About this Issue

We all know that the United States imprisons many more people per capita than otherfree countries. But why? Progressives, conservatives, and libertarians all have their favored theories, but this month’s lead authors, Chris W. Surprenant and Jason Brennan, argue that none of them captures the whole story. Here to discuss with them this month are Clark Neily, the Cato Institute’s Vice President for Criminal Justice; and John Malcolm, the Heritage Foundation’s Vice President for the Institute for Constitutional Government and Director of the Meese Center for Legal & Judicial Studies. The conversation will continue through the end of the month, and comments on posts are open to readers during the same time period.

Lead Essay

Perverse Incentives Created Our Terrible Criminal Justice System

The U.S. criminal justice system is worse than you think. Horrific incidents like the recent video of a Minneapolis police office kneeling on the neck of George Floyd happen more often than most people believe. Almost daily, we see police officers beating or shooting subdued citizens, judges or prosecutors imposing unreasonable punishments, or county sheriffs seizing cash and goods from people doing nothing more than driving on the interstate. You can watch cops play murderous games of Simon Says with civilians. Police are rarely punished for excessive force. Prisoners who serve their time return to society as second-class citizens who often cannot vote and have little hope of securing good long-term employment.

The so-called Land of the Free is the Land of the Imprisoned. In 2018, the U.S. imprisoned approximately 2.3 million people, about 1 in 100 adults. The United States has 5% of the world’s population but 25% of the world’s known prison population. 27% of all people incarcerated in the United States, including 60% of all females incarcerated, have not been convicted of a crime. An additional 4.6 million people—2% of the population—are under state control through probation or parole. Over 7 million people, or nearly 2.5% of the population of the United States, are under some form of “carceral supervision”—that is, incarcerated, on parole, or on probation.[1] Only 11 U.S. states incarcerate citizens at a lower rate than the Russian Federation.[2] 35 U.S. states, including overwhelmingly white states like Wyoming and Montana, put people in prison at a higher rate than repressive, authoritarian Cuba.[3]

It wasn’t always like this. As late as 1971, the U.S. incarceration rate was only about 150 per 100,000 people—higher than other OECD countries, but within the same cluster. Since then, it’s exploded to over 700 per 100,000. It’s hard to find reliable data on police violence over time, but what data we do have suggests they are far more violent than before. For instance, in 1981, for example, SWAT teams were deployed about 5,000 times across the United States, mostly to respond to violent crime. By 2015, no-knock SWAT teams were deployed about 100 times per day; 93% were for non-violent drug arrests and searches.[4]

How did we get here? What do we do about it?

Three Incomplete/Mistaken Theories

People of different background ideologies come ready to give the same answers they give everywhere else. Progressives want to blame racism and poverty. Conservatives want to blame family and social collapse and the (resulting?) crime. Libertarians want to blame government overreach, especially the drug war. They are each only partly right, and their stories don’t quite work.

Progressives are right that the U.S. criminal justice is systematically racist. Police officers act more violently toward blacks in ways that cannot be explained away by differences in crime rates. Black Americans are disproportionately and excessively fined by police,[5] are arrested at disproportionately higher rates for non-violent offenses like marijuana possession,[6] and black men receive longer sentences than white men for the same crimes.[7]

Nevertheless, the trends go the wrong way. Neither poverty nor racism have disappeared, of course, but they went on a strong downward trend while the criminal justice system’s degree of harshness and punitiveness went on a strong upward trend. Further, racism does not easily explain why the U.S. system became unusually punitive and harsh toward the white majority.

The conservative story faces similar problems. Conservatives are correct that violent criminals tend to come from one-parent households. The breakdown of families partly explains the rise in crime.[8] Yes, crime rose in the United States (for reasons no one quite knows), and so we might expect increased incarceration.

But the conservative story does not quite work either. Yes, crime rose, but then, starting around 1994, it fell. Yet the punitiveness of the system continued to rise steadily the entire time. Single-parent household rates have been generally stable since 1994, but the criminal justice system has become harsher. U.S. imprisonment rates are not proportional to our higher crimes, and the United States gives far longer sentences than other countries do for equivalent crimes. There is little evidence that adding extra length to sentences does much to deter criminals.[9] In fact, some empirical work finds that increased sentences promote recidivism; the longer a prisoner is in jail, the more he becomes institutionalized and the weaker his ties to his community become.[10]

Libertarians want to blame the War on Drugs, and they are partly right. But if jails released all the prisoners convicted or accused only of drug offenses, the United States would still have far more prisoners than any other country, would still incarcerate at a higher rate than almost any other country, and would still impose longer sentences and harsher punishments.[11] Only about 1 in 5 U.S. prisoners are there only for drug crimes.[12] Some of crimes result from conditions created by the Drug War—just as alcohol prohibition lead to increased gang violence in the 1920s—but eliminating the Drug War wouldn’t solve the problem on its own. There are too many other crimes not connected, even indirectly, to conditions created by the Drug War.

Perverse Incentives: Follow the Money

In Injustice for All, we argue that the United States is unusually dysfunctional because the key players in the system—from voters to politicians, town aldermen to state legislatures, local police to prison guards, prosecutors to district attorneys to public defenders—face unusually dysfunctional incentives.

Cash-strapped municipalities can make money by demanding fines for every small infraction—and by expanding what counts as infractions. Local jurisdictions can save money by hiring fewer police and sending more people to state prisons.

Electing prosecutors, DAs, and judges is largely unique to the United States. Thanks to the problem of rational ignorance, voters are systematically misinformed about crime statistics (they persistently believed crime has been increasing though it actually has been falling) and have no incentive to research more effective criminal justice methods. Prosecutors, district attorneys, and politicians win elections by showing they are “tough on crime” and on criminals, not by defending more humane—or more effective—approaches to criminal justice. They are in an arms race to push for more convictions and ever longer sentences.

State legislators have every incentive to keep jails and prisons open in their home districts. Prisons systems have become a workfare and fiscal stimulus program for impoverished rural areas. Prison guard unions, prison supplies, and poor towns lobby for more prisons.

The U.S. Census counts incarcerated persons as residents of town where they are imprisoned, not the town they lived in before incarceration. In Connecticut, this is responsible for creating nine (majority white) state representative districts that would not meet minimum population requirements but for their prison populations.

In 1981, the Military Cooperation with Law Enforcement Act authorized the military to cooperate with domestic law enforcement agencies. The military began training police in military tactics—the kinds of home invasion techniques SWAT teams use come from the army’s playbook. The 1990 National Defense Authorization Act authorized the military to donate “excess” military equipment to local law enforcement.[13] As of 2005, 80% of towns with populations between 25,000 and 50,000 people had a SWAT team.[14] Thanks to this and other measures, the federal government has inadvertently—or perhaps on purpose—incentivized and subsidized training local U.S. police to deploy military-style suppression tactics during routine calls—except that police face fewer restrictions on violent force than U.S. soldiers.

Public defender offices are poorly funded, while prosecutors work hand-in-hand with police. Prosecutors threaten to throw dozens of charges at suspects for one crime, and then scare suspects into accepting plea bargains. Pre-trial incarceration is common: Citizens who cannot afford cash bail rot in jail while awaiting their trail, and thus lose their jobs, their homes, and often their families.

These and dozens of other unusual perverse incentives explain why the U.S. system is an outlier. It’s true that external shocks made a difference. Rising crime in the 1970s and the moral panic of the Drug War led to an overreaction (to put it charitably), but the overreaction has continued even as crime dropped away.

The key players have a monetary and power-based stake in keeping things going as is. The benefits of bad behaviors are concentrated among the few while the costs are diffused among the many.

Fixing the Incentives

If we want to fix the problems, we need to fix the incentives.

In some cases, that means admitting we can’t really fix the incentives, and so we should instead strip government of some of its power. In general, something should be a crime only if it causes a wrongful harm to a victim, only if criminalization works (by reducing rather than increasing overall crime), only if there are no superior alternatives to criminalization for reducing the problem, only if enforcement passes cost-benefit analysis, and only if government agencies show they are competent to handle the problem.

The drug war fails each part of this test. The U.S. government has proven itself incompetent and highly corrupt in how it fights the drug war. It blights our cities by incentivizing cartels to run violent black markets. The drug war has become a cash cow for both governments and cartel. Reform is unlikely. We should end the war on drugs.

Similar remarks apply to SWAT teams. They are rarely used for their proper purpose. We should at least eliminate them in any town smaller than 100,000 people.

Further, rather than merely asking how to reform prisons, we should look for alternatives to imprisonment. We tend to assume incarceration is the normal, natural way to punish criminals. But historically speaking, it’s unusual and recent. Prison should generally be reserved for convicted violent offenders who represent a continuing safety threat to others, and only if there is no better way to contain them.

We could instead require the guilty to pay restitution to their victims—perhaps with exemplary damages in some cases. We could expand house arrest. We could even substitute caning/public flogging for imprisonment. While we do not endorse this as the best punishment, in the book, we argue that caning is both more effective and more humane than imprisonment. Even our own criminal code agrees: If we beat you with a cane, we might receive a 3-month jail sentence, but if we kidnapped you and forced you to live in horrible, prison-like conditions for 3 months, we’d receive a far longer sentence.

Beyond that, here are other reforms, some we should clearly implement, and others more speculative:

  1. Towns may no longer keep monetary fines. Police may not keep confiscated goods or cash. All fines collected go to restitution funds to make victims of crime whole.
  2. End career prosecutors. Instead, government lawyers will alternate between serving as public defenders and prosecutors, and they will be judged by factors other than their win rates.
  3. End the election of judges and district attorneys. Have them appointed, as in most of Europe. Voters have proven themselves incompetent to judge this matter.
  4. Remove laws immunizing police and others from bad behavior. Make it easier to sue them. Have damages paid not from general taxes, but from salary pools, merit raise pools, and even from retirement funds. Incentivize the police to police each other rather than to look the other way.
  5. Eliminate cash bail. There’s almost no evidence showing that assigning cash bail makes it more likely for defendants to appear for their court dates. New Jersey passed sweeping bail reform measures in 2014,[15] implementing them in 2017. Since these measures were implemented violent crime is down by more than 30%,[16] pretrial detention is down by 44% or about 6,000 people at any given time,[17] and there was a statistically insignificant percentage change in the number of defendants who failed to appear for their court dates.[18]
  6. Eliminate plea bargaining. In practice, prosecutors threaten suspects with multiple charges to incentivize them to plead guilty, even when they are innocent.
  7. Bring back mens rea and test for fair notice. Require prosecutors to prove to grand juries that a person could reasonably have known something was a crime before they committed it. Before even hearing the case, jurors are given the law, and then given a wide range of behaviors, including not only the behaviors attorneys are thinking of prosecuting as crimes, but a range of other related behaviors. Unless a supermajority of the jurors agrees the behavior in question is clearly a violation of that law and that a reasonable citizen would know this ahead of time, then the charges are automatically dropped.
  8. Change who pays for prison: Make cities pay the full cost of any prisoners they send to state or county prisons.
  9. Adopt the ancient Athenian model for punishments: After a conviction, both the prosecution and the defense propose a punishment. If the jury does not unanimously accept the prosecutor’s punishment by private vote, the suspect receives the defense’s proposed punishment instead

The system is broken from top to bottom. There is no singular law or policy responsible for its problems, and so there is no one law or policy that can fix everything. But proposals like these should be taken seriously.


[1] Prison Policy Initiative, “Mass Incarceration: The Whole Pie 2019.”

[2] BBC, “World Prison Populations,” 2019; Prison Policy Initiative, “Following the Money of Mass Incarceration,” 2017.

[3] Ibid.

[4] Jonsson, P. “After Atlanta Raid Tragedy, New Scrutiny of Police Tactics,” The Christian Science Monitor; Hohmann, L, “No-Knock Police Raid Ends in Blazing Tragedy.”

[5] Sances, M. and H. You, (2017) “Who Pays for Government? Representation and Exploitative Revenue Sources,” The Journal of Politics 79(3): 1090-1094.

[6] Drug Policy Alliance, “From Prohibition to Progress: A Status Report on Marijuana Legalization.”

[7] United States Sentencing Commission, “Demographic Differences in Sentencing: An Update to the 2012 Booker Report.”

[8] Harper, C. C., and S. S. MacLanahan, (2004) “Father Absence and Youth Incarceration,” Journal of Research on Adolescence 14:369-397.

[9] Nagin, D. S. (2013), “Deterrence in the Twenty-First Century,” in Crime and Justice: A Review of Research, vol 42: Crime and Justice in America, 1975-2025, ed. M. Tonry. Chicago: University of Chicago Press, pp. 199-263; Levitt, S. D. (2004), “Understanding Why Crime Fell in the 1990s: Four Factors that Explain the Decline and Six that Do Not,” Journal of Economic Perspectives 18:163-190; Doob, A. and C. Webster. (2003), “Sentence Severity and Crime: Accepting the Null Hypotheses,” Crime and Justice 30(2003):143-195; Levitt, S. D. (1998), “Why Do Increased Arrest Rates Appear to Reduce Crime: Deterrence, Incapacitation, or Measurement Error?” Economic Inquiry 36:353-372; Gendreau, P., T. Little, and C. Goggin (1996), “A Meta-Analysis of Adult Offender Recidivism: What Works!” Criminology 34:575-607; Farrington, D. P., P. A. Langan, and P. O. Wikström (1994), “Changes in Crime and Punishment in America, England, and Sweden between the 1980s and 1990s,” Studies on Crime and Crime Prevention 3 (1994): 103-131.

[10] Gendreau, P., T. Little, and C. Goggin (1996).

[11] Pfaff, J. (2017), Locked In: The True Causes of Mass Incarceration and How to Achieve Real Reform. New York: Basic Books.

[12] Prison Policy Initiative, “Mass Incarceration: The Whole Pie 2019.”

[13] Andrzejewski, A. (2016), “War Weapons for America’s Police Departments: New Data Shows Feds Transfer $2.2B in Military Gear,” Forbes, May 10, 2016.

[14] Fund, J. (2014), “The United States of SWAT?National Review.

[15] State of New Jersey 216th Legislature (2014), “An Act Concerning Court Administration, Supplementing Titles 2A and 2B of the New Jersey Statutes, and amending P.L. 1995, c. 325.”

[16] Ibarra, R. (2018), “Crime Rates Plunge in New Jersey, and Bail Reform Advocates Are Gloating,” WNYC.

[17] Parmley, S. (2019), “Two Years Into Bail Reform, Pretrial Detention Much Lower, AOC Report Says,” New Jersey Law Journal.

[18] New Jersey Judiciary 2018.

Response Essays

Do We Have a Mass Incarceration Problem? Compared to What?

I would like to thank the Cato Institute for giving me the opportunity to respond to the provocative essay by Professors Jason Brennan and Chris W. Surprenant, and to engage with my friend Clark Neily on the important issues raised by this article. As is often the case with provocative essays, I agree with some of what it says, while disagreeing with much more.

Brennan and Surprenant begin: “Horrific incidents like the recent video of a Minneapolis police office kneeling on the neck of George Floyd happen more often than most people think. Almost daily, we see police officers beating or shoot subdued citizens ….”

Let’s break this down a bit. While there is no denying that the murder of George Floyd – and I do believe it was murder – was a horrific event, the authors cite no data to support their hyperbolic and inflammatory statement that police beatings and shootings of already-subdued citizens are an almost daily occurrence. While impossible to prove or disprove, my guess is that, while such incidents occur, they are relatively rare. There are roughly 375 million police-civilian interactions every year, yet very few result in injury to either the officer or the civilian. The incidents of police brutality that do occur are isolated examples of misconduct, not standard operating procedure for police departments. While the authors hyperlink to a horrific video of an officer shooting a civilian (white, by the way) who was clearly trying to cooperate, there are many more videos of police officers being shot and killed by civilians.

The main theses of this article are that we incarcerate too many people and that our criminal justice system is suffused with racism. As to the former charge, the authors note that per capita incarceration rates have risen significantly since the early 1970s and are much higher than other countries around the world, leading many to cite a “mass incarceration” problem. While it is certainly true that we have the highest incarceration rate in the world, there are many countries where crime, especially violent crime, is out of control and corruption is endemic. Such countries, I suspect, have an under-incarceration problem, and I am glad that we incarcerate more people than they do.

The authors compare today’s per capita incarceration rates to those that existed in 1971; however, it is hardly surprising that our incarceration rates have risen since then. After all, crime rates in this country rose precipitously in the ‘70s, ‘80s, and ‘90s, leading to many more arrests and convictions. Crime (until recently) has dropped since then, thank God! And while well-respected economists such as Steven Levitt and William Spelman believe that increased incarceration is responsible for only part of the decline, it is not an insignificant contributor.

Moreover, some of the “fixes” offered by the authors to our “mass incarceration” problem would not, in my view, improve the situation. For example, they say we should “look for alternatives to imprisonment” and that prisons should “be reserved for convicted violent offenders who represent a continuing safety threat to others, and only if there is no better way to contain them.” Many jurisdictions are now utilizing diversionary courts, such as drug courts, for certain offenders. The D.C. Advisory Committee to the U.S. Commission on Civil Rights, which I chair, is finalizing a report about D.C.’s highly successful mental health court. Nonetheless, if the authors believe that prison should house only violent offenders, they have obviously never met fraudsters like Bernie Madoff, Bernie Ebbers, James Lewis, Tom Petters or other scam artists such as the telemarketers I used to prosecute who preyed on the elderly depriving them of their life savings. Yes, their victims were still alive, but they were devastated, and the perpetrators deserved to be imprisoned. Should petty thieves who continue to recidivate or refuse to comply with the conditions of diversionary courts or supervised release get a free pass? And if a stalker kills his intended victim but clearly means no harm to anyone else, and thus is not a “continuing safety threat,” is he to remain free too? I don’t think so.

As to the charge of systemic racism, having never been stopped for “driving while black” or followed by a security guard as I sauntered down the aisles of a department store, I approach the issue with some humility. Nonetheless, it again seems to me that the authors have not proven their case.

To support this charge, they hyperlink to an article by Prof. Roland Fryer, but fail to note that with respect to officer-involved shootings, Fryer found no evidence suggesting racial bias, and that while racism might explain other findings, “the statistical evidence doesn’t prove it.” Indeed, the odds of an officer being shot by a black person are significantly higher than the odds of an unarmed black person being shot by an officer of any race.

Brennan and Surprenant seem to be perplexed and troubled by the fact that racism and poverty “went on a strong downward trend while the criminal justice system’s degree of harshness and punitiveness went on a strong upward trend.” To me, this is not perplexing at all, especially when one considers that the vast majority of crime is intra-racial and much of that crime takes place in the inner cities in communities of color.

Black-on-black crime has been and remains a naggingly persistent and seemingly intractable problem. In neighborhoods where looting, gang activity, and drive-by shootings are commonplace, people are trapped inside their homes and businesses suffer. When crime rates go down, people are liberated from their homes and more businesses open and flourish, improving economic prospects, safety, and emotional well-being within the community. Consistently high crime rates in communities of color also militate in favor of assigning more police officers to those communities, not fewer. This will, of course, lead to more police-civilian encounters, some of which will inevitably go awry, although the overwhelming majority will not.

Contrary to the beliefs of those misguided individuals pushing the defund/dismantle movement, while many individuals in those communities are wary of the police, most residents still want more, not less, of a police presence. This was evident most recently in New York City where local officials dissolved a special anti-crime unit in the wake of protests, sparking a backlash from black community leaders who are now asking for more police protection. The interests and views of these concerned residents are entitled to respect too. Indeed, one suspects that if the police departments made a conscious decision to spend more resources protecting predominantly white communities, they would be accused of racism for ignoring black victims.

While the authors suggest, or at least allude to, some proposals that I support, such as mens rea reform, addressing overcriminalization, and civil asset forfeiture reform, some of their other “fixes” strike me as being off the mark.

They call, for example, for appointed prosecutors, arguing that elected prosecutors and other politicians win by running on “tough on crime” platforms, as if that is a problem. While that may have once been true, there are now dozens of prosecutors and politicians who ran on “let’s end mass incarceration and systemic racism” messages and won. Even our current presidential candidates are touting criminal justice reform and disavowing past “tough on crime” positions. Moreover, many of the criticisms that the authors level at elected prosecutors (for seeking high sentences) and judges (for imposing them) have also been leveled at federal prosecutors and judges, who are appointed.

The authors call for eliminating qualified immunity, without any discussion of the potential deleterious consequences that could result. They say civil damages assessed against bad cops should be paid out of funds set aside for good cops who put their lives on the line every day to protect us without explaining why this is fair or just. They contend juries, rather than judges, should be presented with sentencing options from the prosecutor and the defendant’s attorney, but the prosecutor’s recommendation should be imposed only if the jurors unanimously endorse it, without explaining why this task should be left to juries and why unanimity should be required. While I agree with a couple of their proposals and am sympathetic to others, space limitations (and an awareness of the patience of my readers) prevent me from delineating the many concerns I have with their other proffered solutions.

The authors end their article by stating that our “system is broken from top to bottom.” Surely there is room for improvement, but I fundamentally disagree with this conclusion. At the outset, they state that “Only 11 U.S. states incarcerate citizens at a lower rate than the Russian Federation. 25 U.S. states, including overwhelmingly white states like Wyoming and Montana, put people in prison at a higher rate than repressive, authoritarian Cuba.” But given the choice, would anyone subject themselves to a trial in Russia or Cuba or any of the dozens of other countries that fail to provide what we consider basic procedural safeguards over a trial in any of the 50 states? I didn’t think so.

Overcriminalization and Plea Bargaining Make Criminal Justice Like Shooting Fish in a Barrel

It’s tempting to simply aver that Jason Brennan and Chris W. Surprenant are right about everything except for the Drug War and leave it at that. But that would leave me nearly a thousand words short of what I agreed to produce, and like any self-respecting libertarian, I believe in contracts. So I will first explain my caveat about the Drug War and then propose that Jason and Chris consider whether the judiciary deserves even more blame for mass incarceration than their account appears to provide. 

Beginning their post with an agreeable “we-a-culpa” tone, Jason and Chris argue that progressives, conservatives, and libertarians are all partly right and partly wrong in their takes on mass incarceration. Thus, progressives are right that the criminal justice system is systemically racist, but they can’t explain why the system’s “degree of harshness and punitiveness went on a strong upward trend” at the same time that overt racism “went on a strong downward trend.” Similarly, conservatives are right that social factors such as single-parent households partly explain crime trends, but they falter when it comes to rationalizing the imposition of increasingly harsh punishments in a system where crime rates have been falling for decades. As for libertarians, they are partly right in blaming the War on Drugs, but even without prohibition, we would still have a much higher rate of incarceration than most other countries. Thus, say Jason and Chris, “eliminating the Drug War wouldn’t solve the problem on its own” because there are “too many other crimes not connected, even indirectly” to that policy.

I do not propose to wade into the question of just how criminogenic the Drug War has been (extremely), nor will I discuss the rampant corruption it engenders, nor its immiserating effect on people in developing nations. My colleagues have addressed those issues elsewhere. Instead, I will focus on an aspect of the Drug War that makes it qualitatively different from most other criminal justice pathologies.

Unlike racism, which Americans almost uniformly reject and are doing our best to root out, and unlike misguided social policies that systematically shortchange and exploit the needy (looking at you, failed public schools), drug prohibition represents a conscious decision to embrace a morally indefensible public policy. In that sense, the Drug War is not simply an example of the state being misguided or inept; instead, like slavery, Jim Crow, and eugenics, it is an example of the state being evil. And that makes it impossible for people like me, Jason, and Chris (if I may presume to speak for them) to defend the system, even if we believe it has many good features and good people working within it. Simply put, a system that does substantial amounts of unjustifiable violence to peaceful citizens has no plausible claim to their confidence or support.

And this brings us to the thrust of my argument, which is that the powerful economic incentives Jason and Chris so persuasively catalog should not, by themselves, have been sufficient to trigger mass incarceration. Instead, that required two distinct acts of judicial abdication involving, first, the government’s ability to criminalize conduct, and second, its ability to criminalize people.

Among the most important questions in political theory is whether there are any limits on the power of democratic majorities to outlaw particular conduct, and if so, what those limits are. After more than two hundred years of constitutional adjudication, the Supreme Court’s response can best be described as a skeptical shrug. Ban books? Generally not. Punish flag burning? Not right now, but check back in a couple of terms. Lock people up for growing a relatively harmless plant that stimulates the appetite, relieves anxiety, and helps many veterans with PTSD sleep at night? Well of course you can!

As I have argued, along with Roger PilonRandy Barnett, Timothy Sandefur, and others, the Supreme Court has largely abdicated its duty to protect Americans from arbitrary government power, allowing legislatures to criminalize vast swaths of non-wrongful conduct—from betting on basketball to wearing saggy pants—on a whim. Combine that dereliction with the public-choice dynamics that Jason and Chris describe, and you end up with a situation where nearly anyone may plausibly be accused of some criminal infraction. Think you’re safe from prosecution just because you haven’t knowingly broken any laws? Think again.

But even with rampant overcriminalization, there was still one missing ingredient—a key constitutional cop-out before America could become the undisputed “lock-em-up” champion of the world. And what was this “killer app” for mass incarceration? It was to morph the expensive and inefficient mechanism for adjudicating criminal charges prescribed by the Constitution into a sleeker, more efficient process better suited to modern carceral appetites. We refer to that new process euphemistically as “plea bargaining,” but it is far more akin to the brutal system of judicially sanctioned torture that various European regimes used during the middle of the last millennium to avoid the inconvenience of unscripted acquittals.  

I have written extensively on the subject of coercive plea bargaining, and I will not replicate the argument here. Suffice it to say that American prosecutors are lavishly equipped with various tools they can use to induce defendants to waive their constitutional right to require the government to prove their guilt beyond a reasonable doubt to the satisfaction of a unanimous jury and simply condemn themselves instead. Indeed, so effective have prosecutors become in employing those levers that 95 percent of all criminal convictions today are obtained through guilty pleas, and fewer than two percent of all federal prosecutions go to a trial. You might well ask what on earth would persuade a defendant to trade the possibility of acquittal and freedom for the certainty of conviction and punishment. The answer is pressure, and lots of it. What kind of pressure? Well, the Supreme Court has found nothing wrong with threatening to bring new charges carrying a life sentence in order to motivate defendant to take a five-year plea offer, and various lower courts have held that prosecutors may threaten to indict a defendant’s family members for the purpose of exerting plea leverage—an abominable practice that any experienced criminal defense attorney will tell you is perfectly routine in our system.  

From an economic standpoint, there is of course a perfectly understandable rationale for this barbarism. Prosecutors must demonstrate productivity just like anyone else, and guilty pleas are a far more efficient way to rack up convictions than trials. As Professor Darryl K. Brown explains in his masterful article “The Perverse Effects of Efficiency in Criminal Process,” when the cost of a particular good—whether it be cars, corn, or convictions—comes down, consumption will rise. And if the process of convicting alleged lawbreakers becomes no more difficult than shooting fish in a barrel, then you can count on having a nearly inexhaustible supply of prisoners. Which, not coincidentally, is precisely what we do have.

The Conversation

It All Goes Back to Incentives

John Malcolm raises three issues with our piece: (1) police violence towards U.S. citizens really isn’t as bad as we say it is, (2) race and racism is not a significant factor in how people are treated by our criminal justice system, and (3) the incarceration rate in the United States isn’t too high when you take into account our violent crime rate. We don’t have too much to say to these three points beyond providing the relevant data—you can decide if things are as bad as we claim.

So far in 2020, nearly 600 Americans have been killed by our own police officers, roughly three people per day. Over the past five years, we’ve averaged approximately 1,000 police killings per year, and so we’re currently on pace for a 10% increase over that average in 2020. This number is unacceptably high, especially when considering the number of police killings in other countries. For example, our rate of citizens killed by police officers is marginally higher than what is seen in Nigeria, the Democratic Republic of Congo, and Kenya; 50% higher than Mexico; double what is seen in Argentina and Egypt; five times higher than Canada; and 15 to 20 times higher than France, the Netherlands, New Zealand, and Australia.

John argues that violence in the United States accounts for our high number of police killings. He’s correct that the United States is a very violent nation when you look at national statistics, especially when we start comparing ourselves to the European nations we often like to compare ourselves to. When you isolate certain parts of the United States—cities like St. Louis, Baltimore, Detroit, and New Orleans—and then isolate certain parts of those cities, you see violent crime rates that approach the most dangerous cities in the world. But there is no correlation in the United States between violent crime in a city and how many people are killed by police in that city.

Worse than just the large number of police killings is how those killings are distributed. Black Americans are three times more likely to be killed by police than white Americans, even though white Americans outnumber black Americans 4.5 to 1. John claims that disproportionately higher crime in the black community is the cause of this disproportionate number of police killings and that police officers actually show a bias towards black Americans, citing work by Harvard economist Roland Fryer. Sounds incredible, right? That’s because it is.

Fryer himself has become dismayed that his “work has been widely misrepresented and misused by people on both sides of the ideological aisle. It has been wrongly cited as evidence that there is no racism in policing, that football players have no right to kneel during the national anthem, and that the police should shoot black people more often.” Fryer found that “there are large racial differences in police use of nonlethal force” and “compliance by civilians doesn’t eliminate racial differences in police use of force.” While police shootings don’t appear to be affected by race, many officer-involved killings—such as those of Eric Garner and George Floyd—were not shootings, and we should be concerned about the disproportionate use of force by our police against black Americans.

One problem is how our police in many cities around the country are trained. They’re taught to approach all encounters with citizens as if they could be life or death situations—for the officers. While being a police officer in the United States is more dangerous than, say, being a schoolteacher, it’s nowhere near as dangerous as we make it out to be. Each year, less than 100 of the over 800,000 U.S. police officers die on duty, and around half of those deaths are vehicle accidents. That’s roughly 11 deaths per 100,000 officers. By comparison, U.S. sanitation workers are four times more likely to die while working, and commercial fisherman are 10 times more likely to die while working.

John may be correct that the amount of violence in the United States justifies our seemingly large prison population. Even if you’re inclined to believe that—and we’re not completely unsympathetic to this position—there are lots of steps we can take to reduce our prison population without increasing the risk of crime. We go into some detail in our book. One easy step is to release most people from prison who are 40 years of age or older, an age at which most people have “aged out” of committing violent crime. Nearly half of the people in our jails and prisons are over 40 years old. Our position isn’t that these people shouldn’t be punished, but rather that it makes financial sense to pursue alternative methods of punishment for people who are unlikely to reoffend due to their age.

Clark Neily’s response raises numerous points about the power of the prosecutor’s office that are incredibly important to understanding the rise in U.S. incarceration rates. Here, we would encourage you read John Pfaff’s excellent book on this topic. Otherwise there is no question that the drug war has done more harm than good when it comes to increased crime and encouraging all sorts of bad incentives up and down our criminal justice system. When it comes to how decriminalizing or legalizing drugs would affect crime and our incarceration rates, the truth is that no one knows, and imagining what might happen if we decriminalize or legalize drugs is one of the parlor games played frequently by those of us involved in criminal justice reform. But, minimally, the data surrounding police violence suggests that we should be treating drug offenses as medical issues, not criminal issues, and approaching them in a very different way than we do right now.

We are much closer to Clark Neily’s position, and our debate with him here is closer to being intramural rather than between schools. That said, while we agree with much of what he says—the drug war and overcriminalization are very bad—for us, these are not so much fundamental explanations but things that need to be explained. Why did the United States have such a bad response to its drug problem? Why did it empower its police so much, remove so many supposedly constitutional constraints, and criminalize so many things? Here, again, the difference seems to be an artifact of the way district attorneys and others are elected, how police and prisons are paid for, how people are counted, and dozens of other institutions which create perverse incentives and which we do not find analogs of in Europe or Canada. So, the United States had a genuine crime shock—for reasons no one quite understands—and this lead the system’s bad incentives to produce an overreaction/misreaction and a rampant rent seeking.

Before We Criminalize: A Checklist

We think that too many people are in jail for too long. But that statement needs to be unpacked. When we say it, we mean each and all of the following:

1. Certain things should not be a crime at all, and shouldn’t be subject to any kind of civil or criminal penalty. Some people shouldn’t be in jail because what they did should instead by legal. For instance, it should be legal to own and sell marijuana.

2. Certain things are wrongful and should face a penalty of sorts, but not a criminal penalty per se. For instance, if you negligently rear-end my car, you should pay damages, but generally speaking, your action should be treated as a tort, as a civil wrong rather than a criminal wrong.

3. Certain behaviors should be subject to some kind of criminal penalty, but imprisonment is not the right way to punish that behavior. In our book, we argue prison should be used only in special circumstances; it should not be the default form of punishment.

4. Certain criminal behaviors should result in someone going to jail, but the person is in jail for too long and under too harsh or bad conditions. One reason why the United States has such a large prison population is that it generally gives out longer sentences for crimes compared to other liberal democratic countries.

What Should Be a Crime?

Here, Chris and I don’t expect we’ll solve the 2,500-year-old philosophical debate about what exactly the purpose of criminal punishment is. Instead, we offer some heuristics:

1. Nothing should be made illegal unless it’s at some point permissible to kill people, or at the very least severely incapacitate them, to stop that kind of thing from happening.

Every act of criminalization is an implicit death threat. It annoys power-loving conservatives and progressives to no end for libertarians and anarchists to say that, but it’s true. As journalistConor Friedersdorf (neither a libertarian nor an anarchist) says:

Law professors and lawyers instinctively shy away from considering the problem of law’s violence. Every law is violent. We try not to think about this, but we should. On the first day of law school, I tell my Contracts students never to argue for invoking the power of law except in a cause for which they are willing to kill. They are suitably astonished, and often annoyed. But I point out that even a breach of contract requires a judicial remedy; and if the breacher will not pay damages, the sheriff will sequester his house and goods; and if he resists the forced sale of his property, the sheriff might have to shoot him.[1]

He continues:

It is an argument for a degree of humility as we choose which of the many things we may not like to make illegal. Behind every exercise of law stands the sheriff – or the SWAT team – or if necessary the National Guard. Is this an exaggeration? Ask the family of Eric Garner, who died as a result of a decision to crack down on the sale of untaxed cigarettes. That’s the crime for which he was being arrested. Yes, yes, the police were the proximate cause of his death, but the crackdown was a political decree.[2]

Should we threaten to kill people—or ruin their lives—for smoking pot? Eating mushrooms? Snorting cocaine? Doing meth? For selling or exchanging these drugs?

2. The mere fact that something is immoral is not sufficient to make it illegal.

Lots of behaviors are morally wrong, but not all of them or even most of them should be illegal. It’s wrong to cheat on your spouse. It’s wrong to make a false promise to pick your friend up from the airport. It’s wrong to cheat at tennis. It’s wrong not to give some money to an effective charity if you’re rich. If you’re an egalitarian philosopher, it’s wrong not to give most of your money away to others.[3] It’s wrong to grandstand. It’s wrong to waste your own life. It’s wrong to wear a Che Guevera T-Shirt or carry a Nazi flag. Nevertheless, while all these behaviors are morally wrong, it does not follow that they ought to be criminalized or punished.

3. Criminalization has to work

Imagine an evil demon announced that if we criminalize, police, and prosecute murder, it will use its dark magic to ensure that even more murders take place. For every criminal we charge with murder, it will make two more murderers appear.

This wouldn’t change the moral status of murder. It would still be wrong. But it would change what we should do about murder. In this case, we should not use criminal law to try to stop murders…precisely because murder is so bad.

This is a silly hypothetical, but real-life analogs exist. The drug war is a perfect example. Criminalizing drugs not only fails to reduce drug use, but in fact generally increases it, increases the strength of the drugs on the market, increases the profits to be made from selling drugs, decreases the safety of the drugs, and leads to criminal violence as drug sellers try to police their territories and enforce private contracts. In this case, we should not use criminal law to try to stop drug use…precisely because drug use is so bad.

4. Criminalization has to be worth the price we pay.

Imagine a different evil demon announces that it will use its dark magic to ensure that prosecuting a murder would cost us $1 trillion per case. Again, this would not change the moral status of murder, but it would change what we should do about it.

If something is a matter of justice, then it feels callous or monstrous ask whether we can afford it. How can you put a price on justice?

The economist and philosopher David Schmidtz responds, “Some things are priceless. So what?”[4] We might decide that dolphins are the world’s priceless heritage. But if it costs $2 billion to save one priceless dolphin, that’s still $2 billion we’re not spending saving all the other priceless things out there.

Again, the silly demon thought experiment has real-life analogs: Take the drug war again. The United Nations estimates that three quarters of imported drugs would need to be seized before the profitability of the illicit drug trade is significantly compromised. At current levels of police power, only 13 percent of heroin and on average 25 percent of other drugs are confiscated.[5] This means that effectively enforcing the law—enforcing it to a level that would actually significantly reduce drug use—is prohibitively expensive.

Before criminalizing something, you need to ask, how much will it cost and is it worth that cost? That also means asking, of all the ways to enforce and punish that behavior, which way gets us the most bang for the buck?

5. Will the government be competent to police the action?

Someone should be given (significant) power only if they will use that power competently and in good faith. We take this to be a theory-neutral minimal condition of the right to exert power over others. Power: use it well or lose it.

When we decide how much power to give people, we shouldn’t imagine them to be omniscient angels who always and only do the right thing. Instead, what government is allowed to do depends on how well they’ll actually do it. There are things we might want a government of angels to accomplish but which we shouldn’t dare let a government of real people even attempt.

6. Can something be handled as a tort instead?

Instead of criminalizing something, can it be handled through a tort system? Might we instead make it easier for people to sue each other for certain kinds of harms—perhaps even by using money now spent on criminal justice toward this end—and then require wrongdoers to compensate their victims?

Note that we can still use exemplary damages to ensure the rich don’t see this as a license to harm anyone they like. (Paying $10,000 to punch someone in the face might seem like a good deal to some.)

We might also use fines, fees, and the like to pay into victim restitution funds.

We haven’t given you a full theory of criminal law here. However, we think these heuristics tell us the scope of criminal should be narrowed. We can debate just how much.


[1] https://www.theatlantic.com/politics/archive/2016/06/enforcing-the-law-is-inherently-violent/488828/

[2] Ibid.

[3] https://link.springer.com/article/10.1007/s10892-020-09342-2?fbclid=IwAR29uBI7ZFzJwAv3O2LAmbj1t5_B-tWKE04fIVa2L7NTDA7LHu8cGgLnVfw

[4] https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1758-2237.2001.tb00042.x

[5] BrennanJason, “Marijuana,” in Ciment, James, ed.,Social Issues in America (Armonk, NY: M.E. Sharpe, 2006), 1044–54.