Will Platonic Guardians End Mass Incarceration?

In property law, abandonment occurs when someone intentionally relinquishes any right to control over an item of property, thereby leaving it for someone else to claim ownership. In her 2025 book, Justice Abandoned, Rachel Barkow, the Charles Seligson Professor of Law at the New York University School of Law, expands that concept to include the abandonment of a legal duty, leaving its performance to others. In her view, the responsible party is the Supreme Court of the United States, and the abandonment is the Court’s willful abnegation of its duty to interpret the Constitution honestly and administer justice impartially in criminal cases. To prove her case, she analyzes six Supreme Court decisions: Terry v. Ohio, Bordenkircher v. Hayes, United States v. Salerno, Harmelin v. Michigan, Rhodes v. Chapman, and McCleskey v. Kemp. In each one, she concludes that the Court abjured its sworn duty to serve as an impartial arbiter of justice in favor of becoming a member of the law enforcement team. The result has been to help create and exacerbate the problem that goes by the sobriquet “mass incarceration.”

Each decision that Barkow discusses addresses an issue or practice that commonly arises in criminal prosecutions, particularly ones involving “street” or “blue collar” crimes. The practices are known to every member of the criminal justice system, every criminal defense attorney, and everyone who winds up going through the criminal process: the police practice of stop-question-and-frisk, which can lead to the “unexpected” discovery of illegal drugs or firearms; a prosecutor’s decision to put the hammer down by charging an offense carrying an ungodly long sentence that is brought to induce the accused to plead guilty to a lesser charge; a magistrate’s decision to confine someone in jail pending trial, however far away that might turn out to be; a judge’s decision to follow a legislature’s orders by sentencing a convicted offender to a mandatory lengthy imprisonment term, however unjust that sentence might seem; a correctional system’s decision to overstuff a prison with convicted offenders until it looks like ten pounds of beef jammed into a five-pound bag; and the criminal justice system’s refusal to come to grips with the adverse effect that it has on minorities.

All that, Barkow has argued, has led to a massive number of imprisoned Americans. To understand how those actions have led to mass incarceration, think of the late Stanford law professor Herb Packer’s two models of the criminal justice system. The one he labeled the Due Process Model sought to use rigorous application of procedural safeguards and respect for constitutional rights to distinguish the innocent from the guilty. The other model, which he called the Crime Control Model, saw fit to process each case come what may with the same attitude that a state trooper displays when waiving traffic past a highway motor vehicle accident: “Nothing to see here. Just keep on moving. Move on.” As far as the latter model goes, imagine Henry Ford’s assembly line applied to the criminal process, not to construct a desirable consumer product, let alone to pick the innocent from the guilty, but to identify, scoop up, and warehouse undesirable individuals, principally black and brown, placing them out of sight and out of mind until the day before our sun burns itself out. Today’s criminal process follows the latter model, and it has generated our current unprecedented number of prisoners.

Like Barkow’s first book Prisoners of Politics, Justice Abandoned addresses a subject—overincarceration—about which she cares deeply. If that phenomenon even exists—and, in my opinion, several scholars have done a major league job of puncturing that balloon—most of the literature criticizing it has focused on how the legislative and executive players in the American criminal justice system have cooperated to imprison a shockingly large number of offenders. Justice Abandoned takes that criticism as given and shifts the reader’s focus to the Supreme Court.

In her opinion, the Court has given the political branches in the federal and state governments an Article III papal blessing to engage in a scorched earth approach to crime. Each decision analyzed in Justice Abandoned, she argues, reflects the Supreme Court’s quite intentional choice to surrender to the political processes any effort to regulate the criminal justice system’s operation via application of the Bill of Rights to the parties she sees as the system’s most potentially dangerous officials: police and prosecutors. In so doing, she maintains, the Court abandoned any pretense of relying on any recognizable constitutional interpretative methodology other than “The government always wins,” which is an approach that the Court has adopted to appease a misinformed and misguided public demanding ever stricter criminal law enforcement, come hell or high water. This jihad’s victims are the less fortunate and less powerful—but more darkly complexioned—people regularly seen in state criminal courthouses. To Barkow, labelling as shameless what the Supreme Court has done in the name of public safety does not begin to describe the real-life harm the criminal process has done to those individuals.

As her scholarship evidences (and as I learned from our discussions about clemency, a subject of common interest), Barkow has a very sharp mind, she is a talented writer, her heart is in the right place, and she possesses true compassion for people whom she believes are the victims of an unjust criminal justice system. Like her earlier book Prisoners of Politics, Justice Abandoned is a well-written, very thoughtful, and highly critical analysis of our criminal justice system. I encourage everyone to read it, along with its elder sibling, to learn what a very bright author thinks about where the criminal justice system has gone wrong and how we might reform it. Books like Justice Abandoned can only contribute to an honest and mature debate about how to improve our criminal justice system.

But Barkow appears to have missed the mark with at least some of her case-specific criticisms. Rather than march through Justice Abandoned in its entirety, let me offer one example: the Supreme Court’s 1978 decision in Bordenkircher v. Hayes. Why should we condemn that ruling? Because it “did more to pave the way for mass incarceration than any other Supreme Court decision.” How could that one decision have had such a disastrous effect? Because “[y]ou cannot get mass incarceration without mass case processing, and you cannot get mass case processing without destroying the constitutional right to a jury.” And how do you destroy that constitutional right? By empowering prosecutors to “threaten[] defendants with punishments orders of magnitude greater than anyone thinks appropriate for their conduct if they exercise their jury trial right.” All that makes plea bargaining (and its partner in crime, overcharging) “the rotten core of mass incarceration.” Bordenkircher v. Hayes, therefore, deserves the place of honor at the villain’s banquet.

Does that decision truly deserve such reproach? I don’t think so.

Consider the facts of that case: The grand jury charged Paul Hayes with uttering a phony $88.30 check, a crime punishable under Kentucky law by 2-10 years’ imprisonment. The prosecutor made Hayes’s counsel a plea agreement offer: if Hayes pleaded guilty, the prosecutor would recommend a parolable sentence of five years’ imprisonment, but if Hayes went to trial, the prosecutor would urge a grand jury to charge him with a violation of the state habitual criminal law (Hayes had two priors), which would create the possibility that Hayes would receive a mandatory (albeit parolable) sentence of life imprisonment. The ensuing events unfolded like a Law & Order episode (but without any closing-scene self-recrimination): Hayes rejected the plea offer, the prosecutor obtained a recidivist charge from the grand jury, the jury convicted Hayes at trial, and the judge sentenced him to life imprisonment. In the Supreme Court, Hayes did not claim to be innocent, nor did he challenge the fairness of his trial; he argued only that the prosecutor’s decision to up the ante simply because he went to trial was unconstitutional because it made his decision to stand trial unduly costly. The Supreme Court was unpersuaded, but Barkow says that the Court got it wrong.

I don’t deny that the outcome was uber harsh for Hayes. His fraud lacked a Bernie Madoff-like scale, yet he wound up with a sentence of life imprisonment because, as Barkow argues, he demanded a right that the Sixth Amendment guaranteed him cost-free. But a harsh or ruinous outcome is not necessarily unconstitutional, because not every choice a defendant must make is cost-free.

The Constitution, for example, guarantees the defendant a right to testify at his trial; the Supreme Court so held in Rock v. Arkansas. If he elects to take the stand, however, the defendant has no immunity from cross-examination, nor any shield against the introduction of proof that he has a criminal record, that he has an untrustworthy character, or that he is a liar. As the Supreme Court put it in 1895 in Reagan v. United States, once a defendant takes the stand, “his credibility may be impeached, his testimony may be assailed, and is to be weighed as that of any other witness. Assuming the position of a witness, he is entitled to all its rights and protections, and is subject to all its criticisms and burdens.” That means, in deciding whether the accused is credible, “[t]he jury [may] properly consider his manner of testifying, the inherent probabilities of his story, the amount and character of the contradictory testimony, the nature and extent of his interest in the result of the trial, and the impeaching evidence.” That is settled law—reaffirmed in Raffel v. United States and in Jenkins v. Anderson—even though each government response to a defendant’s exercise of his or her right to testify makes that decision costly.

Similarly, the accused has a Sixth Amendment Compulsory Process Clause right to introduce witnesses in his defense at trial, but a defendant cannot prevent them from being impeached for bias in the defendant’s favor. In United States v. Abel the Supreme Court—unanimously, I should add—held that the prosecution may impeach a defense witness with proof that he and the defendant belonged to the same Aryan Brotherhood prison gang, “a secret prison sect” whose “tenets required its members to deny its existence and ‘lie, cheat, steal [and] kill’ to protect each other.” With remarkable self-restraint, the Court explained that impeachment was permissible because “[t]he tenets of this group showed that [the defense witness] had a powerful motive to slant his testimony towards [the defendant] or even commit perjury outright.”

The point is that the defendant is not the only party with an interest at stake in a criminal trial. As Chief Justice William Rehnquist once wrote, “justice, though due to the accused, is due to the accuser also. The concept of fairness must not be strained till it is narrowed to a filament. We are to keep the balance true.”

The same principle applies to plea bargaining. Once the Supreme Court upheld the constitutionality of that practice in Brady v. United States and Santobello v. New York, there can be no doubt that what the prosecutor did in Hayes was entirely legitimate, certainly foreseeable, and equally constitutional. As Justice Potter Stewart explained in Hayes, “by tolerating and encouraging the negotiation of pleas, this Court has necessarily accepted as constitutionally legitimate the simple reality that the prosecutor's interest at the bargaining table is to persuade the defendant to forgo his right to plead not guilty.” Justice Stewart also quite reasonably concluded that it made no difference whether the prosecutor added the recidivist charge after Hayes rejected the plea offer or obtained that charge beforehand and declined to drop it once Hayes went to trial; the effect on Hayes would have been the same either way. The prosecutor did “no more than openly present[] the defendant with the unpleasant alternatives of forgoing trial or facing charges on which he was plainly subject to prosecution.” The villain (if any) in Bordenkircher v. Hayes was the Kentucky legislature, which decided to empower a prosecutor to charge an offender with an offense carrying a very stiff mandatory penalty. It is not the Supreme Court’s fault that a state legislature empowered a prosecuting attorney to wield a very big stick.

Besides, even if you believe that plea bargaining is evil, it was a necessary evil in 1970, 1971, and 1978, and it remains so today. Plea bargaining disposes of approximately 95 percent of today’s criminal cases, so doubling what we spend on the criminal trial process—an expense that no one encourages, and no legislature would approve—would increase the number of trials to only ten percent. That is hardly an improvement over the current situation. That result is also unlikely to mollify critics of the current number of prisoners—which, by the way, has steadily decreased since 2009, a fact that critics generally don’t mention. After all, so-called mass incarceration’s critics have not identified a “non-mass” or acceptable level of incarceration, which allows them to move the goalposts at will to avoid being credited with success and being ignored thereafter. Of course, if we increased the number of people we execute beyond aggravated murderers to include simple murderers, fentanyl traffickers, and recidivist sex offenders, we could reduce the number of people imprisoned for long terms, but I doubt that the critics of “mass incarceration” have that remedy in mind.

Besides, the Unconstitutional Conditions Doctrine on which Barkow relies is not as clear-cut or monolithic as she lets on. A  number of pre- and post-Bordenkircher v. Hayes Supreme Court decisions—such as Brady, Colton v. Kentucky, Chaffin v. Stynchcombe, Corbitt v. New Jersey, Jenkins, and Alabama v. Smith (in the interest of full disclosure, I represented the United States as an amicus curiae in Alabama v. Smith)—also rejected unconstitutional conditions arguments, making it clear that putting someone to a choice between unfavorable alternatives is not automatically and always impermissible. The belief that plea bargaining is unconstitutional because constitutional rights should be entirely cost-free has gone the same way as Justice Hugo Black’s 1956 ex cathedra pronouncement in Griffin v. Illinois that “[t]here can be no equal justice where the kind of trial a man gets depends on the amount of money he has.” The latter proclamation might sound great when uttered on the U.S. Senate floor, but it has no real-life purchase. In law, as in life, sometimes there is no good option from which to choose, and we must select the best of the available bad options. The same principle applies to plea bargaining. Like it or not, that is an unavoidable concession to reality.

Barkow tries to offer a way out of that dilemma, but her proposal creates numerous application problems without guaranteeing anything other than that we will continue to select from a list of unfavorable outcomes.

Recognizing that prohibiting all plea bargaining is a nonstarter, she would allow a plea deal to go forward if it is “the least restrictive means” of achieving “a “compelling governmental interest,” is “not coercive,” does not create “‘an intolerably high risk of erroneous punishment,’” and has neither “a purpose or effect” of causing a defendant to forego a trial. She concedes that her standard “admittedly may sound difficult to pin down,” but argues that it should be no more difficult to apply than the “nexus” test that the Supreme Court requires courts to apply when the government imposes a land use condition on a homeowner or developer.

I don’t see the workability of that approach or its parallel to land use decisions.

Start with this question: What is a “compelling governmental interest” in this context? Must the defendant be a leader of a terrorist cell? A spy for the People’s Republic of China? An international arms trafficker? Someone who steals or purchases weapons-grade plutonium? A serial or mass killer with victims numbering in double digits? The leader of a Mexican drug trafficking organization or someone else at a senior level in the organization (and what is a “senior level” in one of the cartels)? What about a small-scale drug trafficker who created counterfeit pills to look like Adderall but were one hundred percent fentanyl and wound up killing scores or hundreds of people? Or a senior corporate executive who used phony data to persuade the Food and Drug Administration to approve his company’s product, which sickened thousands of people but did not prove fatal? Or a mid-level corporate executive who dumped R.B.S. (don’t ask) into a small town’s water supply to save money, threatening thousands with dysentery in the short run and cancer over time?

Moreover, how do we make those decisions by any objective means other than by a majority vote of elected legislators? Legal education teaches budding attorneys how to read various types of legal documents, including judicial decisions, statutes, agency rules, contracts, and wills. Practice in the criminal justice field enables neophyte lawyers, over time, to become adept at navigating the complexities of dealing with clients, police officers, jailers, prosecutors, witnesses, judges, probation and parole officers, and others who become involved in the pretrial, trial, and post-trial processes. Neither that education nor that experience trains a lawyer to identify what a “compelling governmental interest” is.

Also, where do we turn for evidence and guidance in resolving that inquiry? Lawyers can read the Constitution’s text and history to learn what guarantees the Framers thought were “compelling” because the members of the Convention of 1787 discussed them and wrote them into the constitutional text. Once we leave that Convention behind, however, we have only the Supreme Court’s precedents to consider, and what they teach is hardly uniform. As one scholar has noted, “the Supreme Court has never given a general account of what makes some ends that government may pursue compelling and others not.” What scale must a particular interest enjoy, for example, to qualify as “compelling”? Does it encompass only indispensable nationwide concerns (such as national defense) or may it include a desire for completeness and consistency in every application (such as a desire for uniformity with no exceptions to a rule)? And what objective basis is there for making any of those judgments that does not rest on a political and moral philosophy that might be highly disputed? There is no obvious answer to those questions, but there should be one if the government must establish a compelling government interest to engage in plea bargaining.

There is more.

What does the term “least restrictive means” entail in this context? Must the prosecution persuade the judge (presumably in secret, to keep the defendant from choosing not to plead guilty) that a plea is critical because the prosecution’s proof is weak since its witnesses are of dubious credibility? What is an “erroneous punishment”? Is it one that is too high (or too low) to promote retribution, deterrence, isolation, rehabilitation, and whatever other purposes of punishment a judge finds legitimate? Is there a number or range that a consensus would accept (and is that a consensus of academics, judges, cops, crime victims, or average citizens)? What is a “high risk” of an erroneous punishment? Fifty-one percent? Thirty-three percent? Ten percent? What is an “intolerably high risk” of an erroneous punishment? Ninety percent? Seventy-five percent? Eleven percent? How will a judge know that a particular plea agreement has neither a “purpose” nor an “effect” of causing a defendant to plead guilty? After all, isn’t that every plea bargain’s very purpose? Also, how will the trial judge make those decisions? Will he question the prosecutor under oath? Defense counsel? The defendant—assuming that he is willing to waive his Self-Incrimination Privilege before receiving the benefit of a judgment incorporating the plea deal? Is a guess as to what those people might say good enough?

Keep in mind that pre-Santobello guilty plea hearings were travesties of dishonesty played out repeatedly in trial courts. Prosecutors, defense lawyers, and defendants would willfully lie that no one had “coerced” the accused to plead guilty, no one had “promised” him anything, no one had made him “an offer he couldn’t refuse”—with the judge acting like he believed what he heard from the players. Even the Supreme Court once described plea bargaining as “a sub rosa process shrouded in secrecy and deliberately concealed by participating defendants, defense lawyers, prosecutors, and even judges.” It is therefore critical to ask why the approach Justice Abandoned suggests won’t lead to the same Kabuki play that we left behind and why that would be a better-run criminal justice process than the one we now have. Even if we were willing to bite our tongue now and then, is there a limit to the number of such travesties we will accept from a particular judge, a particular court, and a particular jurisdiction over a particular time? Barkow’s remedy would certainly keep lawyers and judges hopping, but it cannot answer those questions in an objective manner other than the ol’ totality of the circumstances approach to legal decision-making. That won’t guarantee any more “justice” than we have now.

The parallel that Barkow draws to land use conditions is also unpersuasive. If you want to change your single-family home into a factory or multi-home apartment building, you will burden the existing local garbage and sewage facilities. It therefore makes sense for a county to ask you to offset the increased cost to every other landowner for a publicly funded service. What is comparable about plea-bargaining? A jury trial protects a defendant by transferring the fact-finding function from someone who might be biased because he is on the government’s criminal justice payroll to twelve of the accused’s cohorts who are not. But if the defendant has no serious defense to the charge, a trial would be just a “slow plea hearing,” as a former U.S. district court judge once described it to me. Third parties somewhat bear the cost of making that decision, as it pushes aside trial time for other cases. However, the accused bears it primarily, as they must spend the agreed-upon time in prison in exchange for waiving a trial. Unlike a landowner, the defendant is not making a decision that is all benefit and no cost to him. There is a cost either way, but the accused is in the best position to make it. And whatever experience judges acquire from evaluating Takings Clause claims in local land use disputes won’t be transferrable to whatever analysis judges use (I say “whatever” because what the professor’s standard entails is a mystery to me) when deciding whether a specific plea bargain is lawful or coercive. And if pleading guilty to avoid being sentenced to death under an unconstitutional statute is not impermissibly coercive because counsel represents the defendant—as the Supreme Court held in Brady v. United States—then it is difficult to see, as a practical matter, what would amount to unlawful coercion. After all, as Professor Albert Alschuler once pithily noted, “A guilty plea entered at gunpoint is no less involuntary because an attorney is present to explain how the gun works.”

Finally, suppose that the Kentucky recidivist statute merely changed the potential sentence for Hayes from a discretionary term of 2-10 years’ imprisonment to a mandatory sentence of ten years flat, the maximum that could be imposed without the additional recidivist charge. Would that have generated any feeling that the government had overreached, let alone any outrage? I doubt it. Five years’ imprisonment for crimes like the ones Hayes committed is not unconstitutionally “cruel and unusual,” as the Supreme Court held in Badders v. United States, so a ten-year term for a recidivist surely is not. American courts, including the Supreme Court, have upheld recidivist laws since the Massachusetts Supreme Court’s 1824 decision in In re Ross & Riley, and legislatures may treat recidivists like Hayes more punitively than first-time offenders. Only the length of Hayes’s incarceration makes his case bothersome, and that is a consequence of the legislature’s moral judgment about what punishment is permissible, not the prosecutor’s plea bargaining actions in Hayes’s case.

That brings me to my more fundamental disagreement with Barkow: I do not believe that we can fault the Supreme Court for declining to do the work of the political branches. In fact, we should applaud the Court for staying in its lane.

If the Rehnquist and Roberts Courts have abandoned anything, it is not “justice,” but rather the notion that Supreme Court justices should serve as the guardians Plato’s Republic describes: the class of people whose superior education and wisdom gives them responsibility for making society’s decisions. That has not always been how justices have seen their role; some were (in)famous for taking up that responsibility. But that is not the case for a majority of today’s Court. If the Constitution’s text and history do not objectively answer a question, the Roberts Court has chosen to leave the regulation of our criminal justice system to the branches of our government that people can boot out of office when they go too far or not far enough: legislatures and chief executives. Expressing not arrogance but humility, coupled with an understanding of the judicial role that is firmly tethered to the Constitution’s text and history, a majority of today’s Supreme Court justices have refused to take the law into their own hands and impose their own favored version of it on the nation.

Consider what happens when the Supreme Court does not treat the text and history as dispositive. The classic example is Roe v. Wade. That ruling tossed out the laws governing abortion in each of the fifty states without even the pretense of finding a home in the Constitution’s terms or background for doing so. As the late Professor John Hart Ely trenchantly observed, “a fetus may not be a ‘person in the whole sense,’ but it is certainly not nothing,” despite what the Court thought. After all, “[d]ogs are not persons in the whole sense,” he noted, “nor have they constitutional rights, but that does not mean the state cannot prohibit killing them.” While most people would object to “letting a drug peddler go unapprehended” even if the only way to stop him from plying his trade is “letting the police kick down his door without probable cause,” the “Constitution, which legitimates and theoretically controls judicial intervention, has some rather pointed things to say about this choice,” but it “says nothing, clear or fuzzy, about abortion.” No matter. The then-regnant—and certainly “fuzzy,” both then and now—non-textual concept of “privacy” entranced the Roe Court, which gave it constitutional status without offering a shred of supporting evidence in the charter’s text or history. “What is frightening about Roe,” Ely concluded, “is that this super-protected right is not inferable from the language of the Constitution, the framers' thinking respecting the specific problem in issue, any general value derivable from the provisions they included, or the nation's governmental structure.” Roe v. Wade was “bad,” Ely concluded, “because it is bad constitutional law, or rather because it is not constitutional law and gives almost no sense of an obligation to try to be.” When the Supreme Court later wrote in Dobbs v. Jackson Women’s Health Organization that “Roe was egregiously wrong from the start,” the Court did no more than state the obvious.

Have the political processes sometimes authorized sentences that are too long or give judges too little sentencing discretion? You betchum. To be sure, in some cases, long sentences—even life imprisonment without the possibility of parole, whether imposed on a mandatory basis by statute or as the product of a trial court’s informed discretion-are more than appropriate; they are necessary. That would be true for terrorists, spies, traitors, murderers, sex offenders, and large-scale drug traffickers. Repeat offenders are often in that category, too. Yet long mandatory sentences often serve the interests of only those legislators who vote for them for their later use, not in court, but in political campaign commercials. Academics (along with people who work for a living) need to call out such hucksters as being the worst type of charlatans, people who insult our intelligence by making unsupported and unjustified promises that only long mandatory sentences will guarantee our safety, rather than improvements to the investigative and adjudicatory processes and hiring more qualified people to staff them for the police, the prosecution, and indigent defense counsel. Barkow has already done just that in her first book Prisoners of Politics, which I also highly recommend. More people need to read her work and criticize elected officials who use long prison sentences as bait when fishing for votes.

If the goal is to improve the criminal justice system, sign me up. But don’t ask me to blame Supreme Court justices for not acting like politicians. We already have too many of the latter, and they generally sound, not like George Washington, who couldn’t tell a lie, but like Professor Harold Hill, who couldn’t tell the truth. We don’t need nine more like him.

John, Barbara & Victoria Rumpel Senior Legal Research Fellow, The Heritage Foundation. I want to thank Rachel Elise Barkow, John G. Malcolm, and Zack Smith for helpful comments on an earlier version of this article. The views expressed in it are my own and should not be construed as representing any official position of The Heritage Foundation. Any mistakes are mine.