin actually says is that “there is little evidence that increases in the length of already long prison sentences yield general deterrent effects that are sufficiently large to justify their social and economic costs.”[64] If we assume for the moment that Nagin is correct, this supports only the proposition that laws like Three Strikes, which impose life sentences on felons who would otherwise receive, say, ten years, do not have enough deterrent effect to justify the cost of imprisonment. But Three Strikes was always about incapacitation, not deterrence, so this proves nothing at issue. It does not negate Kessler and Levitt’s result that an additional five-year term imposed on a repeat robber or burglar who would otherwise be getting several years does have a deterrent effect.

Earlier in the book, Pfaff maintained that the primary engine of prison growth was prosecutors seeking prison time for more of those who committed violent felonies, but that those convicted did not serve excessively long terms. When he turns to his critique of criminal sentencing, though, he tries and fails to refute the proposition that long sentences have a deterrent effect and save innocent people from victimization.

When we examine this chapter critically, what we see is that releasing violent felons is a political third rail for good reason. Some excesses, such as the original version of California’s Three Strikes, can be pruned back, but Pfaff fails to demonstrate that overall the status quo of prison sentences for violent crimes is not needed for incapacitation and deterrence, as well as just plain justice.

D. A Dog’s Breakfast of Proposals

After following this long road of statistics and studies, where do we go next? The end of the book is an untidy and unsatisfying scattering of proposals, a dog’s breakfast.[65]

1. Guidelines

To deal with the perceived problem of too much democracy, Pfaff advocates adoption of sentencing guidelines.[66] Continuing with the third rail metaphor, one might say that sentencing commissions are offered as an insulated glove. Politicians can touch the third rail through the insulation of a law that creates a sentencing commission rather than directly reducing sentences, and the unelected commission does the unpopular work of releasing large numbers of violent felons.

However, whether a guidelines system increases or reduces sentences depends on how the system is designed, how the commission operates, and who is on it. Pfaff makes clear that whether a guidelines system is good or bad depends, in his view, entirely on whether it achieves the result he considers desirable: reduced sentences.[67] “It comes down to design,” he says, but he does not tell us anything about the design other than that the sentences should be shorter.[68]

2. Increased Parole for Violent Felons

Increased use of parole is also a possibility, but Pfaff notes that nearly all proposals to date have been limited to prisoners designated nonviolent, meaning those whose current offense of commitment is for a crime labeled “nonviolent.”[69] To expand parole to more violent felons while insulating legislators and parole officials from the third rail of a Horton-type incident, Pfaff advocates the use of “quantitative risk-assessment tools”—parole by computer algorithm.[70] He deems the victimization of additional innocent people to be an acceptable cost of paroling more violent criminals, but he wants to insulate the people who make that value judgment from the political consequences.

Recent enactments suggest that most politicians do not believe that a program that forthrightly embraces greater parole opportunities for violent felons is politically viable. California’s Proposition 57 was a proposal so soft on crime that Governor Jerry Brown did not think he could get it through even that state’s very soft-on-crime legislature, so he resorted to an initiative, which passed in 2016. He marketed it to the people as limited to inmates “with nonviolent convictions” and with a promise that it “[k]eeps the most dangerous offenders locked up.”[71] That marketing was dishonest in two ways. First, the state’s statutory definition of “violent felony” was written for a limited purpose and leaves out a great many crimes that most people would consider violent, including rape of an unconscious person, assault with a deadly weapon, and drive-by shooting.[72] Second, inmates with a present conviction for a nonviolent felony but a slew of violent priors are eligible. The fact that savvy politicians intent on softening criminal punishment considered the deception necessary indicates that a proposal expressly including violent felons would be unlikely to pass.

A further step in the direction Pfaff proposed, enacted after publication of his book, is the federal First Step Act, which provides a system of credits for early release from federal prison. That act has a list of 52 exclusions from eligibility for its credits program.[73] It was also marketed as excluding “violent and high-risk criminals” from the time credits program,[74] though the list, long as it is, is not comprehensive. The act also requires for eligibility a classification by the Bureau of Prisons as “low-risk,” so its actual application to felons convicted of violent crimes remains to be seen.

Neither of these programs fully embraces Pfaff’s proposal of expanded parole for violent felons, and both were marketed as rejecting it. Even with the cloaking device of parole-by-algorithm, it seems unlikely that his proposal will be able to get much traction. Indeed, when, inevitably, violent felons are released under the programs already enacted and commit more serious violent crimes, it seems more likely that support for further expansions will deflate. Pfaff recognizes the danger that the people will turn against his proposal once they see its actual results, but he hopes that the difficulty of repealing legislation will insulate it from the backlash.[75]

3. Changing the “Reform” Conversation

Pfaff is concerned that marketing “reforms” for nonviolent felons as freeing up space for violent ones will inhibit future expansion to the violent ones.[76] He asks his fellow “reformers” to stop making that sales pitch. But they must say those things to get even limited “reforms” enacted. Changing the conversation in the way he suggests might result in the “reform” movement grinding to a halt.

4. Requiring Cost-Benefit Analysis

Pfaff acknowledges that the non-incarceration alternative punishments he favors require greater initial cost outlay than imprisonment, so to make this go down easier he proposes use of broader cost-benefit analysis. One problem with that proposal is that it requires quantifying the benefit of the supposedly more effective alternatives to imprisonment. It is doubtful that many “reformers” actually want rigorous quantitative assessment of their alternatives. The last thing they want is a repeat of the debacle of the 1970s. At that time, a study of studies found that many studies of the effectiveness of rehabilitation programs lacked the basic elements of methodological validity.[77] Evaluating the studies that could be considered acceptable, there was virtually no evidence that any of the programs had a significant impact on recidivism rates.[78] It is probably for this reason that the First Step Act defines “evidence-based recidivism reduction program” without any requirement of methodological rigor whatever, throwing open the door to “junk science.”[79]

I would be delighted to see cost-benefit analysis done right. That means the benefits of alternative programs really would be determined by strictly valid methods. It would also mean that the costs of crime are fully accounted for, not just direct, tangible losses. I’m not holding my breath for either of these.

5. Regulating the Prosecutor

The biggest set of proposals has to do with prosecutors because of Pfaff’s dubious assertion that prosecutors are the “main engines driving mass incarceration.”[80] From this premise, he considers regulating them to be a top priority.

Pfaff’s first proposal is to beef up funding for public defenders so that they can “regulate” prosecutors by doing a better job of opposing the prosecution.[81] This is one proposal that most prosecutors, along with nearly everyone in the criminal justice system, would agree with. Prosecutors uniformly tell me that they prefer that the defendant be well represented. The opposition to better funding for public defenders comes from those seeking the same government dollars, so perhaps “reformers” and advocates of more effective law enforcement could make common cause in seeking a larger slice of the budget pie for criminal justice generally.[82] The notion that better indigent defense would actually have a significant impact on imprisonment rates, though, is not supported by any real data.[83] Most cases end in plea bargains. Would more money for defendants mean bargains with shorter sentences on average? Maybe a little, but the idea that it would make a substantial difference, and that the difference would be a good thing, requires firmer support than it receives here.

Pfaff also wants grant programs to “encourage prosecutors to focus on more serious offenses.”[84] Does he think that they are not focusing on serious offenses now and that money is the reason? He offers only one example of one odd grant program from the previous century. Most prosecutors are motivated by justice, and justice cries most loudly in the most serious cases.[85] Pfaff argues that higher murder clearance rates in richer than in poorer portions of Los Angeles County result from prosecutors focusing more on solving crimes where rich people are victimized, but he offers nothing but rank speculation to support this argument.[86] The most likely reason is that witnesses are much more afraid to come forward in gritty gangland than in leafy suburbs, and they are right to be afraid. Pfaff has made no case at all that prosecutors’ focus is not presently where it should be, so a proposal to change it with grant money falls flat.

Pfaff proposes gathering more data on how prosecutors exercise their discretion, but he offers few specifics and acknowledges the difficulty and danger of quantifying such a complex exercise of discretion.[87] If people are evaluated on the basis of what is easily quantifiable rather than what is really important, their efforts can be skewed in undesirable ways. Just as the No Child Left Behind Act resulted in “teaching to the test,” so data-driven evaluation of prosecutors could subvert justice in pursuit of numbers. The devil is in the details, and we do not have enough detail to evaluate this proposal.

The next proposal is guidelines for prosecutors. To the extent that Pfaff proposes that elected local district attorneys have their discretion controlled by the state attorney general,[88] he is proposing doing away with the localization of authority and accountability that the people of all but a few states have found to be needed. Local control affords each person a greater voice in decisions that affect the quality of life in their community. It also means greater freedom for an individual to move away from a jurisdiction that makes bad choices, since it is easier to move to another city than it is to move to another state. These are not small matters to be given up lightly, and a much stronger case would need to be made before taking such drastic action.

To the extent that Pfaff proposes that a district attorney’s office create internal guidelines so that junior prosecutors are not “winging it” with major decisions about people’s lives,[89] he is right. That work is already underway. It began, in many offices, with a formalized process for deciding whether to seek the death penalty in cases where it is legally available.[90] A less formal, but still not completely untethered, process is appropriate for less serious cases. Small counties with only a handful of prosecutors might not develop their own guidelines, but they can use those from larger offices as a starting point.

Pfaff proposes tackling the issue of locally elected prosecutors making decisions which impose costs on the state. As discussed earlier,[91] he calls this a “moral hazard,”[92] but the fact that prosecutors do not consider costs when deciding how to charge is a feature, not a bug, of the current system. Justice should not be decided on a price tag. The notion that this results in funding prisons instead of other measures, such as more police, is based on an unrealistically compartmentalized notion of public finances. If more funding for police is needed—and I agree that it is—there are other places to get the money, as discussed further in the next part.

After proposing to make prosecution discretion less local with statewide guidelines, Pfaff flips and proposes to make it more local by electing prosecutors from smaller districts. Specifically, he thinks that suburban voters in counties containing both high-crime inner cities and lower-crime suburbs have too much influence, and that they disproportionately favor tougher crime policies. He offers no evidence for this hypothesis. He notes that there are suburb-free prosecutorial jurisdictions in Baltimore, St. Louis, and New York City, but he provides no evidence that these cities have less of the problem he perceives.[93]

There is some danger in going too small with prosecutor offices. Particularly in places with extensive organized crime, the possibility of intimidation or bribery looms larger. Baltimore is the city that resembles Pfaff’s proposal most closely, and to say that it is not a model of excellent prosecutorial practice would be an understatement. Again, a much stronger case would need to be made before this proposal could be seriously considered.

As his final proposal for prosecutor regulation, Pfaff calls for activists to mobilize to vote out tough prosecutors. He cites the example of Kim Foxx ousting Anita Alvarez in Cook County (Chicago and vicinity) in 2015.[94] However, his discussion of that election has a glaring omission. After putting arguably excessive emphasis on the role of money in his criticisms of prosecutors, Pfaff fails to mention the massive infusion of campaign cash by George Soros toward the election of softer prosecutors, including Foxx.[95] It is a billionaire swamping relatively low-budget races to give one candidate an overwhelming funding advantage, much more than mobilized activists, that has elected these “reform” prosecutors. Notably, Soros hit a brick wall in California in 2018, after this book was published.[96] One of his 2016 winners has already thrown in the towel on reelection,[97] and Foxx is on shaky ground.[98] This tide may be starting to turn the other direction.

6. Disenfranchising the People

If convincing voters to elect “reform” prosecutors is unfeasible, another alternative for the infinitely wiser people is to just cut them out of the decision process. Pfaff proposes that we appoint both judges and prosecutors instead of electing them, as is done in the federal system.[99] The people still elect the appointing official at some point, of course, but as long as crime rates do not return to 1980s levels, elections will still be decided largely on other issues. Pfaff cites the recall of Santa Clara judge Aaron Persky for an appallingly lenient sentence for a rapist as an example of why elections for judges are a bad idea.[100] In so doing, he commits a fallacy that he denounces elsewhere in the book: citing an example that is newsworthy because it is so rare, but treating it as an instance of a widespread problem. California judges are almost never recalled, and they are not commonly opposed in their regular re-election bids. He notes that while there is some movement to eliminate elections for judges, there is almost none to eliminate elections for state and local prosecutors.[101]

If the people of almost all of our states have settled on the same system of locally electing prosecutors, there must be a good reason for it. Yet if Pfaff understands the reasons that counsel against his proposal, he gives us no sign of it. Decentralization and direct accountability to the people are important elements of our system of government. Centralization of power gives rise to a host of evils. One would need a powerful case to overturn the collective wisdom of our nearly unanimous states, and mere disagreement with charging policy does not come close to that mark.

As for sentencing laws, Pfaff fortunately does not go so far as to suggest appointment of legislators. Instead, he suggests insulating the legislature from the political consequences of subordinating public safety to other goals by outsourcing the hard decisions on sentencing to a sentencing commission. Pfaff is commendably candid in admitting his anti-democratic motivation for supporting commissions. He quite forthrightly (one might even say blatantly) defines the “success” and “effectiveness” of sentencing commissions solely in terms of whether they reduce prison populations, not whether they make the sentencing systems of their jurisdictions fairer, with sentences proportioned to actual culpability.

This is not to say that sentencing commissions are necessarily a bad idea. The problem with making criminal law by legislation is not excessive democracy, or even lack of expertise, but instead sporadic interest. Legislators react to headline news in both directions—exceptionally horrible crimes on one side and exceptionally unjust convictions or sentences on the other. Reexamining the whole system to see that punishments are actually proportional to crimes is boring work and not on most legislators’ agenda. This is not a new problem; Blackstone complained of this legislative “want of attention” two and half centuries ago.[102]

A sentencing commission’s advantages of expertise and sustained interest without the disadvantage (in my view, not Pfaff’s) of letting legislators off the political hook could be achieved by outsourcing to the commission the committee work but not the floor vote. A commission could produce a set of recommendations, along with suggested amendments to them by any dissenting commissioners, and the legislature could alter its rules so that these recommendations are brought directly to the floor for brief discussion and a prompt vote. Pfaff disagrees with this kind of approach because letting legislators off the political hook (i.e., giving voters less say in the criminal law) is precisely his purpose. If the commission proposals are to come before the legislature at all, he recommends that they take effect if the legislature does nothing because this will facilitate “scaling back punishments” by creating a “less risky” (i.e., less democratic) path for legislators.[103]

The idea of delegating legislative authority to bureaucrats who are not accountable to the people has always been controversial. The U.S. Supreme Court upheld the constitutionality of delegating guideline-issuing authority to the Sentencing Commission thirty years ago, but it did so over a strong dissent by Justice Antonin Scalia;[104] his view of delegation has gained many adherents since that decision. The Supreme Court’s recent Gundy decision, in which four of the eight participating Justices indicated views along the lines of Justice Scalia’s, may indicate that such broad delegations are on shaky ground.[105] Constitutional or not, it is contrary to the spirit of democracy to have laws that govern punishment enacted by unaccountable appointees.

Pfaff has a number of other proposals, stated briefly and mostly in general terms. These include prison closing commissions similar to the post-Cold War base closing commissions, private prisons with rehabilitation incentives, sunset clauses on “crime of the month” laws, the Justice Reinvestment Initiative to fund alternatives to incarceration, cultural changes in attitudes toward crime, and social impact bonds.[106] These are all debatable as to their cost-effectiveness, but the debate is beyond the scope of this review given the brevity of their treatment in the book.

The bottom line of the second half of the book is Pfaff’s proposition that “our attitude toward violent crime needs to change if we hope to end mass incarceration.”[107] If weakening both the deterrent and incapacitative effect of our current laws with a resulting increase in victimization of innocent people really is the only way to end mass incarceration, then I suspect that an overwhelming majority of the American people would just stick with so-called mass incarceration. If there is a path forward to reducing prison populations, it must not be stained with innocent blood.

III. The Road Not Taken

But maybe our long-run choice is not between the current high incarceration rate and inadequately punishing violent criminals. Perhaps there is another way. There is no Part III to Pfaff’s book, but there are some tantalizing tidbits that cry out for further exploration.

In the most curious passage of the book, Pfaff notes that in the three years ending in 2013 the number of people in prison for violent crimes fell almost as much as the number in prison for drug crimes.[108] Yet, as noted earlier, substantially all “reform” efforts have been directed to prisoners convicted of nonviolent crimes, with those convicted of violent crimes expressly excluded. “There’s no explanation for how this happened that I’ve seen,” Pfaff says. “[I]n fact, no one really seems to have commented on it at all.”[109]

A plausible explanation fairly screams off the page. The rate of violent crime has fallen dramatically in the last quarter century. The bottom rate of 2014 was less than half of the horrific peak rate of 1991.[110] One would expect a falling crime rate to eventually produce a falling incarceration rate, although it may take considerable time. If an overloaded system was underpunishing violent criminals in 1991, in the view of the people making the punishment decisions, then the fall in incarceration rates would not happen immediately, particularly if the capacity of the system were expanded, as indeed happened. Eventually, though, between capacity expansion and falling crime rates, the system would reach the correct level of punishment; further declines in crime would further decease the prison population. I do not have the data to test whether this hypothesis is correct, but Pfaff’s statement that no academic is even bothering to ask the question is striking.

If we have reached the point that incarceration rates will fall as crime rates fall, then those whose true goal is reducing the incarceration rate[111] should consider measures that reduce the crime rate as a primary means of achieving the goal. Conversely, if reducing the punishment for violent crime increases the crime rate through diminished deterrence or incapacitation, such measures may be counterproductive to the goal of reducing the incarceration rate.

Pfaff himself recognizes this effect, but curiously only for murder. He says that if police put more emphasis on solving murder cases, then incarceration rates will go up in the short term but down in the long term as more effective enforcement lowers the murder rate,[112] presumably through deterrence, incapacitation, or both. There is no need to limit this to murder. Pfaff points to the recent declines in imprisonment in New York State, noting that the declines came only from New York City, while upstate New York actually increased prison admissions.[113] The decline therefore did not come from statewide legislative changes softening the state’s notoriously harsh drug laws. New York City has seen even more dramatic reductions in crime than the country as a whole, largely because of more effective policing and prosecution.[114]

The intriguing possibility is that prison populations can be brought down through more effective law enforcement without going soft on violent crime. We could invest in more police and more effective policing. More police will, of course, cost a lot of money. More effective use of existing police can be achieved through innovations such as New York’s Compstat and “community policing,” in the original and correct meaning of that term.[115]

While “reformers” often call for funding more law enforcement with the savings from reduced incarceration, it is neither necessary nor practical to expect that we can reduce incarceration before improving law enforcement. Furthermore, savings from reduced incarceration do not come from population reductions alone. It will take closure of facilities and layoffs of personnel to produce real cost savings.[116] It is going to be slow. And if the reductions are achieved through sentence reductions, the resulting increase in crime will offset the gains from better law enforcement to some extent.

Fortunately, it is not necessary to wait for savings, if any, from reducing prisons to fund better law enforcement. The recent improvement of the American economy has state governments flush for now.[117] That money can and should be used for improvements to the institutions that protect people from crime, the first purpose of state and local government. The economy will turn down again eventually. But by the time it does, we should be reaping the benefits of lower crime rates, one of which will be lower imprisonment rates. That cost reduction will help sustain better law enforcement into the future.

Of the factors affecting crime rates, law enforcement is the one most easily changed by government policy, but it is not necessarily the most important. Barry Latzer’s extensive examination of crime rates across time and across groups[118] convincingly demonstrates that culture is at least as important as any other factor in determining crime rates.[119] Pfaff calls for a cultural shift in our attitudes, changing “hearts and minds.”[120] The change he calls for, though, is only in attitudes about punishing crime, not attitudes toward committing crime. A culture of respect for the law and respect for the rights of others would likely do more than any government program to bring down crime rates, which would in turn bring down incarceration rates. Yet Pfaff omits any mention of this possibility from his culture change discussion.

IV. Conclusion

The first half of Locked In is a very valuable contribution to the field. Pfaff’s thoroughly documented demolition of the Standard Story should be read by everyone concerned with the problems of crime and punishment. Knowing where we really are at present is essential for deciding which direction to go from here. As for our new heading, we can do better than Pfaff’s misguided proposals. The best way to bring down incarceration rates is to bring down crime rates, and more attention is needed on ways to achieve both reductions, not trading one for the other.