From the Bench: How Congress, the U.S. Sentencing Commission, and Federal Judges Contribute to Mass Incarceration

Mass incarceration produces deep social transformations in communities that are disproportionately poor, urban, and African American. In particular, the punitive criminal justice policies that American policy makers adopted over the last half-century have resulted in a collective experience for young black men that differs dramatically from the experience of other Americans. Many young black males must contend with long terms of forced confinement and the stigma of criminality, which pervades their post-imprisonment lives. They are marginalized, unable to vote in most states, unlikely to obtain decent jobs, to marry, or even to be included in conventional measures of economic status.

Origins of Mass Incarceration

The policies that this country adopted were inextricably connected to the 1960s urban rebellions against discrimination, the lack of jobs, and the actions of law enforcement. Politicians sought to discredit the protests. Federal officeholders from both major political parties routinely equated urban disorder with criminality. The Republican National Convention of 1964 reeked of hostility to a civil rights bill at the same time that its presidential nominee, Barry Goldwater, denounced the growing menace to personal safety. In 1965, President Lyndon Johnson created the Office of Law Enforcement Assistance, providing additional funding to law enforcement and prisons, and a bureaucracy to fight crime. The Law Enforcement Assistance Act of 1965 and the Omnibus Crime Control and Safe Streets Act of 1968 provided more federal money for the same purpose. Supreme Court decisions such as Terry v. Ohio, which gave the police wide latitude to stop and frisk citizens without probable cause, increased the intensity of the policing of poor and African American neighborhoods and led to record arrest rates.

In his 1970 state of the union address, President Richard Nixon spoke of winning the war against “criminal elements which increasingly threaten our cities, our homes and our lives.” But this rhetoric inaccurately portrayed America’s crime problem. In 1964, the crime rate in the United States was historically unremarkable. The murder rate was 5.1 per 100,000, whereas in 1921 it had been 8.1.

Yet the movement toward harsher crime laws continued. In 1984, Congress enacted the misnamed Sentencing Reform Act of 1984 (SRA) and subsequently passed legislation imposing mandatory minimum sentences in drug cases and others. The SRA abolished federal parole and established a commission to promulgate mandatory sentencing guidelines, thus nearly eliminating judicial sentencing discretion. Congress went on to enact the 1994 Violent Crime Control Act, which provided states with financial incentives to impose longer prison sentences, eliminate parole and good time, and build more prisons. Subsequently, Congress passed the Anti-Terrorism and Effective Death Penalty Act, significantly curtailing state prisoners’ ability to obtain habeas corpus relief for federal constitutional rights violations.

Most criminal justice scholars agree that the narrative that started over 40 years ago—that harsh sentences were needed to incapacitate offenders and deter violent crime—was false, that the federal prison population is too large, that most prisoners could serve shorter sentences with no adverse impact on public safety, and that mass incarceration has a devastating effect on many families, particularly in minority communities. Yet Congress, the U.S. Sentencing Commission, and the federal judiciary, despite being well aware of these problems, have done remarkably little to alleviate them. Congress has not eliminated or reduced mandatory minimum sentences. While the commission made one important change in the drug guidelines and took small steps to improve several others, it has largely ignored the issue of mass incarceration. And though the Supreme Court no longer requires federal judges to follow the very harsh Federal Sentencing Guidelines, most judges continue to do so anyway.

Much of the federal over-punishment problem stems from the SRA. Before the SRA, federal judges enjoyed broad sentencing discretion. In the 1970s, however, a number of legal academics led by Marvin Frankel, a judge and former professor of administrative law, proposed creating an agency to establish numerical sentencing guidelines. They argued that judges had too much sentencing discretion, leading to excessive sentencing disparity. The notion that disparity was widespread, however, was supported by little evidence. And to the extent that disparity existed, it had multiple causes aside from judicial discretion, including differing prosecutorial practices and regional traditions.

Moreover, Frankel and his colleagues gave little thought to the likelihood that supplanting judicial discretion and individualized sentencing with numerical guidelines would produce much harsher sentences. Like mandatory minimums, numerical guidelines are influenced by political considerations and are therefore always higher than they should be. Diminished judicial discretion also increases the power of prosecutors—who decide what charges to bring and, therefore, what guideline will apply. Frankel’s group also rejected the idea shared by many students of criminal justice that numerical guidelines were, by definition, rigid and inflexible and thus incapable of capturing important differences between offenders.

Criminal justice is such a politicized issue that it is hard to legislate on the subject without getting enmeshed in the politics of law and order. The SRA was no exception. Its main proponent, Senator Ted Kennedy, had no knowledge of or experience with sentencing. Worse, Kennedy was overly eager to be regarded as an effective legislator and thought that the way to achieve this goal was to pass a lot of bills. Thus, there was almost no compromise that he would not enter into, no amendment he would not accept, if it would help to get a bill that he was sponsoring passed.

Kennedy accepted numerous harmful amendments to the SRA, including one making the guidelines binding instead of advisory. As the bill went through Congress, it essentially became another law-and-order bill, whose principal feature was that it would impose additional punishment on virtually all defendants. Nevertheless, Kennedy strongly supported it, as did Joe Biden, who, as chairman of the Senate Judiciary Committee, played an important role in passing many of the brutal anti-crime laws of the 1980s and ’90s. Besides Kennedy and Biden, the SRA’s greatest supporters were President Ronald Reagan, who touted it as legislation that would crack down on criminals, and far right senators like Strom Thurmond and John McClellan, who knew that curtailing judicial discretion would lead to harsher sentences. These conservative senators also had long-standing complaints about federal judges stemming from decisions in the areas of racial segregation and criminal procedure and, for this reason too, were happy to limit judicial power. Although a few insightful legislators such as John Conyers spoke out against eliminating individualized sentences, the SRA passed overwhelmingly.

Unsurprisingly, the commission that the bill created proceeded to promulgate very harsh guidelines. Only one of the seven commissioners had any previous sentencing experience, and all seven regarded the Justice Department and the most law-and-order members of Congress as their principal constituencies. Although the commission stated that it based the guidelines on past sentencing practice, it had limited data on past practice and, in many cases, ignored past practice entirely. To the extent that the commission relied on past practice, it counted only past prison sentences, ignoring that approximately 50 percent of pre-guideline sentences were sentences of probation. Strongly biased in favor of prison sentences, the commission created guidelines that barred judges from putting defendants on probation except in the rarest instances. To this day, the guidelines do not affirmatively recommend probation in any case; rather, the penalty for every offense references the number of months in prison.

Nearly all of the guidelines were excessively severe. The drug guideline is an example. Before the commission formulated the guidelines, Congress enacted the Anti-Drug Abuse Act of 1986, which established a three-tier sentencing structure calling for sentences of 10 years to life, five to 40 years, and zero to 20 years, depending on the quantity of drugs. The commission, which was supposed to be independent of Congress, quickly followed suit, establishing punitive guidelines based on the identical quantities. The result was increased prison terms far above previously typical lengths. The child pornography guideline is another example. Over time, a combination of congressional directives and guideline changes caused the mean sentence for possession of child pornography to increase from 36 to 110 months. As for fraud and theft offenses, like drug offenses, the commission treated amount as a proxy for seriousness. Focusing so exclusively on quantity, however, often led to the unfair oversimplification of complicated facts as well as to many enormously long sentences.

The commission also discounted congressional directives that could have softened the guidelines. For instance, the SRA stated that a probationary sentence would generally be appropriate for first offenders who had not been convicted of a crime of violence or an otherwise serious offense. The commission viewed this language as a problem, which it solved by simply reclassifying as “serious” many offenses for which defendants had previously received probation.

Over the next 20 years, the commission regularly amended the guidelines to make them even more severe. As one scholar put it, the guidelines were subject to a one-way upward ratchet in which sentences were raised easily and often, and lowered rarely and with difficulty. Thus, the average federal sentence increased from 28 to 50 months. With the SRA’s abolition of parole, the average time that a defendant served increased from 13 to 43 months. While in the pre-guideline era, judges sometimes imposed excessively harsh sentences, under the guidelines, harshness became a rule of law. Between 1987 and 2013, the federal prison population increased from about 50,000 to almost 220,000, before dropping slightly.

In its landmark decision in United States v. Booker, the Supreme Court in 2005 struck down the mandatory feature of the guidelines, thus giving judges the opportunity to establish a less punitive sentencing regime. In subsequent decisions, the Court made clear that judges had no obligation to follow the guidelines. Unfortunately, however, judges have almost totally failed to take advantage of Booker to ameliorate the harshness of the federal sentencing regime. After Booker, judges slightly reduced the length of sentences, from 47.9 months in fiscal year 2003, to 44 months in 2016. Shockingly, however, the number of offenders receiving prison-only sentences actually increased, from 83.3 percent in fiscal year 2003 to 88.1 percent in fiscal year 2016.

Sentencing Process

The only inference that can be drawn from these numbers is that, notwithstanding the punitive and prison-oriented nature of the guidelines, judges are not troubled by them. How can this be? It may be that judges have become so used to the guidelines that relying on them seems normal and appropriate. Possibly, the answer lies in the psychological phenomenon known as anchoring, the tendency of someone who must decide an inherently subjective question and is given a numerical baseline to give great weight to the numerical baseline regardless of its merit. Also, judges are people who follow rules and may, therefore, tend to follow guidelines even if they don’t have to. Finally, judges may have consciously bought into, or less consciously assimilated, the punitive attitudes that have been so prevalent in this country for so long.

Other factors also contribute to the problem. To some extent, appellate courts have pushed judges toward the guidelines. When the guidelines were mandatory, courts of appeals overzealously enforced them. And even after Booker, when reviewing sentences for reasonableness, courts of appeals have tended to under-police guideline sentences and over-police non-guideline sentences.

Other players in the sentencing process, such as probation officers who provide judges with presentence reports on defendants, also push judges toward the guidelines. When the mechanistic approach to sentencing called for by the guidelines commenced, the role of the probation officer shifted dramatically. Under the SRA, probation officers, who formerly were the social workers of the criminal justice system, became the guardians of the guidelines. Instead of preparing reports geared toward understanding the causes of an offender’s behavior and assessing the possibility of change, the officer’s main task was to determine what numerical guideline applied to the facts of the case. Even after Booker, this role has changed little. Working closely with the commission, probation officers almost invariably recommend guideline sentences. The same is true of prosecutors, who almost uniformly continue to advocate guideline sentences.

In recent years, much has been written on mass incarceration. Advocates have called for changes in the policies that cause it, and policy makers are well aware of the issue. President Obama became the first president to visit a federal prison and granted clemency to a number of offenders, albeit a very modest percentage of the federal prisoner population. And in a recent session of Congress, bipartisan sentencing reform legislation scaling back mandatory minimum sentences was approved by both the Senate and House judiciary committees, before stalling due to opposition from a handful of conservative senators.

Clearly, each of the major players in federal sentencing could do much to reduce mass incarceration. It seems unlikely, however, that much progress will be made in the near future. In his campaign, President Donald Trump emphasized law and order and said nothing suggesting that he regarded mass incarceration as a problem. Attorney General Jeff Sessions, a strong law-and-order advocate and proponent of mandatory minimum sentences, was one of the senators to kill the bipartisan sentencing reform bill in Congress. As attorney general, Sessions has revived the long-disproved narrative that most drug offenders are violent and is unlikely to encourage Congress to cut back on mandatory minimums or to reduce any federal criminal penalties.

As an independent agency, the commission could take meaningful steps in the area of mass incarceration regardless of the political ideology of the administration in power. And, as stated, the commission has recently taken several positive steps. For the commission to make a major dent, however, it would have to recognize how serious the problem of severity in federal sentencing actually is and that too many defendants are serving sentences longer than necessary to satisfy any reasonable purpose.

Reducing Mass Incarceration

Reducing mass incarceration is conceptually simple. It means sending fewer defendants to prison for shorter periods of time. Thus, the first thing the commission could do is to review the guidelines from top to bottom with an eye toward recommending as many non-custodial sentences as possible and substantially reducing sentences. Although the commission should examine every guideline, it might begin with those that, based on sentencing practices, judges consider most problematically harsh.

The likelihood that the commission will seriously address the problem of the severity of the guidelines, however, is small. This is so partly because, despite its independence, the commission is inevitably somewhat political. Its members are nominated by the president on the recommendation of the political parties. More importantly, however, the commission has focused excessively and single-mindedly on the issue of inter-judge disparity. Like Congress, the commission has ignored the fact that the most comprehensive empirical study of the SRA concluded that such a focus was fundamentally misguided. In fact, the study found that the SRA itself, as well as the guidelines, was based on the misconception that justice in sentencing is synonymous with the minimization of inter-judge disparity. Rather, the study concluded that the ultimate objective of an effective and credible sentencing system should be reasonable sentences, and that this objective could only be achieved by the exercise of informed discretion by judges and not by a system of formal rules yielding a quantitative measure of justice. Finally, the study determined that the sentencing system established by the SRA was based not only on a misconception but also actually on hostility to the practice of judging, and that a system based on this attitude could never achieve justice.

Thus, neither Congress nor the Sentencing Commission is likely to significantly reduce mass incarceration in the next four years. Sadly, the possibility that federal judges will do much to alleviate the problem is also remote. Judges have not only adhered too closely to the guidelines but have also been infected by the preoccupation with inter-judge disparity. Law professor Nancy Gertner, a former judge, described this problem well: “The Guidelines essentially supplanted everything. It was almost as if we could no longer speak about anything else. . . . ‘It’s as if the only thing we are talking about is whether I am doing the same thing as Judge Adelman is doing, even if we are both wrong.’” Federal Sentencing under “Advisory” Guidelines: Observations by District Judges, 75 Fordham L. Rev. 1, 5, 7 (2006).

The response to one of my own decisions also illustrates the problem. The main point of the decision, which involved an immigration issue, was that a punishment enhancement called for by the guidelines was excessively harsh. A subsidiary point was that the sentence called for by the guidelines was higher than those imposed on defendants in certain districts near the Mexican border that were employing special “fast-track” programs. Many judges responded to the decision, but every respondent commented on the subsidiary point relating to the disparity in the practices of different districts and none to the critique of the harshness of the enhancement.

One can only hope that a recent speech by commissioner and Eleventh Circuit Judge William Pryor, entitled “Returning to Marvin Frankel’s First Principles in Federal Sentencing,” is not a sign of things to come. Unfortunately, like Frankel, Pryor, a former Alabama attorney general and longtime associate of Sessions, measures fairness in sentencing primarily by the presence or absence of disparity. To reduce disparity, he seeks to eliminate individualized sentencing. Pryor objects to the discretion conferred on judges by the Supreme Court in Booker, objects to allowing judges to consider individual characteristics of defendants such as their backgrounds and family histories, and advocates a return to mandatory guidelines. While Pryor acknowledges that severity is a problem in the present system, his insistence on mandatory guidelines means that many defendants would serve even longer prison sentences than they do now and that defendants who could appropriately receive sentences of probation will be sent to prison. It is ironic that in a country where economic and political inequality has emerged as an issue, one of the few areas where federal policy makers actually attempt to achieve greater equality is in corrections, yet their effort is to ensure that criminal offenders spend an equal—and very substantial—amount of time in prison.