While the American criminal justice system was once known for its impressive features, like the jury trial and an independent judiciary, the system’s most notable feature in modern times is its high incarceration rate. The trend began in the 1970s when approximately 360,000 people were locked up in America’s prisons and jails. Those numbers corresponded to an incarceration rate of just over 100 per 100,000 people, a rate that paralleled the incarceration rates of Western European countries. The rate had remained relatively constant for decades, prompting two criminologists to hypothesize, in 1973, that even dramatic societal changes could not alter this “constant level of punishment.”1 The scholars could not have been more wrong.
As my new book, Mass Incarceration Nation: How the United States Became Addicted to Prisons and Jails and How it Can Recover, explains, the 1970s began a process of persistent incarceration growth. By 2002, the United States locked up more than 2 million people in prisons and jails, with the incarceration rate reaching over 700 per 100,000. This rate eclipsed that of every other country as well as the United States’ own historical norms. Incarceration plateaued across the 2000s until disruptions to law enforcement created by COVID-19 and budding reforms brought the number down to around 1.7 million in 2020.
Now, the United States has reached an inflection point. The question is whether we will see a continued decrease in incarceration, a second plateau, or a reversal of the recent downward trend. A return to low incarceration rates depends on legislators, police, prosecutors, and, yes, judges, learning the lessons of the incarceration rates’ precipitous rise. One of the key themes of Mass Incarceration Nation is that it was the remarkable consensus of law enforcement actors — legislators, police, prosecutors, and judges — all enthusiastically embracing punitive measures that caused incarceration rates to skyrocket and stay high even after crime fell.
While people readily recognize the contributions of legislators, police, and prosecutors to incarceration rates, the role played by judges is obscured by a variety of factors. One obscuring factor is that the American criminal justice system is not a system at all, but the illusion of a system generated by the interconnected decision-making of a series of independent officials, each with distinct roles. This independent interdependence makes it hard to hold any single official actor responsible for over-incarceration. Legislators enact criminal laws, police make arrests, prosecutors charge, and judges sentence; but each actor can point to someone else to explain why any particular individual gets locked up.
Another factor that obscures judges’ role is that judges themselves regularly downplay their agency. In 2015, federal Judge Jed Rakoff called out this judicial shyness, arguing that “[f]or too long, too many judges have been quiet about an evil of which we are a part: the mass incarceration of people in the United States.”2 And as federal Judge Lynn Adelman wrote in a co-
authored article for this publication that same year, judges can do more than speak out. Judge Adelman urged judges to act, stating that “if we are actually to achieve a significant reduction in the federal prison population, federal judges will have to make major changes in their sentencing practices.”3 What is true for federal judges is true for state judges as well.
While their individual contributions can get lost in the day-to-day bureaucracy of the modern American “system,” judges play a substantial role in how many people are locked up. And it is helpful to spotlight that role so that judges can be conscientious in assessing their contributions to the country’s incarceration rate and potential role in reducing it.
The clearest role that judges play in incarceration — and the role spotlighted by Judge Adelman — occurs at sentencing. Increased time served in prison and jails is one of the key contributors to the growth in incarceration in the United States. Over the past 50 years, sentence lengths increased across virtually all crime types.
While judges often decry the reduction in their sentencing discretion as a result of mandatory sentencing laws enacted in the past 50 years, legislatures also enhanced judicial sentencing authority in that period. In the 1970s, most jurisdictions relied on parole to keep prison populations low, making parole boards the true sentencing authorities. Under these indeterminate sentencing regimes, judges imposed essentially placeholder sentences with a maximum term. But parole boards ultimately determined the actual time prisoners served, typically well below the maximum. Over time, legislatures severely restricted and, in many jurisdictions, abolished parole. The “truth in sentencing” moniker for this movement reflected the enhanced importance of judicial sentencing decisions as the sentences imposed by judges became much closer to the actual time that defendants served in prison. This was true even in jurisdictions that retained the possibility of parole for some or all prisoners, like Texas and California, as parole boards — themselves caught up in a “tough on crime” mindset — became more reluctant to offer early release.
As I explain in the book, the change from indeterminate to determinate sentencing increased the importance of judicial sentencing, while also enhancing scrutiny on the sentences judges selected.
Determinate sentencing . . . created a scoreboard to judge the judges. Since judges had to issue precise, meaningful sentences, it became easy to identify and replace judges who were “soft on crime.” Conversely, judges who handed down harsh sentences could do so with great fanfare and be rewarded through re-election and promotion.4
Despite the increase in mandatory minimums, most statutes still left judges with discretion to select sentences within a significant range. This was true even in the federal system, where mandatory sentences were most prevalent. In 2011, for example, the Federal Sentencing Commission reported that only 27 percent of sentencings involved a defendant subject to a mandatory minimum and almost half of those avoided the mandatory sentence through substantial-assistance/safety-valve provisions — and that this proportion had remained relatively stable over the past 20 years.5 In other words, mandatory minimums could account for only about 14 percent of sentencings. Even the most impactful American mandatory sentencing law, California’s “three strikes law,” did not remove judicial discretion entirely. Instead, California courts preserved their freedom to dismiss strikes, along with other sentencing enhancements, “in furtherance of justice” — while only rarely exercising their authority to do so.
None of this is to say that judges were the sole reason that time served increased dramatically in the past decades. As Jenia Turner and I explain in a new article, sentencing in the United States is “a dynamic process with substantial contributions from multiple actors.”6 Legislators set sentencing ranges and create mandatory minimum sentences, particularly for crimes that they think judges punish too leniently. Prosecutors select charges and enhancements and agree to plea deals. But judges play an important role at sentencing, too, and they were, and continue to be, part of the same punitive consensus. Some of this was an inevitable consequence of the general ratcheting up of the system’s severity after the 1970s. As I explain in Mass Incarceration Nation,
Increasing sentences for some crimes — especially for crimes that might not seem gravely serious like transporting drugs or possessing weapons — creates pressure to increase the sentences for others. With precise sentences, judges, reporters, and the public naturally came to view sentences as signals of how seriously society viewed a crime (or offender) relative to other crimes (or offenders).7
The result was a kind of “punishment auction” that pushed sentences for everything steadily higher, with judges and other officials using the severity of the sentences imposed to demonstrate the fairness of the American system.
Importantly, the United States expanded the use of incarceration beyond the core crimes, like homicide and sexual violence, that had spurred the nation to increase penal severity in the 1970s, and that most naturally lent themselves to retributive “justice.” The American public and their elected officials embraced the intuitively attractive but empirically mistaken idea that incarceration was also an effective response to urgent policy problems such as drug abuse, gun possession, domestic violence, and drunk driving. As I explain in the book, there is no empirical evidence that long sentences deter these crimes — and given the infrequency with which all crimes are detected and solved in this country, there is no reason to think that they would.
A key theme of Judge Adelman’s 2015 article was that judges should return to the humbler view of sentencing that had governed for “much of this country’s history.” He urged judges to “start[] with the ‘in/out’ question and impose[] sentences of imprisonment only if they were convinced that no less restrictive alternative sufficed to satisfy the purposes of sentencing.” This ties into a theme of Mass Incarceration Nation, that a core set of “justice” crimes may require severe punishment. And there may be additional offenses for which incarceration is necessary to keep society safe. But beyond this minimal core, incarceration serves little purpose — and yet is regularly imposed. As Judge Adelman recognized, judges have many tools besides incarceration, including probation. And even when those tools prove inadequate, that does not mean that incarceration will work any better.
For more sophisticated audiences, judges sidestep responsibility for the explosion of the American incarceration rate by pointing to plea agreements as the primary vehicle by which defendants end up in prison and jail. And it is true that (1) the vast majority of convictions result from guilty pleas not trials, and (2) parties usually negotiate plea deals without direct judicial involvement. But judges can and do shape plea deals both formally and informally.
The suggestion that judges are not responsible for plea deals is misleading in two respects. First, in many jurisdictions, including two of the most substantial contributors to mass incarceration, Florida and California, judges can (and do) participate in the negotiation process.8 And even when judges are not involved, plea deals often leave the ultimate sentence to the judge. True, many plea agreements include recommendations of specific sentences. But judges are not bound by the parties’ suggestions. Every American jurisdiction’s laws require judicial approval of plea deals because it is the judge who bears the ultimate responsibility for sentences. And as the case law reflects, judges do reject plea deals but, more often than not, they do so because they perceive those deals to be too lenient, not too strict.
The second reason that the prevalence of plea deals does not absolve judges of responsibility for over-incarceration is that plea deals do not occur in a vacuum. The parties negotiate with an eye toward the outcome that will result if the defendant rejects the plea deal. And that depends on the judge. Judges signal their sentencing preferences indirectly through their comments and rulings, and directly by imposing sentences when they have discretion to do so, such as after trials. In this manner, judges set the parameters of the parties’ negotiation. As judges sentence more severely, particularly after trials, defendants face increasing pressure to plead guilty on prosecutors’ terms. Indeed, one of the main reasons that defendants seek plea deals is to avoid more severe punishment imposed by a judge.
One of the most dramatic increases in the American incarcerated population occurred in jails, and specifically the number of people held pending trial. In 1970, about 160,000 people were locked up in jail; by 2019, that number was over 700,000. The majority (65 percent) were locked up pending trial, a shift from the 1970s when most people in jail were serving sentences. While police arrest decisions and prosecutor requests factor into these numbers, judges determine whether someone is jailed pending trial. And again, the country’s heightened reliance on pretrial detention resulted not from any empirical evidence of the efficacy of such detention. Instead, more and more people were locked up pending trial because of a paradigm shift in the assumptions (of judges and other officials) about the propriety of using incarceration beyond those few situations in which it is absolutely necessary.
While most convictions result from guilty pleas, trial outcomes remain important because they shape the plea-bargaining landscape. If defendants think they have a realistic chance of acquittal, they become resistant to overly punitive plea deals. But the opposite is true as well. When judicial rulings make it hard for defendants to succeed, prosecutors gain plea-bargaining leverage and can impose more severe plea terms. Across jurisdictions, judges have decreased the prospect of success at trial with their rulings. A key example is rulings on the admissibility of prior-crime evidence.
A surprising percentage of defendants have a prior record. A 2009 Bureau of Justice Statistics study of arrestees in large cities found that 75 percent of those arrested for a felony had a previous arrest and 60 percent had a prior conviction.9 This evidence of prior crimes can influence jurors. In a case where conviction is in doubt, the introduction of a defendant’s prior record can tip a jury toward conviction. Consequently, whatever the merits of the defense, success at trial often depends on the admissibility of prior-crime evidence.
The forecast should be favorable for defendants when it comes to prior crimes. American jurisdictions continue to embrace the common-law tradition that “disallow[s] resort by the prosecution to any kind of evidence of a defendant’s evil character to establish a probability of his guilt.”10 As then-Judge Benjamin Cardozo explained in 1930: “In a very real sense a defendant starts his life afresh when he stands before a jury, a prisoner at the bar.”11
But defendants can no longer be confident that juries won’t become aware of their prior record — for two reasons. First, judges admit prior-crime evidence under Evidence Rule 404(b) (or state analogues) purportedly to prove things other than the defendant’s “evil character” or criminal predisposition. Yet many rulings are best described as allowing prior-crime evidence so long as it is called something besides character evidence, rather than it actually being used for a noncharacter purpose. (A common example is a prior drug sale admitted as evidence of a defendant’s “intent” to sell drugs, even though the conviction only supports that inference by showing the defendant’s “predisposition” to sell drugs.) One judge saw this happening so frequently, he felt “compelled to write separately because I believe that the use of ‘bad acts’ evidence under Rule 404(b) in criminal trials is now routinely used to convince the jury that they should convict the defendant because he or she is not a nice person.”12 By easing the pathway to the admission of prior crimes, judges reduce the prospect of acquittal, pushing defendants to accept more severe plea deals.
Second, judges allow prosecutors to introduce prior-crime evidence when defendants testify. As an evidentiary matter, this happens through Evidence Rule 609 (and state analogues). But it shouldn’t happen nearly as often as it does. The evidence rules only permit the introduction of most prior convictions to impeach the defendant’s “character for truthfulness . . . if the probative value of the evidence outweighs its prejudicial effect to that defendant.”13
On its face, that balance strongly favors the defense. Even if we accept that prior convictions speak to credibility — a contested assumption — all defendants have a powerful incentive to lie to avoid conviction: Their freedom is at stake. Thus, there is no real benefit for the jury to learn that, in addition to this powerful incentive to lie, a testifying defendant also has a prior conviction. In evidence terms, that means that a prior conviction offered to impeach a testifying defendant’s credibility has little “probative value.”
Considering the other side of the Rule 609 balance, there is almost always a powerful prejudicial effect. Indeed, this danger — that the jury will relax the burden of proof when it learns that a defendant has a prior record — is the reason for the general ban on evidence of prior crimes. An evenly balanced test of probative value versus prejudicial effect, like the one set out in Rule 609, should lead to the invariable rejection of prior convictions offered as impeachment.14 Yet that is not the reality. And that is the fault of judges. The result is that defendants with prior convictions frequently face an impossible choice: Either decline to testify at their own trial or suffer the introduction of prior-crime evidence. Both options substantially reduce the prospect of acquittal.15 Even for defendants with a plausible defense, pleading guilty becomes the rational choice, especially if the outcome after a trial conviction will be a severe sentence imposed by the judge.
Another area where judges contribute indirectly to increased penal severity is in the interpretation of criminal law and constitutional rights. The criminal law is supposed to be interpreted narrowly; the long-established concept of the “rule of lenity” operationalizes this idea. But across jurisdictions and time, the courts have broadened as much as narrowed criminal statutes. Every jurisdiction has its own examples. When I practiced, one of the notable examples was the broad range of items, including sneakers, that courts deemed “deadly weapons,” or another line of cases that found unsuccessful efforts to open a door or window to count as “entry” for purposes of establishing felony burglary. Perhaps the strongest example that cuts across jurisdictions is the litany of trial errors deemed “harmless” by appeals courts. (One wonders why prosecutors seek to introduce so much evidence that, courts later rule, made no difference to the outcome.) Every ruling for the government makes convictions easier, increasing prosecution leverage in later plea deals. That leads to more guilty pleas on more severe terms.
Judicial interpretation of constitutional rights offers another example. Sometimes the courts interpret state and federal constitutions to place real obstacles in the prosecution’s path. But over the era of increasing incarceration, pro-defense rulings were the exception, not the norm. And that trend is hard to square with the rights themselves. As I explain in Mass Incarceration Nation, the constitutional rights provided to the accused should provide a robust check on the government’s ability to fill prisons — but, in practice, they do not.
The nation’s founders grew up on stories of English treason prosecutions of famous figures like Sir Walter Raleigh, who was beheaded in 1618. And when folks like George Mason (who lodged in the celebrated Raleigh Tavern in Williamsburg, Virginia) and Thomas Jefferson (who had a portrait of Raleigh in his Monticello home) heard these stories, they identified with Raleigh, not the Crown. Worried that they could fall prey to similar injustices, the country’s founders placed a number of rights designed to undermine the government’s ability to throw them in jail in the federal and State constitutions. Among these rights were a right to a jury trial, a right to an attorney, a right to confront witnesses, the right not to incriminate themselves, a right to due process, and a prohibition of cruel and unusual punishments. It is a great irony that Mass Incarceration arose in a country whose founding documents included such seemingly powerful obstacles to criminal prosecution. But rights must be interpreted. And almost every step of the way, modern judges interpreted these rights in a manner that expanded the reach of the criminal law.16
Of course, judges do not bear sole responsibility for incarceration rates. Mass Incarceration Nation highlights the role of all the officials in the criminal justice system, including legislators, police, and prosecutors, as well as the important role — spanning the book — played by a transient crime spike and the American public’s reaction to that spike. One of the book’s core arguments is that “it takes a village to send someone to prison.” Every official actor had to cooperate to fill the nation’s prisons. In a nutshell, that’s what happened.
But the book includes chapters on the important role played by judges. After all, it is hardly controversial to recognize that judges matter; elevation to a judicial post is a much sought after, and properly celebrated, pinnacle of a legal career. No one would seek the position if judges didn’t have a substantial influence on the cases that came before them. Thus, it should be no surprise that judges played an important role in the country’s incarceration explosion. The good news is that judges can help the country return to its historical norm of low incarceration rates. When that happens, the American criminal justice system will once again be recognized for its best features, not its bloated incarceration rate.
Jeffrey Bellin is the Mills E. Godwin, Jr., Professor at William & Mary Law School where he teaches criminal procedure and evidence. He is the author of Mass Incarceration Nation: How the United States Became Addicted to Prisons and Jails and How it Can Recover (Cambridge Univ. Press 2023).
ENDNOTES:
1 Alfred Blumstein & Jacqueline Cohen, A Theory of the Stability of Punishment, 64 J. of Crim. L. and Criminology 198, 201 (1973).
2 Jed S. Rakoff, Mass Incarceration: The Silence of the Judges, The New York Review of Books, May 21, 2015 https://www.nybooks.com/articles/2015/05/21/mass-incarceration-silence-judges/.
3 Jon Deitrich & Lynn S. Adelman, How Federal Judges Contribute to Mass Incarceration and What They Can Do About It, Judicature, Winter 2015, at 72, 73. For more recent commentary, see Carissa Byrne Hessick, Judges and Mass Incarceration, 31 Wm. & Mary Bill Rts. J. 461 (2022); Rachel Barkow, The Court of Mass Incarceration, 2021–2022 Cato Sup. Ct. Rev. 11.
4 Jeffrey Bellin, Mass Incarceration Nation: How the United States Became Addicted to Prisons and Jails and How it Can Recover 54 (2023).
5 U.S. Sent’g Comm’n, 2011 Report to the Congress: Mandatory Minimum Penalties in the Federal Criminal Justice System 121–22, 132, 140 (2011).
6 Jeffrey Bellin & Jenia Turner, Sentencing in an Era of Plea Bargains, 102 N.C. L. Rev. 179 (2023).
7 Bellin, supra note 4, at 128.
8 See Nancy J. King & Ronald F. Wright, The Invisible Revolution in Plea Bargaining: Managerial Judging and Judicial Participation in Negotiations, 95 Tex. L. Rev. 325, 396–397 (2016).
9 Brian A. Reaves, Bureau of Just. Stat., Felony Defendants in Large Urban Counties, 2009 — Statistical Tables 10, 12 (2013) (citing table 7 and table 9, respectively).
10 Michelson v. United States, 335 U.S. 469, 475 (1948) (citing Greer v. United States, 245 U.S. 559 (1918)).
11 People v. Zackowitz, 254 N.Y. 192, 197 (1930).
12 State v. Willett, 223 W. Va. 394, 400 (2009) (Ketchum, J. concurring).
13 Fed. R. Ev. 609(a)(1)(B).
14 See Jeffrey Bellin, Circumventing Congress: How the Federal Courts Opened the Door to Impeaching Criminal Defendants with Prior Convictions, 42 U.C. Davis L. Rev. 289, 312 (2008).
15 See Jeffrey Bellin, The Silence Penalty, 103 Iowa L. Rev. 395, 415, 433–34 (2018).
16 Bellin, supra note 4, at 134.
