As quantum theory developed, Erwin Schrödinger began to explore the strange results the theory seemed to predict. Oversimplifying, quantum theory proposed that a single atom could be in two places at once but that observing the atom at one point would cause it to exist at only that point.1 The atom, prior to observation, both existed and did not exist at a particular point.2 In a thought experiment meant to highlight the absurdity of such a result, Schrödinger asked his colleagues to imagine two closed boxes, each of which holds a single atom that is exhibiting this strange behavior. Other than the atom, one box is empty; inside the other is a cat and a Geiger counter that, upon measuring the presence of an atom, would pull the cork from a bottle of cyanide, spilling the poison and killing the cat. Schrödinger suggested that quantum theory’s prediction meant that it was possible to create a scenario in which the cat was simultaneously dead and alive.3 Absurd as this seems, a nearly two-decade-old federal circuit split places federal defendants in an equal state of indeterminacy.

The passage of the Comprehensive Crime Control Act of 19844 created the indeterminacy that this Comment addresses. This omnibus bill marked a major shift in how the federal judiciary dealt with criminal defendants in nearly every phase of the criminal justice process. Most notably, the Act fundamentally altered both the federal bail system and federal sentencing. At bail hearings, federal judges were now empowered not only to impose conditions that would assure a criminal defendant’s appearance at trial, but also to consider the risk a defendant might pose to the community if released.5 The Act also emphasized the need for certainty in sentencing by replacing discretionary federal parole boards with judge-ordered supervised release terms—set periods of time following prison terms during which a defendant is allowed to live in the community but required to adhere to certain conditions.6 Although the aims of both the sentencing and bail reforms are relatively clear, the interaction of the two systems has created substantial uncertainty among the circuit courts. In particular, the federal circuits are divided as to whether pretrial detention can toll a supervised release term.

18 USC § 3624(e), which governs the tolling of supervised release terms, states that such terms “do[ ] not run [that is, are tolled] during any period in which the person is imprisoned in connection with a conviction for a Federal, State, or local crime unless the imprisonment is for a period of less than 30 consecutive days.”7 The question is whether, when a defendant is ultimately convicted, the credit he receives to his prison term for any time spent in pretrial detention creates the necessary “connection.” If it does, it tolls a supervised release term, delaying the expiration of the term for an amount of time equal to the time the defendant spends in pretrial custody.

Consider the following example. Defendant X is convicted by Judge A of possession of a controlled substance and sentenced to a term of imprisonment to be followed by a two-year supervised release term. Sixty days prior to the expiration of his supervised release term, he is arrested and charged with larceny. Unable to secure pretrial release, he is detained while awaiting trial for a total of seventy days. Ultimately, he is convicted of larceny and sentenced to a term of imprisonment. If his supervised release term was tolled while in pretrial detention, he will still need to serve sixty days of that term when he is released after serving his prison term for larceny. If it wasn’t tolled, it expired prior to his larceny conviction, and he is no longer under court supervision for his possession conviction.

Defining the exact contours of a supervised release term has important consequences for criminal defendants. Federal law requires that every supervised release term carry the condition that the defendant refrain from committing another crime.8 Violation of this or any other condition of supervised release may result in revocation of the term and the imposition, in its stead, of a prison term equal to “all or part of the term of supervised release authorized by statute . . . without credit for time previously served.9 If a defendant had been serving a two-year supervised release term, then revocation—no matter when during that term it occurs—could result in a two-year prison sentence. What’s more, the federal sentencing guidelines advise judges that prison terms imposed upon revocation “shall be ordered to be served consecutively to any sentence of imprisonment that the defendant is serving, whether or not the sentence of imprisonment being served resulted from the conduct that is the basis of the revocation of probation or supervised release.”10 The power to revoke lies with the court that imposed the sentence, but it remains with that court only so long as the supervised release term is running.11 While the term runs, the court also has the power to extend the length of the term and may modify any of its conditions.12 The most important limit on the scope of a court’s near-plenary power in this area, then, is temporal. Determining whether pretrial detention tolls supervised release thus determines when a defendant is released from the power of the sentencing court.13

Consider again Defendant X from above. As a reminder, sixty days before the expiration of his two-year supervised release term imposed by Judge A for possession, X was arrested and charged with larceny.14 At his bail hearing, Judge B places him in pretrial detention, in which he remains for seventy days. At his larceny trial, X is convicted and sentenced to another year in prison. Because in almost all jurisdictions there is a statutory requirement that prison sentences receive credit for any time spent in pretrial detention,15 his new prison sentence will be credited with the seventy days he spent in jail awaiting trial. After learning of this new conviction, Judge A initiates revocation proceedings on the theory that X violated the conditions of his supervised release. X objects and claims that his supervised release expired as scheduled. Judge A disagrees. He believes that when X received a credit to his sentence for his pretrial detention, that detention became “connected” with his conviction and thus tolled his supervised release term. If X is correct, Judge A has no jurisdiction, and thus no power to impose any additional punishment. If Judge A is correct, however, he may revoke X’s supervised release term and require that X serve up to two additional years in prison. Although this hypothetical seems as though it would arise only infrequently in the real world, a Bureau of Justice Statistics study of recidivism rates among the supervised release population suggests otherwise. That study found that within a year, nearly 20 percent of offenders under supervision were arrested for a new crime.16 Within five years, 43 percent of the sample population was arrested on suspicion of a new crime.17 Clearly, a substantial number of the inmates awaiting trial in jails around the country may simultaneously be serving supervised release terms.

Since 1999, six separate circuits have considered this question. While each circuit proclaims the statutory language provides an unambiguous answer, no consensus has developed. Four circuits believe, like Judge A above, that the “connection with a conviction” requisite to toll supervised release exists when a defendant receives a sentence credit for time spent in pretrial detention, thereby connecting the conviction and the period of pretrial detention.18 In these circuits, defendants who are serving supervised release terms while they sit in pretrial detention are in Schrödinger’s cell. As they await the disposition of their second trial, they cannot know whether their pretrial detention will ultimately be connected to a possible conviction, tolling supervised release. Until they receive judicial “observation” in the form of a verdict, they remain both under and not under court supervision. Two other circuits have held that pretrial detention is incapable of ever tolling supervised release. One circuit contends that the statute unambiguously requires tolling only for periods of detention that follow convictions.19 The other circuit reasons that the statute unambiguously precludes the kind of backward-looking analysis in which the majority position engages.20 Their reasoning, however, offers insufficient rebuttal to the majority position.

This Comment resolves the confusion by focusing on the purposes of the statutes in question. A proper resolution of this issue demands an understanding of not only the tolling provision, but also the other statutes with which it interacts. Part I of this Comment investigates the text and the legislative history of each of those statutes. Part II canvasses the cases that have confronted this question and analyzes the various positions of the circuits. Finally, Part III proposes a purpose-driven resolution to the interpretive problem the tolling provision has posed for courts. Contrary to every court to have considered the issue, this Comment accepts that the statutory text is ambiguous at best. Thorough analysis of the legislative history, combined with consideration of the quasi-constitutional ramifications of the majority position, indicates that pretrial detention should not toll supervised release.

I. Statutory Background

The tolling of supervised release is controlled by a single statutory provision, 18 USC § 3624(e), but the current controversy also concerns several other statutes. Namely, this issue implicates federal bail statutes and 18 USC § 3585, which credits the sentences of convicted defendants for time spent in pretrial detention.21 Each of these statutes was created or amended by the Comprehensive Crime Control Act of 1984, passed at the end of “a decade long bipartisan effort . . . to make major comprehensive improvements to the Federal criminal laws.”22 The drafters of this omnibus legislation claimed it would “restore a proper balance between the forces of law and the forces of lawlessness.”23 Understanding two of the Act’s component chapters—the Sentencing Reform Act of 198424 and the Bail Reform Act of 198425 —is vital to determining whether Congress intended pretrial detention to toll supervised release terms.

Part I.A examines the Bail Reform Act and the history of bail reform. It discusses the legislative history and logistics of pretrial detention, noting the judicial reaction to these reforms. Part I.B analyzes the Sentencing Reform Act. In particular, it examines one of the Act’s major innovations—supervised release. It also reviews the legislative history of one of the Sentencing Reform Act’s minor facets—the sentence-credit provision. Al­though that provision is perhaps an unassuming piece of the overall project of these Acts, it has become the lynchpin for the majority of circuits that have considered the central question of this Comment.

A. The Bail Reform Act of 1984

Bail and pretrial detention are deeply rooted in Anglo-Saxon history and have been a part of American criminal procedure since before the beginning of the republic.26 In this Comment, the term “bail” refers to the conditions attached to the pretrial release of a defendant. From its Anglo-Saxon origins on, the most common condition was financial. The presiding magistrate would determine an amount (based on the severity of the alleged crime) for which the defendant was required to secure a personal surety who would be responsible for payment should the defendant flee.27 This practice was seen as a way to balance the need to ensure a defendant’s appearance at trial with the desire to limit pretrial detention.28

The colonists brought this tradition with them to North America, where defendants would have “a friend or neighbor take a pledge, backed by property, and assume responsibility for [them] until trial.”29 Although the Framers were certainly familiar with the practice of bail,30 they neglected to explicitly guarantee a right to bail in the Constitution. Instead, the nation’s charter provides only that “[e]xcessive bail shall not be required.”31 It is difficult to say with any certainty, then, whether the Framers believed bail was a necessary practice.32 Despite the Constitution’s implied approval of bail, the Supreme Court has consistently affirmed that bail is essential to the fundamental presumption of innocence in American criminal procedure.33 These declarations notwithstanding, American judges have always retained the power to deny bail and detain defendants as they await trial, though they typically did so only when the defendant was charged with a capital offense.34

Over time, ordinary citizens became more reluctant to act as sureties for criminal defendants, giving rise to the “commercial bondsman system.”35 Because “[t]he pecuniary commitment of a commercial bail bondsman did not necessarily reflect the defendant’s own stake in appearing at trial,” the breakdown of the personal surety system led to the imposition of “financial conditions of release that exceeded [ ] defendant[ ] ability to pay.”36 In effect, the dawn of the commercial bondsman ushered in the commonplace use of financial conditions as “sub rosa pretrial detention.”37

In response to this development, Congress enacted the Bail Reform Act of 1966,38 which attempted to emphasize pretrial supervision as a way to secure the release of defendants without the requirement of excessive financial conditions.39 The reliance on financial conditions in the federal system continued to wane until the Bail Reform Act of 1984 finally prohibited sub rosa pretrial detention by declaring that judicial officers “may not impose a financial condition that results in the pretrial detention” of the defendant.40 The 1984 Act, however, did more than prohibit excessive financial conditions. It fundamentally altered the purposes to which judges were permitted to put pretrial detention. It allowed judges to consider not only the risk that a defendant might flee to avoid trial, but also the risk the defendant posed to the members of the community into which he would be released.

1. The history of the Bail Reform Act of 1984.

Prior to the Bail Reform Act of 1984, the Court limited bail to a single purpose, namely, “assuring the presence of [the] defendant” at trial.41 In fact, “[b]ail set at a figure higher than an amount reasonably calculated to fulfill this purpose [was] ‘excessive’ under the Eighth Amendment.”42 In 1984, Congress felt the purpose of bail needed expansion. At least one representative of Congress believed that “over half of those out on bail [were] committing crimes” and that the bail system must account for this danger.43 The 1984 Act required judges to consider not only what conditions would assure a defendant’s appearance at trial but also the safety of the community into which the defendant might be released.44 This marked a major shift from the Bail Reform Act of 1966, the legislative history of which clearly indicates a belief that deciding to detain a defendant because of “predicted—but as yet unconsummated—offenses” was “extra-legal.”45 Some members of Congress were cognizant of this remarkable change, noting that the 1984 Act marked “a significant departure from the basic philosophy of the Bail Reform Act [of 1966], which is that the sole purpose of bail laws must be to assure the appearance of the defendant at judicial proceedings.”46 But many in Congress advocated for the change on the grounds it would “address the alarming problem of crimes committed by persons on release.”47

United States v Salerno48 upheld the constitutionality of the Bail Reform Act of 1984’s requirement that judges consider community safety when setting bail and when authorizing pretrial detention.49 In that case, the Court stated that “when Congress has mandated detention on the basis of a compelling interest other than prevention of flight, as it has here, the Eighth Amendment does not require release on bail.”50 In holding thus, however, the Court was careful to note that “[i]n our society liberty is the norm, and detention prior to trial or without trial is the carefully limited exception.”51 Prior to Salerno, the Court had held that “under the Due Process Clause, a detainee may not be punished prior to an adjudication of guilt.”52 In order to reconcile this precedent with its finding that the Bail Reform Act of 1984 was constitutionally permissible, the Court held that Congress’s intent was “not [to] formulate the pretrial detention provisions as punishment for dangerous individuals.”53 Accordingly, pretrial detention, even when ordered for predicted “but as yet unconsummated” crimes, can never be penal.54

2. Logistics of bail and pretrial detention.

The Bail Reform Act of 1984 set up an intricate statutory scheme to guide judges in determining whether a defendant should be detained or released pretrial. When a defendant is charged with a federal crime, the district court must determine if or how that defendant should be monitored pending trial.55 The court may release a defendant on his personal recognizance or through bail “upon execution of an unsecured appearance bond.”56 Alternatively, the court may craft conditions for a defendant’s release that attempt to both “reasonably assure the appearance of” the defendant and provide for “the safety of any other person and the community.”57 If, however, the court finds “that no condition or combination of conditions” will suffice, it may decide that pretrial detention is appropriate and order that the defendant be returned to jail to await trial.58 In deciding which of these routes to choose, a judge must consider “the nature and circumstances of the offense charged,” “the weight of the evidence against the person,” the “person’s character,” and, importantly, whether at the time of the charge the person was “on [ ] release pending . . . completion of sentence for an offense under Federal, State, or local law.”59 If scrupulously followed, this last factor makes it all the more likely that those defendants charged while serving a supervised release term will be held pretrial and find themselves in the circumstances described in this Comment.

If a defendant is detained before trial, his time in pretrial custody will be automatically credited to his sentence if he is ultimately convicted. 18 USC § 3585 provides that “[a] defendant shall be given credit toward the service of a term of imprisonment for any time he has spent in official detention” as a result of the offense of conviction or any other charge for which the defendant is arrested after the commission of the offense of conviction.60 The only limit on these sentence credits is that a period of pretrial detention cannot be credited against more than one sentence.61 This process is discussed in greater detail in Part I.B.3.

Because the tolling provision in 18 USC § 3624(e) provides that federal supervised release does not run during periods of imprisonment in connection with state or local convictions, as well as federal convictions, this issue also implicates state and local pretrial detention.62 The courts that have held pretrial detention tolls supervised release, however, rely on the sentence credit to create the requisite connection. Simply put, these courts believe that when pretrial detention is credited to an ultimate sentence, it becomes connected with the underlying conviction and tolls supervised release terms. Therefore, although this Comment explores the interaction of these federal statutes, because state procedures differ with respect to crediting pretrial custody to ultimate sentences, this issue may not arise in some jurisdictions.63

B. The Sentencing Reform Act of 1984

Congress felt federal sentencing “lack[ed] the sureness that criminal justice must provide if it is to retain the confidence of American society and if it is to be an effective deterrent against crime.”64 To address this, the Sentencing Reform Act fundamentally reconceived how federal defendants were sentenced in two ways. First, the Act created the US Sentencing Commission which promulgated the then-mandatory (and now-advisory65 ) sentencing guidelines that required judges to sentence defendants within strict sentencing ranges.66 Second, the Act abolished parole and replaced it with supervised release.67 The drafters had as their goal “a comprehensive and consistent statement of the Federal law of sentencing, setting forth the purposes to be served by the sentencing system and a clear statement of the kinds and lengths of sentences available for Federal offenders.”68 More specifically, the drafters hoped their reforms would “assure that the offender, the Federal personnel charged with implementing the sentence, and the general public [were] certain about the sentence and the reasons for it.”69

Congress hoped that supervised release would be a more consistent and predictable substitute for parole. Parole decisions were traditionally made at hearings conducted by examiners selected by the US Board of Parole.70 Although the Parole Board attempted to inject some certainty into this decision-making process through the creation of guidelines for the hearing examiners, those guidelines did not “completely eliminate opportunities for unstructured discretionary judgments.”71 As a result, hearing examiners often provided inmates with “spurious explanations [that] conceal[ed] . . . the true basis for the parole decision.”72 The exercise of such wide discretion also meant that neither the defendant, the court, nor the victim could know how much of a term of imprisonment would actually be served. According to the Senate Judiciary Committee, under the new system of supervised release, “the question whether the defendant will be supervised following his term of imprisonment is dependent on whether the judge concludes that he needs supervision.”73

Congress used supervised release to do away with the uncertainty that plagued the parole system in two ways. Supervised release is imposed by a judge at the time of initial sentencing and commences only once a defendant completes his prison term.74 More importantly, a supervised release term is a fixed period of time (subject, of course, to the judge’s power to modify the term described in the Introduction).75 Taken together, these innovations meant that, following the imposition of a sentence, defendants, victims, and the government knew with reasonable certainty the exact contours of the punishment.

1. Legislative history of the supervised release tolling provision.

Legislative history sheds little light on the supervised release tolling provision. One possibility is that the inclusion of a tolling provision was simply an attempt to align probation and supervised release. Probation, which also contains a tolling provision, mirrors supervised release in other important respects.76 For instance, the conditions attached to probation are essentially identical to those attached to supervised release.77 Furthermore, the power of the court to modify those conditions is identical to the power of a court to modify the conditions of supervised release.78 The court’s jurisdiction over and power to revoke a probation term is, like its jurisdiction over supervised release terms, temporally limited.79 Furthermore, the Supreme Court has discussed the purposes served by probation in much the same way that it has discussed the purposes of supervised release.80

Initially, the supervised release tolling provision read, in pertinent part:

The term runs concurrently with any Federal, State, or local term of probation or supervised release or parole for another offense to which the person is subject or becomes subject during the term of supervised release, except that it does not run during any period in which the person is imprisoned, other than during limited intervals as a condition of probation or supervised release, in connection with a conviction for a Federal, State, or local crime.81

In describing the provision, the Senate Judiciary Committee essentially went no further than recapitulating the terms of the statute.82

Two years later, the provision was amended to its current form, which provides that “[a] term of supervised release does not run during any period in which the person is imprisoned in connection with a conviction for a Federal, State, or local crime unless the imprisonment is for a period of less than 30 consecutive days.”83 The Senate Report for this amendment stated that its purpose was simply “to conform a provision concerning the running of a term of supervised release . . . with a similar provision about probation in 18 U.S.C. [3564].”84 A more illuminating comment was made by the House Report on the same amendment. It noted that the amendment was intended to make clear that “a term of supervised release does not run while the person is serving a term of imprisonment in excess of 30 days for any offense (Federal, State, or local).”85 The House’s conception of the necessary causal relationship between an offense and the kinds of imprisonment that toll supervised release terms is of vital importance to this Comment’s position.86

2. Logistics of supervised release terms.

Following a defendant’s conviction, a district court has authority to include a supervised release term with any prison term it chooses to impose.87 The court must consider the nature of the offense, along with the capacity of supervised release to deter recidivism, to incapacitate the defendant, or to rehabilitate the defendant when deciding whether a supervised release term is appropriate.88 Apart from determining the length of the release term, the district court also has discretion to include a number of conditions if the court believes they are reasonably necessary to serve the penal interests of supervised release.89 In addition to these discretionary conditions, federal law makes several conditions mandatory. Among those is “that the defendant not commit another Federal, State, or local crime during the term of supervision.”90 While on supervised release, the offender is placed under the supervision of a probation officer who, among other things, is required to “keep informed . . . as to the conduct and condition of [the offender], who is under his supervision, and report his conduct and condition to the sentencing court.”91

The Supreme Court has unequivocally held that supervised release commences the day the defendant is released from the custody of the Bureau of Prisons.92 After a term begins, the sentencing court has enormous discretion over its operation. A court may terminate the term of supervised release,93 extend the term,94 or “modify, reduce, or enlarge” the conditions of release.95 Most important, the court may revoke a supervised release term and require the defendant “to serve in prison all or part of the term of supervised release authorized by statute for the offense . . . without credit for time previously served on postrelease supervision.”96 Accordingly, if a defendant is serving a two-year supervised release term and violates a release condition ten days before its expiration, the judge may revoke the term and sentence him to serve two years in prison. Revocation proceedings are typically initiated at the direction of the sentencing court with jurisdiction over a given term of supervised release. When a probation officer files a report describing conduct that may amount to a violation, the court may issue a warrant for the defendant’s arrest.97 Revocation proceedings are governed by Rule 32.1 of the Federal Rules of Criminal Procedure, which provides that upon execution of the warrant the offender should promptly be given a hearing in front of a magistrate to determine whether probable cause exists to believe a violation occurred.98 Should the magistrate find probable cause, he then holds a revocation hearing at which he will determine whether revocation is warranted.99 A US Attorney will prosecute the offender, calling witnesses and presenting evidence supporting the allegation of a supervised release violation.100 Revocation is proper only when a court finds, by a preponderance of the evidence, that the offender violated a release condition.101

The sentencing court typically retains authority to modify or revoke a supervised release term from the day it commences until the day it expires. According to statute, the court may only revoke a term of supervised release after its expiration “if, before its expiration, a warrant or summons has been issued on the basis of an allegation of [ ] a violation.”102 Thus, the power of a court over a defendant’s release term is defined, in large part, by that term’s expiration date. If no warrant or summons has issued prior to that date, the court lacks any power to revoke the term.

3. Legislative history of the sentence-credit provision.

Some courts believe that the credit a defendant receives to his ultimate sentence for time in pretrial custody creates a connection between a defendant’s pretrial detention and his conviction and thus should toll a supervised release term.103 In the current federal system, 18 USC § 3585(b) mandates that such credits be applied to the sentences of convicted defendants.104 A consideration of the history of the sentence-credit provision is helpful to understand both its purpose and how it is best interpreted. Initially, the sentence-credit provision provided a “credit . . . for any days spent in custody prior to the imposition of sentence . . . for want of bail set for the offense under which the sentence was imposed where the statute requires the imposition of a minimum mandatory sentence.”105

This credit provision had two conditions that are no longer in the statute. First, the statute provided credit only when a person was detained because they could not afford bail. Second, credits were available only to defendants who were convicted of crimes that carried mandatory minimums.106

The primary purpose of the sentence-credit provision was “to eliminate the disparity in sentences under certain statutes requiring mandatory terms of imprisonment.”107 The 1960 Congress that enacted this statute sought to respond to the fact that defendants who did not pay bail and whose crimes carried mandatory minimums were spending far longer in custody, pretrial or otherwise, than their peers who were able to secure release prior to trial.108 In other words, because mandatory minimums meant judges could not account for time already spent in custody, Congress hoped to provide a statutory alternative.109

The Sentencing Reform Act of 1984 changed the sentence-credit provision to its current form.110 The current sentence-credit provision states “[a] defendant shall be given credit toward the service of a term of imprisonment for any time he has spent in official detention” as a result of the offense of which he is convicted or any other charge for which the defendant is arrested after the commission of the offense of which he is convicted.111 The drafters did not explain why they chose to change the provision, except to point out that it provides the defendant with a credit for time in custody for the charge on which the sentence was ordered or “was a result of a separate charge for which he was arrested after the commission of the current offense.”112

* * *

Although this overview of the statutory framework may give the impression that each covers a separate phase in a defendant’s interaction with the criminal justice system, that is not how many criminal defendants experience them. In practice, these mechanisms—bail, pretrial detention, sentence credits, and supervised release—overlap a great deal. For instance, the sentence-credit provision exists only as a counterpart to the availability of pretrial detention. As the example in the Introduction demonstrates, when a defendant serving a supervised release term is arrested on suspicion of a new crime, all three statutes are implicated.113 Indeed, one need not invent hypotheticals to see the frequency of such a situation—18 percent of defendants in “federal community supervision” (either on probation or serving a supervised release term) were arrested during such supervision at least once within a year of being placed under such supervision.114 As shown below, courts have focused exclusively on the statutory text of the supervised release tolling provision housed in 18 USC § 3624(e). This Comment shows how a thorough understanding of this network of statutes provides an answer to the question that has troubled the courts.

II. Current Judicial Interpretation

Courts have had little guidance from the Supreme Court in determining whether pretrial detention can toll a supervised release term. The Court has, however, considered the supervised release tolling provision, albeit briefly, and its comments deserve examination. Part II.A discusses what little Supreme Court precedent exists surrounding the tolling provision. Then, Part II.B discusses the case in which the Ninth Circuit became the first circuit to directly confront the controversy this Comment resolves. Part II.C analyzes the four circuit cases that subsequently rejected the Ninth Circuit’s position and expressed what has become the majority position on this issue. Finally, Part II.D discusses the DC Circuit’s recent consideration of the issue and its rejection of the majority position.

A. The Supreme Court and the Supervised Release Tolling Provision

The Supreme Court has considered the tolling provision only once, in United States v Johnson.115 Although the Court did not consider the effect of pretrial detention on a supervised release term, its comments surrounding the tolling provision help explain how the Court views supervised release more generally. Roy Lee Johnson had been sentenced to 111 months in prison, which were to be followed by a three-year term of supervised release.116 Following a Supreme Court ruling that implicated one of his convictions, he filed an unopposed motion to vacate, and his sentence was reduced to fifty-one months.117 By that time, however, he had already been incarcerated for over fifty-one months, and thus, Johnson believed he had already begun to serve some of his release term.118 The Government contended, however, that no part of a supervised release term could be served while a person is imprisoned. The Court was called on to determine at what point a supervised release term commences.119

The Court held that a supervised release term cannot begin until a defendant has left the custody of the Bureau of Prisons.120 In so holding, the court made several illuminating observations. The Court noted supervised release “fulfills rehabilitative ends, distinct from those served by incarceration.”121 Furthermore, the Court was clear that “Congress intended supervised release to assist individuals in their transition to community life.”122 In its attempt to support its conclusion, however, the Court seems to have spoken rather imprecisely. Congress explicitly instructed judges to consider the extent to which supervised release is necessary “to afford adequate deterrence to criminal conduct” and “to protect the public from further crimes of the defendant.”123 While supervised release may be better at serving rehabilitative ends, that is not its only purpose. The Court’s remark is nonetheless understandable. Congress certainly designed supervised release with an offender’s transition to community life and rehabilitation in mind.124 The Court reasoned that “[t]he objectives . . . of supervised release would be unfulfilled if excess prison time were to offset and . . . reduce terms of supervised release.”125 Although not stated explicitly, the Court suggests that when an offender is in prison he cannot engage in the transition to community life, nor can he be rehabilitated in the way the supervised release system envisions.

Turning to the statutory text, the Court noted that, in providing that “[t]he term of supervised release commences on the day the person is released from imprisonment,”126 the statute “suggests a strict temporal interpretation, not some fictitious or constructive earlier time.”127 The opinion turned on whether the statute provided for concurrent running of a prison term and supervised release.128 The Court held that it did not, stating that “[t]he statute instructs that concurrency is permitted not for prison sentences but only for” terms of probation, parole, and supervised release.129 The Court went on, stating the tolling provision “does address a prison term and does allow concurrent counting, but only for prison terms less than 30 days in length.”130

Although the Court’s opinion suggests the tolling provision refers only to punitive prison sentences and terms, it is plausible that this choice of language was due entirely to the question before the Court. The Court was asked to determine whether it was possible for a prison term to run concurrently with supervised release, and it interpreted the tolling provision as providing a negative response.

The Court’s first pass at 18 USC § 3624(e) in Johnson set the table for the current circuit split. Its two main arguments come into conflict in the cases that more directly address the tolling provision. On the one hand, the Court’s insistence that incarceration is inconsistent with the objectives of supervised release suggests a finding that pretrial detention should toll release terms. On the other hand, the Court’s remark that the statutory text requires a “strict temporal interpretation” suggests otherwise.131 Though none of the courts that have divided over the correct interpretation of the tolling provision explicitly rely on Johnson’s reasoning, this conflict underlies the dis­agreement described below.

B. The Ninth Circuit and United States v Morales-Alejo

In United States v Morales-Alejo,132 the Ninth Circuit was the first court of appeals to consider whether pretrial detention can toll a supervised release term. The facts model the typical circumstances that place this issue squarely before a court. Jose Morales-Alejo, an alien, pleaded guilty to illegally reentering the United States. Judge James A. Redden sentenced Morales-Alejo to a two-year prison term to be followed by a one-year supervised release term.133 Upon release from his imprisonment, Morales-Alejo’s supervised release commenced.134 On the same day, he was deported.135 Eight months later, Morales-Alejo was arrested by an agent of the Immigration and Naturalization Service (INS) and was indicted on another charge of illegal reentry.136 After sitting in pretrial detention for almost four months, Morales-Alejo, just one day before his term of supervised release was set to expire, again pled guilty and was sentenced to a prison term to which his time in pretrial detention was credited by operation of 18 USC § 3585.137 While Morales-Alejo awaited sentencing on this new charge, however, Redden issued a “warrant and order to show cause regarding revocation.”138 Because Morales-Alejo’s supervised release term, absent tolling, would have expired before Redden issued his warrant, Morales-Alejo moved to dismiss the revocation proceedings, claiming that Redden lacked jurisdiction.139 Consequently, the Ninth Circuit was asked to determine whether the term had been tolled for any reason.

The Ninth Circuit concluded that “the intent of Congress is apparent from the language of the statute” and that it had no cause to examine the legislative history.140 The court maintained that a plain reading of the statute “suggests that there must be an imprisonment resulting from or otherwise triggered by a criminal conviction” in order to toll a supervised release term.141 This excluded pretrial detention “because a person in pretrial detention has not yet been convicted and might never be convicted.”142

This interpretation, according to the court, was strengthened by its finding that Congress used the terms “imprisonment” and “detention” in very different ways.143 While the term “imprisonment” “is used to refer to a penalty or sentence,” the term “detention” describes “a mechanism to insure a defendant’s appearance [at trial] and the safety of the community.”144 Because the statute tolls release terms only when a defendant is “imprisoned in connection with a conviction,” it was clear that Congress did not contemplate pretrial detention.145

The prosecution in Morales-Alejo raised the argument that, because the defendant had received a credit to his sentence by virtue of 18 USC § 3585(b)(1), “the detention period [became] part of the sentence.”146 The Ninth Circuit rejected this argument, stating that the statute “gives no indication that Congress ever contemplated the type of backward-looking analysis suggested.”147 The court also noted the practical difficulties of such a construction. If similar facts existed, a judge would be unable to determine whether she retained jurisdiction over a given supervised release term “in any case in which a judgment of conviction had not yet been entered.”148 For those interpretive and practical reasons, the Ninth Circuit held that pretrial detention could not toll supervised release.149

C. The Majority Position

Beginning in 2008, the Sixth, Fifth, Eleventh, and Fourth Circuits each considered the question in quick succession. Each court determined that the Ninth Circuit had erred.150

In 2008, the Sixth Circuit decided, in United States v Goins,151 that pretrial detention did toll a supervised release term when it was credited to a defendant’s ultimate sentence.152 That court faced facts that mirrored, in important aspects, those present in Morales-Alejo,153 namely a defendant who had been detained pretrial and whose supervised release term had been revoked after it would have expired absent tolling.154 After being convicted of bank fraud, Roderick Goins was sentenced to a one-month prison term to be followed by a five-year supervised release term.155 While serving his release term, Goins was charged with identity fraud.156 He sat in pretrial detention for sixty-three days until he finally posted bond and promptly absconded.157 Then, eighteen days after his supervised release would have expired absent tolling, the relevant court issued an arrest warrant for possible violations of supervised release.158 The question, therefore, was whether the sixty-three days Goins had spent in pretrial detention for the identity theft charges had tolled the release term.159

The Sixth Circuit “respectfully decline[d] to adopt the Ninth Circuit’s interpretation” of § 3624(e).160 The court was required to “apply the plain meaning of the statute if the statute [was] not ambiguous.”161 The court held that the plain meaning of the term “imprison,” according to conventional and legal dictionaries alike, “includes not only confinements as a result of a conviction, but [more broadly] any time the state detains an individual.”162 Furthermore, the Sixth Circuit reasoned that the Ninth Circuit’s definition of the term “imprisonment,” as including only incarceration resulting from or triggered by a conviction, rendered the statute’s phrase “in connection with a conviction” superfluous.163 The Goins court noted that a common canon of statutory interpretation demands that courts “make every effort not to interpret a provision in a manner that renders other provisions of the same statute . . . superfluous.”164

Having rebutted the Ninth Circuit’s contention that imprisonment means only penal incarceration, the Sixth Circuit next determined that the sentence-credit provision in 18 USC § 3585 created a connection between pretrial detention and a subsequent conviction.165 Accordingly, it held that pretrial detention can toll a supervised release term.166 Although the court conceded that its interpretation required a backward-looking analysis, it stated “the phrase ‘imprisoned in connection with a conviction’ eschews any temporal limitations.”167 Finally, the Goins court responded to the Ninth Circuit’s practical concerns. It saw no reason that a district court attempting to determine whether it retained jurisdiction over a supervised release term could not simply initiate revocation proceedings and wait until the defendant was convicted to determine whether those proceedings were valid.168 In the court’s opinion, its interpretation of the statute, “despite its indeterminacy, creates only a rare and remediable problem.”169

In United States v Molina-Gazca,170 the Fifth Circuit agreed with the reasoning of the Goins court.171 It adopted the Sixth Circuit’s position and buttressed that court’s determination that the distinction between Congress’s use of the terms “imprisonment” and “detention” was not so clear-cut as the Morales-Alejo court had made it seem.172 The court noted that 18 USC § 3041, which describes the powers of federal judges and magistrates “seem[ed] to reject an imprisonment-detention distinction.”173 That provision states an offender “may . . . be arrested and imprisoned or released . . . for trial before [a] court of the United States as by law has cognizance of the offense.”174 Congress’s use of the term “imprison” to refer to pretrial detention was evidence, according to the Fifth Circuit, that Congress may have contemplated pretrial detention when it used the same term in the supervised release tolling provision. The Fifth Circuit, like the Sixth, conceded that its interpretation would, in some instances, create uncertainty surrounding a defendant’s supervised release, but concluded that, to the extent the language of § 3624(e) “results in uncertainty as to a defendant’s status, our role is not to imply [ ] limits when Congress could have done so in the first instance.”175 Following the Sixth Circuit’s lead, the Fifth Circuit observed that courts uncertain of their jurisdiction could “wait[ ] to see if a conviction will actually occur.”176

Both the Fourth and Eleventh Circuits have accepted the reasoning of the Fifth and Sixth Circuits wholesale.177 The Fourth Circuit emphasized the fact that the supervised release tolling provision used the phrase “during any period.”178 To that court, this phrase “necessarily includes all time periods, both before and after a conviction, for which an imprisonment is connected with that conviction.”179 The Eleventh Circuit was apparently so confident in the majority position that it did not even mention Morales-Alejo and instead relied entirely on the Fifth and Sixth Circuits’ interpretation of § 3624.180

D. The DC Circuit’s Rebuke of the Majority Position

In 2016, the DC Circuit, in United States v Marsh,181 examined whether Congress included pretrial detention when it provided that supervised release is tolled “during any period in which the person is imprisoned in connection with a conviction for a Federal, State, or local crime.”182 The DC Circuit joined the Ninth Circuit in finding that pretrial detention cannot toll supervised release.183 In concluding that defendant Brian Marsh’s pretrial detention did not toll his supervised release, however, the court did not focus on the phrase “in connection with a conviction.”184 Instead, the opinion proclaimed that its “conclusion rest[ed] on a word that [its] sister circuits and the parties have appeared to ignore—the word ‘is.’”185

The court held that Congress’s use of the present tense “foreclosed the type of backward-looking tolling analysis” that the majority position allowed.186 According to the DC Circuit, the courts expounding the majority position erred in relying on the sentence-credit provision to create the requisite connection.187 The court maintained that upon receiving a credit for time spent in pretrial custody, “it might be appropriate to say that the person was imprisoned or has been imprisoned ‘in connection with a conviction,’” but certainly not that the defendant “is” imprisoned.188 In coming to this conclusion, the DC Circuit relied on 1 USC § 1, which provides that “[i]n determining the meaning of any Act of Congress, unless the context indicates otherwise[,] . . . words used in the present tense include the future as well as the present,”189 and Carr v United States,190 in which the Supreme Court noted that “the present tense generally does not include the past.”191

The Marsh court also dispensed with the majority position’s concern that a contrary interpretation would render parts of § 3624(e) mere surplusage. The court noted that its interpretation gives effect to each of the provision’s terms.192 The phrase “in connection with a conviction” expresses that the statute operates to toll supervised release only when a person’s imprisonment is “triggered by a conviction.”193 The phrase “during any period,” meanwhile, “clarifies that a term of supervised release is tolled not only during the period of imprisonment initially imposed upon conviction, but also any additional period of imprisonment flowing from a conviction.”194

Having addressed the interpretive concerns expressed in the opinions adopting the majority position, the DC Circuit noted that its interpretation also makes the most practical sense.195 The court was critical of the majority position; the panel believed it condoned situations in which district courts would “be unable to determine whether they retain jurisdiction over defendants.”196 The court could not find “any other area of law in which district court jurisdiction [was] similarly contingent on future events,” and was concerned that such a situation would be unfair to defendants “who would have no idea whether they continue[d] to be subject to court supervision.”197 The requirement that a warrant or summons issue prior to the expiration of the supervised release term, the court found, “provides fair notice to the defendant and certainty for all.”198

III. A Purpose-Driven Resolution

Substantial confusion remains as to whether pretrial detention can toll supervised release. This Comment argues that, in order to gain clarity, it is first important to acknowledge the ambiguity of the supervised release tolling provision. Doing so suggests that courts should investigate the legislative history surrounding not only the supervised release tolling provision, but also the statutes with which its interaction has created such confusion.199 Part III.A examines the deficiencies of the majority and minority position alike, demonstrating that further analysis is necessary to determine the proper interpretation of § 3624(e). Part III.B shows that § 3624(e) is, in fact, ambiguous and, at the very least, does not unambiguously carry the meaning the majority position proposes. Part III.C then examines the legislative intent behind the creation of the statutory tolling provision. Finally, Part III.D discusses what the legislative intent behind both pretrial detention and the sentence-credit provision tells us about the proper conception of the supervised release tolling provision.

A. Schrödinger’s Cell and an Insufficient Response

The circuits have divided in the two decades after the Ninth Circuit’s attempt to determine whether pretrial detention may toll supervised release. The one thing on which all of the circuits agree, however, is also the one thing about which they are each mistaken—that the text of § 3624(e) is unambiguous. Because each of these courts found the statutory text clear, their an­alyses went no deeper than the text of the tolling provision. Al­though the courts have come to opposite conclusions, each position is deficient.

1. Schrödinger’s cell.

The majority position’s acceptance of the uncertainty it creates for courts and defendants alike is troubling. First, the claim that this problem is “rare” seems questionable, at the very least, given that six of the federal courts of appeals have had cause to confront the issue. The fact that over one in three defendants sentenced to supervised release are arrested on a new charge within three years supports this belief.200 When combined with the length of the average term of supervised release, it is clear that a large number of defendants will be simultaneously in pretrial detention and supervised release. When supervised release is required by statute, the average length is eighty-two months, and when supervised release is added at the discretion of the sentencing court, the average length is thirty-five months.201

Second, and more importantly, the majority position underestimates the degree of uncertainty its result creates. It both underestimates the cost of the jurisdictional uncertainty it creates for courts, and it completely ignores the costs of uncertainty for criminal defendants. With respect to supervised release terms, temporal limits are also jurisdictional limits. Asking a court to commit already strained judicial resources to cases over which its jurisdiction is uncertain is inappropriate.202 As one commentator put it, “[u]ncertainty also leads to mistaken jurisdictional assumptions and exercises of authority, which, if later discovered, will undo all of the effort expended.”203 At the very least, “jurisdictional uncertainty can surely lead to both a waste of judicial time and added expense to the litigants.”204 Thus, there is certainly a “reason” that resource-conscious district courts would not want to simply “continue the proceedings until a conviction or an acquittal is rendered in the other case.”205

The uncertainty created by the majority position weighs even more heavily on the defendant. A defendant who has been arrested while serving a supervised release term is placed in Schrödinger’s cell. Just like an atom exists in both of Schrödinger’s boxes until one lid is lifted and the atom is observed, until a defendant receives judicial observation—that is, a verdict—his supervised release both is and is not running. The sentencing court has control over whether it will commit resources to initiate revocation proceedings despite the indeterminacy, but the defendant is entirely powerless.

Uncertainty is especially severe in the shadow of a plea bargain. Plea deals essentially present a defendant with two options: take a certain level of punishment or bet that you will receive no punishment at the risk of receiving a substantially greater amount of punishment. The certain punishment and the uncertain punishment, however, are not independent. The higher the uncertain punishment is, the higher a prosecutor will be able to set the level of certain punishment.206 When it is uncertain whether a defendant’s release term has been tolled and thus whether he is at risk of revocation, a great amount of potential punishment may be added to the already substantial punishment the defendant is facing for the instant offense.

Remember Defendant X from the Introduction. He is serving a two-year supervised release term for possession. Sixty days before his release term is set to expire he is arrested for larceny and held pretrial. Sixty-one days into his pretrial detention, the prosecutor comes to him with a plea bargain. In exchange for pleading guilty to the larceny charge with an understanding that the prosecutor will recommend a mitigated sentence, the prosecutor also promises to help reduce X’s exposure to punishment for his violation of his release term.207

Recall that the court with jurisdiction over X’s supervised release term may or may not decide to revoke X’s supervised release.208 As noted in Part I.B.2, however, revocation can only occur after a hearing at which a US Attorney will prosecute the allegation of a supervised release violation. The prosecutor in X’s larceny case may therefore make X a range of offers. The larceny prosecutor may assure X that, should he agree to the larceny plea bargain, he will work with the US Attorney prosecuting his revocation hearing to recommend to the presiding sentencing court that X be given a mitigated sentence for his violation of supervised release.209 He may also offer to consult with X’s probation officer and ask her to refrain from reporting the violation to the relevant sentencing court,210 or encourage the probation officer to offer mitigating evidence at X’s revocation hearing. The larceny prosecutor may even assure X that he will reach out to the sentencing court that would preside over any potential revocation proceedings and urge the judge to refrain from revoking X’s supervised release. X must decide whether any of these offers are worth factoring in to the plea bargain he has been offered in the larceny case.

Already faced with the difficulties of mounting a defense while detained pretrial,211 X now must reckon with the apparent dual state of his detention. Schrödinger’s cell places him both in and not in danger of revocation proceedings that might result in substantial additional incarceration. Instead of simply balancing the likelihood that he will be convicted of larceny at trial and the sentence he is likely to receive against the plea bargain he has been offered, X now faces a different calculus. He must add the likely sentence he will receive at a revocation hearing to the sentence he is likely to receive if convicted of larceny and balance both against the bargain the prosecutor offered. Indeterminacy with respect to the risk of revocation proceedings is undeniably a powerful tool for securing a guilty plea in a prosecutor’s already-full toolbox.212 The prospect of revocation proceedings and the resulting additional prison sentence allows the prosecutor to create a greater downside risk to a defendant’s refusal to agree to a plea bargain. This added risk might allow the prosecutor to obtain a harsher plea bargain or to more quickly obtain a guilty plea.

While projected trial results certainly color how defendants understand plea bargains, other factors also weigh on the decision. For instance, “lawyers’ risk aversion makes them shy away from [ ] uncertainty.”213 Defense attorneys who are uncertain whether a court maintains jurisdiction over their client’s supervised release term may err on the side of caution. That is, in presenting options to their client, they may weigh the prospect of supervised release revocation more heavily than is warranted and give greater weight to offers from a prosecutor to work to mitigate any potential punishment for a supervised release violation. This is important because, “[t]hough appointed counsel may enjoy less trust, clients on average give lawyers’ recommendations a great deal of weight.”214

The uncertainty a defendant faces is not limited to plea bargaining. Imagine Defendant Y who, thirty days before his supervised release is set to expire, is arrested on a new charge. He sits in pretrial detention for thirty-one days until he finally comes up with the money to post bond. Now released, he cannot know whether, over the next thirty days, he remains under supervision or not. Several “standard conditions” of supervised release make this problematic.215 Should he seek permission to leave the judicial district?216 Must he open his home and personal effects to a probation officer?217 Must he maintain full time employment or seek excusal from that requirement?218 Violation of any condition could result in revocation. The uncertainty the majority position creates places defendants in an untenable position, even once outside Schrödinger’s cell.

2. The Marsh court’s insufficient response.

Although a succinct critique of the majority position, the Marsh court’s opinion itself is vulnerable to criticism. In holding that 1 USC § 1 demands a forward-looking construction, it relied on Carr, a case in which the Court considered a law that criminalized the failure of a sex offender to update his registration information if he traveled in foreign or interstate commerce.219 The question presented in that case was whether the law applied to offenders who traveled prior to the act’s enactment, and, provided it did, whether the law ran afoul of the Constitution’s prohibition of ex post facto laws.220 The interpretive question presented in Marsh and its predecessors is distinguishable. The question is not to whom does the tolling provision apply, but to what kinds of “imprisonment.” While the Court’s analysis seems well-fitted to questions of a statute’s retroactivity, its use to analyze a tolling provision seems vulnerable to the criticism Justice Samuel Alito leveled in dissent in Carr. Although it is obvious that laws phrased in present tense do not include the past, for Alito such a simplistic argument ignores the salient question: “At what point in time does [the statute] speak?”221 This is especially relevant to the analysis of a tolling provision, because an inquiry into whether a statute or term of punishment was tolled is inherently backward-looking. That is, a judge uses a tolling provision to understand how a defendant’s past status affects what the law has to say about her court’s present jurisdiction over a supervised release term. Does the law speak at the point of the pretrial detention, or does it speak at the point in time when the judge undertakes the tolling analysis? If it is the former point in time, the Marsh court’s conclusion holds true, but not if it is the latter. This is all to say that the use of present tense may not lend § 3624(e) the kind of clarity the DC Circuit claims it does. Instead of the tense of the statute, a simpler analysis focuses directly on the question whether Congress intended to include pretrial detention in the kinds of incarceration that could toll supervised release.

B. Statutory Ambiguity

How a court should determine whether a statute is ambiguous is, well, ambiguous.222 At least one federal judge has complained that “there is often no good or predictable way for judges to determine whether statutory text contains ‘enough’ ambiguity” to allow the use of legislative history as an interpretive aid.223 Some commentators have suggested that statutory ambiguity exists on a sliding scale.224 Where exactly on that scale a statute becomes clear as opposed to ambiguous, however, is in no way a settled rule.225 This has led some to propose that the test for ambiguity should be whether “a lawyer would litigate the issue in court,”226 while another has argued that the distinction between clarity and ambiguity should be abandoned altogether and instead asks that courts simply “strive to find the best reading of the statute.”227

This Comment does not attempt to apply a definitive test for ambiguity. Instead, it compiles the evidence that tends to weigh in favor of a finding of statutory ambiguity. First, the very fact that six circuits have claimed the text is clear on its face and yet have developed contrary interpretations intuitively belies the notion that the statute is clear. Predictably, perhaps, there is nevertheless a circuit split about whether a circuit split surrounding the proper interpretation of a statute is evidence of statutory ambiguity.228 Additionally, although the majority position relies on the lack of “an imprisonment-detention distinction” within the US Code to buttress its finding that § 3624(e) unambiguously points to its result,229 it seems instead to point to ambiguity. At the very least, statutory text that is used inconsistently undermines a finding of statutory clarity.230

Another way to approach the question whether a statute is ambiguous is to apply what is claimed to be its unambiguous meaning. Should that application produce odd results, it is evidence that the statute lacks clarity, or, at the very least, that it does not carry the proposed unambiguous meaning.231 The odd ramifications of the majority position’s interpretation, therefore, point decidedly away from a finding that it is a proper interpretation. All of the courts that have considered the question discussed the uncertainty into which that interpretation places federal district-court judges, and this Comment has demonstrated the more troubling uncertainty into which that interpretation places criminal defendants.232

Yet another odd result further undermines the majority position. Consider three hypothetical situations. In Situation A, Defendant X is arrested five days prior to the expiration of his supervised release term and is detained pretrial. After ten days in jail, he decides to plead guilty to the new charge. He is sentenced to one year in prison, but this sentence is credited with his ten days of pretrial detention by virtue of the mandatory sentence-credit provision. By the majority position’s logic, his ten days in jail have now become connected with his new conviction, and yet no tolling can have occurred because § 3624(e) states that an imprisonment cannot toll a supervised release term if it “is for a period of less than 30 consecutive days.”233 In Situation B, Defendant X is again arrested five days prior to the expiration of his supervised release, and is detained for ten days. This time, however, his savvy defense attorney secures his release pretrial, though he is, in the end, still convicted of this new offense and sentenced to a year in prison. Again, his ten days of pretrial detention will be credited to his sentence, but no tolling has occurred. Finally, in Situation C, imagine either Situation A or B and simply change the amount of time spent in detention from ten days to thirty-one. In C, X’s supervised release would be tolled because his time in detention exceeded thirty consecutive days.

It is difficult to imagine a possible justification Congress could have had for creating a tolling provision that encompasses only periods of pretrial detention that last longer than thirty days. In reporting the tolling provision, the Senate remarked that its limitation was to ensure that “brief period[s]” of imprisonment served “as a condition of probation or supervised release” did not toll supervised release terms.234 The absence of tolling for those situations is sensible. After all, if brief periods of detention result from conditions placed on supervised release by a judge,235 he is able to account for that in determining the length and design of the term. That is, if he chooses to punish a violation of a minor condition of the release term with short-term imprisonment (say fourteen days), the judge can know that the term will not be tolled and that the overall structure of the term will not be affected. A judge, however, is not able to foresee pretrial detention for as-yet-uncommitted offenses,236 and therefore cannot, ex ante, design the term accordingly. Thus, the odd ramifications the limitation would have as applied to pretrial detention suggest that Congress did not consider it in crafting the tolling provision.

It is possible that Congress wanted to toll the terms of those who were deemed too dangerous to be released into the community in recognition that perhaps such a finding was evidence that the supervised release had not fulfilled its rehabilitative ends. That is, Congress may have contemplated Situation B and decided that tolling should apply only to those detained until trial as a good proxy for those on whom supervised release did not have the desired effect. Situation A, however, highlights the inadequacy of this justification. Whether a person chooses to quickly enter a guilty plea has little to do with how effective supervised release was in meeting its rehabilitative aims.

Another possible justification is that Congress believed periods of time less than thirty days were insignificant disruptions to supervised release terms and should not toll such terms. This justification seems in line with the few comments the Supreme Court has made about supervised release, namely that it was designed to assist defendants in their transition to community life.237 As explained below, however, the tolling provision’s insistence that imprisonment be connected with a conviction shows that this justification is insufficient.238 Without an adequate justification, the seemingly senseless distinction between pretrial detention that lasts less than thirty days and those that are longer than thirty days weighs in favor of a finding of statutory ambiguity and, at the very least, suggests that the statute does not unambiguously require the majority position’s interpretation.239

Taken together, the evidence of statutory ambiguity is significant. The considerable confusion among the circuits, the interpretive errors of the Ninth and DC Circuits, the uncertainty to which the majority position subjects judges and defendants alike, and the several odd results of the majority position all suggest a lack of statutory clarity. This ambiguity makes it possible to examine how the legislative history informs the proper interpretation of § 3624(e).

C. Legislative Intent: Supervised Release Tolling Provision

As noted above, the Supreme Court has weighed in on Congress’s intent in creating supervised release.240 In Johnson, the Court stated that “Congress intended supervised release to assist individuals in their transition to community life.”241 The Court reasoned that such an intent was inconsistent with supervised release running concurrently with a prison term. The Court’s belief was that while a person is imprisoned she is not engaging in the “transition to community life” and, therefore, allowing supervised release to run concurrently would be contrary to the purpose of sentencing her to supervised release in the first place.242

Of course, Congress’s intent in creating supervised release was not necessarily the same intent it had when it enacted the tolling provision. Nevertheless, the Court’s finding necessitates a response to the obvious argument that this logic naturally extends to pretrial detention. That is, it could be said that while a person is in pretrial detention, they are similarly incapable of transitioning to community life.

This argument has one major weakness. If Congress intended for supervised release to run only while a defendant was transitioning to community life, its insistence on a connection with a conviction in order for a term to be tolled makes little sense. Defendants held pretrial are removed from their communities regardless of their guilt or innocence. Some defendants in pretrial detention will never be convicted, yet, while they are in custody they too are unable to engage in the transition to community life. Nevertheless, absent a conviction on the underlying charge, the supervised release terms of such defendants will not be tolled. The connection between presence in the community and the running of supervised release is tenuous at best, and it seems unlikely that the Court’s finding can fully explain Congress’s intent. It thus remains important to examine the legislative history surrounding each of the statutes in question in an attempt to divine whether Congress intended pretrial detention to toll supervised release.

It is evident from a review of the legislative history surrounding the Sentencing Reform Act of 1984 that the paramount concern of legislators was to inject more certainty into the length and contours of criminal sentencing.243 In particular, it is clear that the creation of the supervised release term was done to allow judges to have control over who is supervised after release and for how long, in contrast with the administratively regulated parole system.244 This was seen as preferable to a system that based post-release supervision on good time served.245 Although the tolling provision as enacted in 1984 could be seen as far-reaching (after all, the drafters noted that supervised release would be tolled whenever a defendant “is imprisoned other than during limited intervals as a condition of probation or supervised release”),246 it is the 1986 amendments and the history surrounding them that most illuminate the tolling provision as it currently reads.

In reporting the tolling provision amendment, the House Report stated its purpose was to make clear that “a term of supervised release does not run while the person is serving a term of imprisonment in excess of 30 days for any offense.”247 The Sixth Circuit, the court that first developed the majority position, noted in Goins that “in connection with” is distinct from “for,” implying that the latter would indicate Congress intended for tolling only when imprisonment resulted from or was otherwise triggered by a conviction.248 Thus, legislative history indicating that Congress believed “in connection with” was interchangeable with “for” weighs in favor of an interpretation in line with that proposed by the Ninth and DC Circuits. This is especially true if viewed in the light of the overarching purpose of the Sentencing Reform Act. A Congress intent on creating certainty in federal sentencing law is especially unlikely to enact a tolling provision that results in substantial indeterminacy.249 Instead, Congress set forth a simple mechanism by which a district court can provide itself the necessary time to adjudicate revocation proceedings. It simply asked that the defendant be provided fair notice by means of a warrant or summons prior to the expiration of his supervised release.250

D. Legislative Intent: Bail Reform and the Sentence-Credit Provision

The above discussion shows that tolling supervised release for pretrial detention seems to lack a justification in the stated purposes of the Sentencing Reform Act. The purpose of bail reform further bolsters the argument that pretrial detention should not toll supervised release. An accurate understanding of pretrial detention precludes the use to which the majority position puts the sentence-credit provision. Furthermore, an examination of congressional intent in creating the sentence-credit provision shows that the majority position results in a situation directly contrary to that purpose.

1. The nonpenal nature of pretrial detention.

Although Congress’s clear intent to create certainty in sentencing lends credence to the view that pretrial detention should not toll supervised release, relying on the House Report’s comment that tolling occurs when a person is imprisoned “for” a conviction remains somewhat vulnerable. It could be argued, for instance, that the comment changes little because once the defendant is convicted and, by operation of 18 USC § 3585, receives a sentence credit for her time in custody, that period essentially becomes time served “for” her conviction. This argument rests on a fundamental misunderstanding of how pretrial detention and the sentence-credit provision operate. The notion that time served in pretrial detention could ever be “for” an offense is based on the idea that the sentencing credit retroactively changes pretrial detention into punishment for the resulting conviction. This interpretation of the sentence-credit provision is sensible, especially given the initial impetus for the provision was “to eliminate the disparity in sentences under certain statutes requiring mandatory terms of imprisonment.”251 An understanding of congressional intent and Supreme Court precedent surrounding the Bail Reform Act of 1984, however, forecloses this interpretation.

Prior to the Bail Reform Act of 1984, “the sole purpose of bail laws [was] to assure the appearance of the defendant at judicial proceedings.”252 A major undertaking of the 1984 Act, however, was to expand the purposes that a court must consider in conducting pretrial proceedings. The Act added one significant purpose—the need to consider community safety. The Supreme Court in Salerno made clear that there is a distinction between pretrial detention and sentences of imprisonment. Pretrial detention, the Court said, cannot be punitive.253 The Due Process clause does not allow “a detainee [to] be punished prior to an adjudication of guilt.”254

Seen in this light, it is unlikely that Congress intended, by enacting the sentence-credit provision, to transform pretrial detention to punitive detention. Nor, according to the Salerno Court, would that purpose have been constitutionally permissible.255 Instead, the sentence-credit provision might be seen as a policy choice in favor of limiting incarceration. Or, perhaps, the sentence-credit provision could be a simple recognition of the factual (as opposed to legal) similarity between pretrial and postsentence incarceration, and an instance of government mercy.

This latter interpretation is supported by cases that have sought to determine whether there is a constitutional guarantee of a sentence credit for pretrial detention. As one court put it, “[a]s a general rule, a state prisoner has no federal constitutional right to credit for time served prior to sentence absent a state statute granting such credit.”256 Some courts find an exception to this general rule “when the pre-sentence time together with the sentence imposed is greater than the statutory maximum penalty for the offense,”257 or more narrowly when “a criminal defendant [ ] is confined before sentencing because his indigency prevents him from making bond.”258 In both cases, courts have held that a sentence credit is mandated by the Fourteenth Amendment when a defendant is sentenced to the statutory maximum.259 The history of the federal sentence-credit provision,260 describing how the provision was initially limited to mandatory sentences but then expanded to reach any detention in relation to any offense, is also informative. That Congress was willing to provide sentence credits in cases beyond its constitutional mandate suggests an act of legislative grace.

This understanding raises a more pressing concern about the majority position. That position causes pretrial detention to effectively lengthen punitive supervised release terms, for al­though supervised release has rehabilitative aims,261 it is undoubtedly a form of punishment that meaningfully constrains defendants and places them under the power of the federal courts.262 That is to say, when a judge places a defendant who is subject to supervised release in pretrial detention, thereby lengthening the time that he is under the initial sentencing court’s power, that defendant’s punishment is extended by the operation of the tolling provision. This observation is enough to raise constitutional concerns surrounding the provision. If the majority position essentially converts periods of pretrial detention into punitive detention, that would violate the clear command that, under the Due Process clause, “a detainee may not be punished prior to an adjudication of guilt.”263 The Court has consistently instructed lower courts to avoid interpretations that may raise constitutional concerns when another plausible interpretation is available.264 In other words, “courts will [ ] not lightly assume that Congress intended to infringe constitutionally protected liberties.”265 Given that an interpretation which does not condone the tolling of supervised release terms for periods of pretrial detention is not only available, but eminently sensible, courts should favor its adoption.

Some commentators maintain that the Court was mistaken, or even disingenuous, in Salerno and that pretrial detention is indeed punitive.266 Whether you believe the Court or not matters little for the purposes of this controversy. The government in Salerno was emphatic that Congress did not intend to make pretrial detention punitive by enacting the Bail Reform Act of 1984.267 The government’s position in Salerno essentially estops it from taking the contrary position here in order to put pretrial detention to use as a tolling mechanism for supervised release terms. Because “[w]here a party assumes a certain position in a legal proceeding, and succeeds in maintaining that position, he may not thereafter, simply because his interests have changed, assume a contrary position,” the government may have to choose which conception of pretrial detention it prefers to defend.268 If it chooses to argue for an interpretation that allows periods of pretrial detention to be converted to punitive incarceration, it might need to argue that that pretrial detention should be an exception from the typical pre-adjudication bar on punishment guaranteed by the Due Process clause. If the constitutionality of pretrial detention on the basis of predicted danger to the community depends on the government taking the position that such detention is nonpunitive, it may not insist on its availability while simultaneously asking courts to interpret the tolling provision in a manner making the very same detention punitive in fact. Thus, whatever Congress’s true intentions with respect to the availability of pretrial detention to protect communities from the as-yet-uncommitted crimes of defendants, Salerno forecloses the majority position’s interpretation.

2. The sentence-credit provision and unequal exposure to punishment.

The preceding Section showed that the purpose of the sentence-credit provision is decidedly not to convert pretrial detention into penal imprisonment. This Section focuses on what the purpose is. When Congress first passed a sentence-credit provision, it did so “to eliminate the disparity in sentences under certain statutes requiring mandatory terms of imprisonment.”269 The original sentence-credit provision directly referenced the cause of the disparity it corrected—those unable to afford bail often spent much longer in custody than their fellow convicts who were able to post bond. The statute has since become more capacious, reaching not only those in pretrial detention “for want of bail,”270 but also those held pretrial for any reason.271 Although expanded, perhaps to serve a broader purpose, it maintains its usefulness for correcting such disparities. Federal defendants may no longer be held pretrial on account of burdensome financial conditions,272 but that is not the case in many states.273

State pretrial detention, under the majority position, is just as capable of tolling supervised release as federal pretrial detention. The imprisonment that tolls supervised release may be connected with any state, federal, or local conviction according to § 3624(e). This, however, creates a situation that runs in direct conflict with the purpose of the sentence-credit provision. Imagine defendants X and Y, both of whom reside in a state that allows judges to set financial conditions that result in the detention of a defendant. Both X and Y are serving federal supervised release terms related to convictions for possession with intent to distribute that are five days from expiration. Unfortunately, X and Y are both arrested for larceny. Given their similarity, Judge A sets the bail for both at $5,000. Luckily for X, his rich uncle is able to come up with the money and he only has to spend ten days in jail. X is ultimately convicted and sentenced, and so receives a credit for the ten days spent in jail. No tolling has occurred by virtue of the thirty-day requirement and his supervised release term expired as scheduled. Y, however, has no rich uncle, and he sits in Schrödinger’s cell for forty days before his trial commences. He is convicted, sentenced, and receives credit for his pretrial detention. Under the majority position, his supervised release term has been tolled, and he is now exposed to revocation proceedings. In other words, his inability to afford bail has exposed him to additional punishment. The sentence-credit provision has been put to a use in direct conflict with its purpose.

The situation above provides a concrete illustration of a broader problem with the majority position—it is inappropriate to use the sentence-credit provision in a manner that produces different results on account of a difference in wealth. More broadly still, it seems in tension with a merciful congressional purpose to use the provision to expose defendants to greater punishment. A finding that pretrial detention cannot toll supervised release is more faithful to the purpose of the sentence-credit provision.

This odd result and the concerns expressed in the preceding Section show the virtue of dispensing with the notion that § 3624(e) is unambiguous. The ability to use legislative history to understand the entire network of statutes that gives rise to this problem (as opposed to narrowly focusing on § 3624(e)) shows us that the sentence-credit provision, the lynchpin of the majority position, cannot be properly interpreted as ever creating the requisite connection between pretrial detention and an eventual conviction. Pretrial detention, the sentence-credit provision, and the prison sentences imposed for convictions each serve a distinct purpose. Commingling these statutes without carefully considering the purposes each serve has led the majority of circuits to accept a troubling level of uncertainty. This Comment has instead laid out each statute’s purpose and, in doing so, shown how allowing pretrial detention to toll supervised release is inconsistent with legislative intent.


The operation of the supervised release tolling provision contained in 18 USC § 3624(e), and its interaction with the statutes concerning pretrial detention and the sentence-credit provision contained in 18 USC § 3585,274 has given rise to substantial confusion among the federal circuits. No satisfactory answer has been provided to the question whether pretrial detention can toll supervised release. According to the majority of courts that have considered the question, defendants held in pretrial detention who are serving supervised release terms may as well be sitting in Schrödinger’s cell. As they await their trial, they are in a profound state of indeterminacy, simultaneously subject to and not subject to the power of their sentencing courts. Only upon the disposition of their instant cases will they gain certainty about their status. The DC and Ninth Circuits, while coming to the right conclusion, offer insufficient rebuttals to the Fourth, Fifth, Sixth, and Eleventh Circuits. The fatal error in the reasoning of all six circuits, however, is the same—each assumes that the text of 18 USC § 3624(e) is unambiguous.

Accepting that the supervised release tolling provision is ambiguous will encourage courts to examine the legislative history surrounding each of the relevant statutes. A thorough inquiry into that history reveals that Congress could not have intended for pretrial detention to toll a supervised release term. This Comment has shown that the sentence-credit provision, on which the majority relies to forge the requisite connection, can never properly be viewed as creating a connection between pretrial detention and a later conviction. Instead, Congress intended for the term of supervised release to be tolled only when a defendant is serving a prison term resulting from or otherwise triggered by a conviction.

Hundreds of thousands of citizens are currently serving supervised release terms.275 When they are arrested on new charges and held in pretrial detention, their predicament is one of substantial uncertainty. These defendants, faced with the enormous pressures associated with pretrial detention and pending criminal charges, must also be concerned with whether they remain under the power of their sentencing court. They deserve the kind of certainty Congress attempted to create in adopting the Sentencing Reform Act of 1984.

  • 1See Bruce Rosenblum and Fred Kuttner, Quantum Enigma: Physics Encounters Consciousness 97 (Oxford 2d ed 2011).
  • 2For an experimental illustration of this “quantum enigma,” see id at 87–97.
  • 3Id at 146–48.
  • 4Pub L No 98-473, 98 Stat 1976.
  • 5See Comprehensive Crime Control Act § 203, 98 Stat at 1976–80, 18 USC § 3142(c), (e)–(f).
  • 6Comprehensive Crime Control Act § 212, 98 Stat at 1999–2000, 18 USC § 3583.
  • 718 USC § 3624(e) (emphasis added).
  • 818 USC § 3583(d).
  • 918 USC § 3583(e)(3).
  • 10USSG § 7B1.3(f). While the Sentencing Guidelines no longer carry the force of law after the 2005 case United States v Booker, 543 US 220, 245 (2005), it appears this is an oft-followed policy. See, for example, United States v Jaimes-Benitez, 644 Fed Appx 299, 300 (5th Cir 2016); United States v Smith, 571 Fed Appx 938, 939–40 (11th Cir 2014); United States v Day, 2012 WL 6019113, *4 & n 1 (WD Ark).
  • 11See 18 USC § 3583(e)(3).
  • 1218 USC § 3583(e)(2).
  • 13The statutes governing probation contain an identical tolling provision. See 18 USC § 3564(b). Because no court has analyzed this question as it relates to probation, this Comment focuses on supervised release. Nevertheless, the analysis is identical in both scenarios, and the dual purpose to which this statutory language is put only heightens the importance of finding a resolution.
  • 14According to the terms of § 3624(e), it does not matter whether this crime is federal, state, or local.
  • 15If the second crime is a federal crime, the sentence credit would be required by 18 USC § 3585(b) (“A defendant shall be given credit toward the service of a term of imprisonment for any time he has spent in official detention . . . as a result of the offense for which the sentence was imposed.”). Almost every state has a similar required credit. See Arthur W. Campbell, Law of Sentencing § 9:28 at 444–45 & nn 2–3 (West 3d ed 2004).
  • 16See Joshua A. Markman, et al, Recidivism of Offenders Placed on Federal Community Supervision in 2005: Patterns from 2005 to 2010 *3 (Bureau of Justice Statistics, June 2016), archived at
  • 17See id.
  • 18See Part II.C. See also United States v Goins, 516 F3d 416, 422–23 (6th Cir 2008); United States v Molina-Gazca, 571 F3d 470, 473–74 (5th Cir 2009); United States v Johnson, 581 F3d 1310, 1311–12 (11th 2009); United States v Ide, 624 F3d 666, 669–70 (4th Cir 2010).
  • 19See Part II.B. See also United States v Morales-Alejo, 193 F3d 1102, 1105 (9th Cir 1999) (“A plain reading of [the tolling provision] suggests that there must be an imprisonment resulting from or otherwise triggered by a criminal conviction.”).
  • 20See Part II.D. See also United States v Marsh, 829 F3d 705, 709 (DC Cir 2016) (reasoning that the statute’s use of the present-tense expression “is imprisoned in connection with a conviction” renders it inapplicable to pretrial detention preceding a conviction).
  • 21See 18 USC §§ 3142, 3585.
  • 22Comprehensive Crime Control Act of 1983, S Rep No 98-225, 98th Cong, 1st Sess 1 (1983), reprinted in 1984 USCCAN 3182, 3184.
  • 23Id at 2 (quotation marks omitted).
  • 24Pub L No 98-473, 98 Stat 1987.
  • 25Pub L No 98-473, 98 Stat 1976, codified in various sections of Title 18.
  • 26See June Carbone, Seeing through the Emperor’s New Clothes: Rediscovery of Basic Principles in the Administration of Bail, 34 Syracuse L Rev 517, 519–21, 529–32 (1983).
  • 27See id at 520–21. As initially conceived, the amount required of the surety was equal to the fine that would be imposed if the defendant were convicted. Id. The surety was responsible for paying that fine in full if conviction in fact resulted. Id.
  • 28See id (describing the Anglo-Saxon bail system as “perfectly designed”). See also Betsy Kushlan Wanger, Note, Limiting Preventive Detention through Conditional Release: The Unfulfilled Promise of the 1982 Pretrial Services Act, 97 Yale L J 320, 323 n 19 (1987) (explaining that, in medieval England, pretrial detention without bail could be extremely prolonged because magistrates traveled from county to county and, consequently, were only in particular towns for a few months every year).
  • 29Laura I. Appleman, Justice in the Shadowlands: Pretrial Detention, Punishment, & the Sixth Amendment, 69 Wash & Lee L Rev 1297, 1324 (2012).
  • 30Several colonial charters, in fact, included a guaranteed right to bail. Id at 1325.
  • 31US Const Amend VIII.
  • 32There is some debate as to whether the omission of a right to bail was deliberate or a “historical accident.” See Appleman, 69 Wash & Lee L Rev at 1326 & n 145 (cited in note 29).
  • 33See, for example, Stack v Boyle, 342 US 1, 4 (1951) (“Unless this right to bail before trial is preserved, the presumption of innocence, secured only after centuries of struggle, would lose its meaning.”).
  • 34See John N. Mitchell, Bail Reform and the Constitutionality of Pretrial Detention, 55 Va L Rev 1223, 1225–27 (1969) (noting a “pervasive practice of denial of bail in capital cases when the eighth amendment was ratified in 1791”).
  • 35Appleman, 69 Wash & Lee L Rev at 1329 (cited in note 29).
  • 36Wanger, Note, 97 Yale L J at 324 (cited in note 28).
  • 37Id.
  • 38Pub L No 89-465, 80 Stat 214, codified in various sections of Title 18.
  • 39See Wanger, Note, 97 Yale L J at 325 (cited in note 28).
  • 4018 USC § 3142(c)(2). This is not the case, however, in the many states in which onerous financial conditions may still lead to pretrial detention. See, for example, Gabriel Loupe, Comment, The Lack of Money Is the Root of All Evil: Louisiana’s Ban on Bail without Surety, 77 La L Rev 109, 114, 138–39 (2016) (concluding that a Louisiana law banning recognizance bonds for arrestees charged with certain drug offenses “allows for a situation in which the indigent may languish in jail while their peers, identical to them in all regards save wealth, are freed pending trial”).
  • 41Stack, 342 US at 5.
  • 42Id.
  • 43Anti-Crime Act of 1984, HR 5690, 98th Cong, 2d Sess, in 130 Cong Rec 28595 (Oct 2, 1984) (statement of Rep Sawyer).
  • 44See S Rep No 98-225 at 3 (cited in note 22) (describing assuring both community safety and defendants’ appearances at trial as purposes of the legislation).
  • 45Federal Bail Procedures, Hearings on S 1357 before the Subcommittee on Constitutional Rights and the Subcommittee on Improvements in Judicial Machinery of the Senate Committee on the Judiciary, 89th Cong, 1st Sess 3 (1965) (“1965 Bail Hearings”) (statement of Sen Ervin).
  • 46S Rep No 98-225 at 3 (cited in note 22).
  • 47Id.
  • 48481 US 739 (1987).
  • 49Id at 745 (“We think that respondents have failed to shoulder their heavy burden to demonstrate that the Act is ‘facially’ unconstitutional.”).
  • 50Id at 754–55.
  • 51Id at 755. Whether pretrial detention has remained, or ever was, the “carefully limited exception” to which the Salerno majority referred is debatable. In fiscal years 2008 through 2010, for instance, only 36 percent of defendants appearing before federal district courts were released prior to trial. Thomas H. Cohen, Pretrial Release and Misconduct in Federal District Courts, 2008–2010 *1 (Bureau of Justice Statistics, Nov 2012), archived at
  • 52Bell v Wolfish, 441 US 520, 535 (1979). See also Ingraham v Wright, 430 US 651, 674 (1976) (noting that detainees’ liberty interests are protected by the Due Process Clause rather than the Eighth Amendment); Kennedy v Mendoza-Martinez, 372 US 144, 186 (1963) (identifying, “dating back to the Magna Carta,” a “cherished tradition” that “punishment cannot be imposed without due process of law”) (quotation marks omitted); Stack, 342 US at 4 (“The traditional right to freedom before conviction . . . serves to prevent the infliction of punishment prior to conviction.”); Wong Wing v United States, 163 US 228, 237 (1896) (requiring a trial to establish guilt before alien detainees could be subjected to hard labor); Hudson v Parker, 156 US 277, 285 (1895) (“The statutes of the United States have been framed around the theory that a person accused of crime shall not, until he has been finally adjudged guilty in the court of last resort, be absolutely compelled to undergo imprisonment or punishment.”).
  • 53Salerno, 481 US at 747 (emphasis added).
  • 541965 Bail Hearings, 89th Cong, 1st Sess at 3 (cited in note 45) (statement of Sen Ervin).
  • 55See 18 USC § 3142(a).
  • 5618 USC § 3142(a)(1).
  • 5718 USC § 3142(f).
  • 5818 USC § 3142(e).
  • 5918 USC § 3142(g)(1)–(3).
  • 6018 USC § 3585(b).
  • 6118 USC § 3585(b).
  • 6218 USC § 3624(e).
  • 63In most jurisdictions, like in the federal system, a statute mandates that a sentence credit be given for any detention in relation to an offense. See Campbell, Law of Sentencing § 9:28 at 444–45 & nn 2–3 (cited in note 15). In the “handful of states [that] leave determination of time-served credit to the discretion of sentencing judges,” it would be necessary to determine whether the judge has actually awarded a credit for pretrial detention before one could determine whether the controversy described here is even implicated. Id at 444.
  • 64S Rep No 98-225 at 49–50 (cited in note 22).
  • 65See United States v Booker, 543 US 220, 245 (2005).
  • 66See Sentencing Reform Act § 212, 98 Stat at 1988–89, 2017–26, 18 USC §§ 994(a), 3553(a)–(b). See also note 10.
  • 67See Sentencing Reform Act § 217, 98 Stat at 1999–2000, 18 USC § 3583.
  • 68S Rep No 98-225 at 39 (cited in note 22).
  • 69Id.
  • 70Note, Parole Release Decisionmaking and the Sentencing Process, 84 Yale L J 810, 820, 828 (1975).
  • 71Id at 837.
  • 72Id at 839.
  • 73S Rep No 98-225 at 123 (cited in note 22).
  • 74See 18 USC § 3583(a).
  • 75See 18 USC § 3583(b) (listing maximum authorized terms of supervised release). See also text accompanying note 12 (noting the judge’s power to modify the term for the duration of supervised release).
  • 76See 18 USC § 3564(b).
  • 77See 18 USC § 3563(a).
  • 78See 18 USC § 3563(c).
  • 79See 18 USC § 3565(c).
  • 80Compare Roberts v United States, 320 US 264, 272 (1943) (“[T]he basic purpose of probation [is] to provide an . . . offender an opportunity to rehabilitate himself without institutional confinement.”), with United States v Johnson, 529 US 53, 59 (2000) (“Supervised release fulfills rehabilitative ends, distinct from those served by incarceration.”).
  • 81Sentencing Reform Act § 212, 98 Stat at 2009.
  • 82S Rep No 98-225 at 148–49 (cited in note 22) (noting that the provision “specifies that the term begins on the date of release and that it runs concurrently with any other term of supervised release, probation, or parole unless the person is imprisoned other than for a brief period as a condition of probation or supervised release”).
  • 8318 USC § 3624(e).
  • 84Minor and Technical Amendments to the Comprehensive Crime Control Act of 1984, S Rep No 99-278, 99th Cong, 2d Sess 3 (1986). The actual report contains an error, stating that the amendment was intended to bring the supervised release tolling provision into conformity “with a similar provision about probation in 18 U.S.C. 3624.” Id. Given that the amendment was to § 3624, it is clear that the report intended to reference 18 USC § 3564, which contains the probation tolling provision.
  • 85Criminal Law and Procedure Technical Amendments Act of 1986, H Rep No 99-797, 99th Cong, 2d Sess 21 (1986) (emphasis added).
  • 86See Part III.B.
  • 8718 USC § 3583(a).
  • 88See 18 USC § 3583(c), citing 18 USC § 3553(a)(1), (a)(2)(B)–(D). While the decision as to the length of a term of supervised release is statutorily in the discretion of the court, 18 USC § 3583(b) creates upper limits for the duration terms of supervised release based on the class of felony or misdemeanor with which the term is associated.
  • 89See 18 USC § 3583(d), citing 18 USC § 3563(b). Common conditions include requirements that the defendant notify a judicial officer before leaving the judicial district, that he open his home and effects to a probation officer, and that he maintain full time employment. See USSG § 5D1.3(c)(3), (6), (7).
  • 9018 USC § 3583(d) (requiring further that “the defendant cooperate in the collection of a DNA sample,” “that the defendant not unlawfully possess a controlled substance,” and that the defendant “submit to a drug test within 15 days of release on supervised release and at least 2 periodic drug tests thereafter”).
  • 9118 USC § 3603(2).
  • 92Johnson, 529 US at 57. See also text accompanying notes 126–27.
  • 9318 USC § 3583(e)(1).
  • 9418 USC § 3583(e)(2). The court may only impose this condition “if less than the maximum sentence was previously imposed.” 18 USC § 3583(e)(2).
  • 9518 USC § 3583(e)(2).
  • 9618 USC § 3583(e)(3).
  • 97See 2016 Primer on Supervised Release *11–12 (US Sentencing Commission, June 2016), archived at (listing the appropriate court responses to each grade of defendant conduct). Some courts of appeals have held that probation officers may directly file petitions seeking revocation of supervised release and the initiation of revocation proceedings. See, for example, United States v Cofield, 233 F3d 405, 408–09 (6th Cir 2000).
  • 98FRCrP 32.1(b)(1).
  • 99See FRCrP 32.1(b)(2). See also 28 CFR § 2.50(a) (describing possible consequences of a violation).
  • 100See United States v Burnette, 980 F Supp 1429, 1434 (MD Ala 1997) (describing a typical hearing procedure).
  • 10118 USC § 3583(e)(3).
  • 10218 USC § 3583(i).
  • 103See, for example, United States v Molina-Gazca, 571 F3d 470, 473–74 (5th Cir 2009); United States v Johnson, 581 F3d 1310, 1311–12 (11th Cir 2009) (per curiam); United States v Ide, 624 F3d 666, 669–70 (4th Cir 2010).
  • 104See 18 USC § 3585(b).
  • 105Pub L No 86-691, 74 Stat 738 (1960).
  • 10674 Stat at 738.
  • 107Credit for Time in Custody Awaiting Trial, S Rep No 86-1696, 86th Cong, 2d Sess 3 (1960).
  • 108See id.
  • 109See id.
  • 110Sentencing Reform Act § 212, 98 Stat at 2001.
  • 11118 USC § 3585(b).
  • 112S Rep No 98-225 at 129 (cited in note 22).
  • 113See text accompanying notes 14–16.
  • 114Markman, et al, Recidivism of Offenders at *1, 3 (cited in note 16).
  • 115529 US 53 (2000).
  • 116Id at 54–55.
  • 117See id.
  • 118See id.
  • 119See Johnson, 529 US at 55–56.
  • 120Id at 58–60.
  • 121Id at 59.
  • 122Id.
  • 12318 USC § 3553(a)(2)(B)–(C). 18 USC § 3583(c) directs judges to consider the cited provisions in determining the propriety of supervised release.
  • 124See S Rep No 98-225 at 124 (cited in note 22) (describing the transition to community life and the rehabilitation of offenders as the primary goals of supervised release).
  • 125Johnson, 529 US at 59.
  • 12618 USC § 3624(e).
  • 127Johnson, 529 US at 57.
  • 128Id at 57–58.
  • 129Id at 58 (emphasis added).
  • 130Id (emphasis added).
  • 131See Part II.D.
  • 132193 F3d 1102 (9th Cir 1999).
  • 133Id at 1103.
  • 134Id.
  • 135Id. All federal circuits to consider the question have determined that deportation does not affect the running of a term of supervised release. See Thomas Nosewicz, Watching Ghosts: Supervised Release of Deportable Defendants, 14 Berkeley J Crim L 105, 109 & n 53 (2009). This is sensible given that the supervised release statutes specifically state, “If an alien defendant is subject to deportation, the court may provide, as a condition of supervised release, that he be deported and remain outside the United States.” 18 USC § 3583(d) (emphasis added).
  • 136Morales-Alejo, 193 F3d at 1103.
  • 137See id. Presumably because Morales-Alejo was not sentenced until a few weeks later, the Ninth Circuit assumed that the plea (entered before expiration of the supervised release term) did not make his incarceration connected to a conviction. See id at 1103–06. Perhaps the district court did not accept the guilty plea until the sentencing hearing or at least until a few days after it was entered. The court’s opinion is unclear.
  • 138Id at 1103.
  • 139Id at 1103–04.
  • 140Morales-Alejo, 193 F3d at 1105.
  • 141Id.
  • 142Id.
  • 143Id.
  • 144Morales-Alejo, 193 F3d at 1105. For example, 18 USC § 3559 uses the term “imprisonment” to describe allowable punishment for various classes of crimes, while 18 USC § 3142 uses the term “detention” in the context of pretrial custody, the aim of which is decidedly nonpenal.
  • 145Morales-Alejo, 193 F3d at 1105 (emphasis added), citing 18 USC § 3624(e).
  • 146Morales-Alejo, 193 F3d at 1105.
  • 147Id.
  • 148Id.
  • 149Morales-Alejo, 193 F3d at 1106.
  • 150See United States v Goins, 516 F3d 416, 419–20 (6th Cir 2008); United States v Molina-Gazca, 571 F3d 470, 473–74 (5th Cir 2009); United States v Johnson, 581 F3d 1310, 1311–12 (11th Cir 2009); United States v Ide, 624 F3d 666, 668–69 (4th Cir 2010).
  • 151516 F3d 416 (6th Cir 2008).
  • 152Id at 424.
  • 153This time, however, the court’s delay was seemingly the result of institutional bungling. The sentencing court was unaware that it possessed jurisdiction over Goins’s supervised release. Id at 417–18.
  • 154Id at 417–18.
  • 155Goins, 516 F3d at 417.
  • 156Id at 418.
  • 157Id.
  • 158See id.
  • 159See Goins, 516 F3d at 419.
  • 160Id at 420.
  • 161Id.
  • 162Id at 422.
  • 163Goins, 516 F3d at 421.
  • 164Id (quotation marks omitted).
  • 165Id at 422.
  • 166Id at 424.
  • 167Goins, 516 F3d at 422.
  • 168Id at 424.
  • 169Id.
  • 170571 F3d 470 (5th Cir 2009).
  • 171Id at 473–74, citing Goins, 516 F3d at 421–22.
  • 172See Molina-Gazca, 571 F3d at 473–74, citing Morales-Alejo, 193 F3d at 1105.
  • 173Molina-Gazca, 571 F3d at 474.
  • 17418 USC § 3041 (emphasis added).
  • 175Molina-Gazca, 571 F3d at 473.
  • 176Id.
  • 177See generally Johnson, 581 F3d 1310.
  • 178Ide, 624 F3d at 669 (emphasis added).
  • 179Id.
  • 180See Johnson, 581 F3d at 1311–12.
  • 181829 F3d 705 (DC Cir 2016).
  • 182Id at 707, quoting 18 USC § 3624(e).
  • 183Marsh, 829 F3d at 709–10.
  • 184See id.
  • 185Id.
  • 186Id at 709.
  • 187Marsh, 829 F3d at 709.
  • 188Id.
  • 189Id, quoting 1 USC § 1.
  • 190560 US 438 (2010).
  • 191Id at 448.
  • 192Marsh, 829 F3d at 710.
  • 193Id.
  • 194Id.
  • 195See id (expressing concern about the jurisdictional uncertainty and notice issues arising from the majority rule).
  • 196Marsh, 829 F3d at 710.
  • 197Id.
  • 198Id.
  • 199See Blum v Stenson, 465 US 886, 896 (1984) (“Where, as here, resolution of a question of federal law turns on a statute and the intention of Congress, we look first to the statutory language and then to the legislative history if the statutory language is unclear.”).
  • 200See Markman, et al, Recidivism of Offenders at *3 (cited in note 16).
  • 201Federal Offenders Sentenced to Supervised Release *49–52 (US Sentencing Commission, July 2010), archived at
  • 202While the majority position does not ask courts to actually decide cases before they are sure of their jurisdiction, it comes awfully close. Deciding cases on the basis of “hypothetical jurisdiction” is highly disfavored and arguably impermissible. See Steel Co v Citizens for a Better Environment, 523 US 83, 93–94 (1998) (declining to endorse “hypothetical jurisdiction” in the civil context).
  • 203Scott Dodson, The Complexity of Jurisdictional Clarity, 97 Va L Rev 1, 3 (2011).
  • 204Martin H. Redish, Reassessing the Allocation of Judicial Business between State and Federal Courts: Federal Jurisdiction and “The Martian Chronicles”, 78 Va L Rev 1769, 1794 (1992).
  • 205Goins, 516 F3d at 424.
  • 206See Robert E. Scott and William J. Stuntz, Plea Bargaining as Contract, 101 Yale L J 1909, 1947–48 (1992) (“[A]ll defendants—whether guilty or innocent—are offered a sentence based upon the prosecutor’s estimate of the strength of the case.”).
  • 207Note that entering a guilty plea represents an admission on X’s part that he violated the terms of his supervised release. All release terms carry the condition that the offender must not commit a crime during the term. See 18 USC § 3583(d).
  • 208While the governing statute gives courts discretion in deciding whether conduct warrants revocation (except in a few cases), see 18 USC § 3583(e), (g), the sentencing guidelines suggest courts should treat far more conduct as subject to mandatory revocation. See USSG § 7B1.3.
  • 20918 USC § 3583(e)(3) gives the sentencing court wide discretion in determining the length of imprisonment it imposes for supervised release violations. Nevertheless, the advisory US Sentencing Guidelines provide a detailed system of sentence ranges that depend on the grade of the supervised release violation and the criminal history of the offender. See USSG § 7B1.4.
  • 210The sentencing guidelines suggest that probation officers have some discretion in deciding what conduct they will include in their reports to the sentencing court, at least for crimes punishable by less than a year of prison. See USSG § 7B1.2 (explaining that a probation officer must report a violation to the court unless it is “Grade C” and the officer determines that the violation is “minor, and not part of a continuing pattern” and “will not present an undue risk to an individual or the public”).
  • 211See Andrew D. Leipold, How the Pretrial Process Contributes to Wrongful Convictions, 42 Am Crim L Rev 1123, 1130 (2005) (noting that “pretrial detention can hamper the defense by making it difficult for the suspect and his lawyer to find witnesses, gather and review evidence, and consult about strategy”).
  • 212See Robert Schehr, The Emperor’s New Clothes: Intellectual Dishonesty and the Unconstitutionality of Plea-Bargaining, 2 Tex A&M L Rev 385, 389 (2015) (explaining that, especially for innocent defendants, “the process costs of proceeding to trial often dwarf plea prices”), citing Josh Bowers, Punishing the Innocent, 156 U Pa L Rev 1117, 1132 (2008).
  • 213Stephanos Bibas, Plea Bargaining outside the Shadow of Trial, 117 Harv L Rev 2464, 2528–29 (2004) (discussing factors, including uncertainty, that may affect whether plea bargains actually reflect trial outcomes).
  • 214Id at 2527.
  • 215See USSG § 5D1.3(c) (quotation marks omitted).
  • 216See USSG § 5D1.3(c)(3).
  • 217See USSG § 5D1.3(c)(6).
  • 218See USSG § 5D1.3(c)(7).
  • 219See Carr, 560 US at 441–42.
  • 220Id.
  • 221Id at 462 (Alito dissenting).
  • 222See Gregory E. Maggs, Reducing the Costs of Statutory Ambiguity: Alternative Approaches and the Federal Courts Study Committee, 29 Harv J Legis 123, 125 (1992) (“The term ‘statutory ambiguity’ itself could have several meanings.”).
  • 223Brett M. Kavanaugh, Book Review, Fixing Statutory Interpretation, 129 Harv L Rev 2118, 2136 (2016).
  • 224See Maggs, 29 Harv J Legis at 125 (cited in note 222) (describing a spectrum that ranges from a strict definition of ambiguity encompassing “only those portions of statutes that no court could interpret,” to a loose definition that labels ambiguous “any statutory provision subject to more than one reading, even if no reasonable person would disagree about what it actually means”).
  • 225See Kavanaugh, Book Review, 129 Harv L Rev at 2137 (cited in note 223).
  • 226Maggs, 29 Harv J Legis at 125 (cited in note 222).
  • 227Kavanaugh, Book Review, 129 Harv L Rev at 2144 (cited in note 223) (advocating for a two-step interpretive process in which the first step would require judges to “determine the best reading of the text of the statute by interpreting the words of the statute, taking account of the context of the whole statute, and applying any other appropriate semantic canons of construction” followed by a second step in which a judge may “apply—openly and honestly—any substantive canons . . . that may justify departure from the text”).
  • 228Compare Snell Island SNF LLC v National Labor Relations Board, 568 F3d 410, 420 (2d Cir 2009) (noting a split “suggests that the statute is ambiguous”), with Allapattah Services, Inc v Exxon Corp, 333 F3d 1248, 1254 n 4 (11th Cir 2003) (stating “[t]he mere existence of a split among circuits . . . does not relieve us of our obligation to interpret the statute independently”). Nevertheless, the Supreme Court has noted that the existence of a split among authorities makes it “difficult indeed” to contend that the statutory language prompting the split is unambiguous. Smiley v Citibank (South Dakota), NA, 517 US 735, 739 (1996).
  • 229See Molina-Gazca, 571 F3d at 474.
  • 230See United Air Regulatory Group v Environmental Protection Agency, 134 S Ct 2427, 2441 (2014) (stating that inconsistent use of a term within a broader statutory scheme is “not conducive to clarity”).
  • 231See Corley v United States, 556 US 303, 314 n 5 (2009) (explaining that odd results cut against a finding of statutory clarity). See also Barber v Thomas, 560 US 474, 491 (2010) (suggesting interpretations that produce odd results are disfavored).
  • 232See text accompanying notes 169–77.
  • 23318 USC § 3624(e).
  • 234S Rep No 98-225 at 148–49 (cited in note 22). Note that this limitation was more explicit in the original language of the tolling provision. See text accompanying note 82.
  • 235Intermittent confinement is specifically condoned by USSG § 5D1.3(e)(6) for violations of supervised release.
  • 236While a judge might consider the likelihood that a defendant will reoffend when he designs the term, he has no way of predicting whether pretrial detention will result from the new offense. See 18 USC § 3553(c).
  • 237Johnson, 529 US at 59.
  • 238See Part III.D.
  • 239See Corley, 556 US at 314 n 5. See also Lamie v United States Trustee, 540 US 526, 536 (2004) (suggesting that when the plain meaning of a statutory text leads to “absurd results,” courts are “require[d] [ ] to treat the text as if it were ambiguous”).
  • 240See Part II.A.
  • 241Johnson, 529 US at 59.
  • 242See id at 60.
  • 243See S Rep No 98-225 at 49–50 (cited in note 22).
  • 244See id at 123.
  • 245See id.
  • 246Sentencing Reform Act § 212, 98 Stat at 2009.
  • 247H Rep No 99-797 at 21 (cited in note 85) (emphasis added).
  • 248Goins, 516 F3d at 420–21. The Sixth Circuit first made this observation in an unreported case. See United States v Sturdivant, 1999 WL 1204689, *2 (6th Cir). In that case, the court observed: “It appears that Congress used ‘in connection with’ instead of ‘for’ to cover the scenario of a revocation of parole.” Id at *2. In Goins the court clarified that this opinion, though nonprecedential, should not be interpreted to imply that the phrase “in connection with” covered only revocation of parole. See Goins, 516 F3d at 421.
  • 249This is not to say that Congress always perfectly carries out its stated purposes. Rather, a clear congressional preference for certainty in sentencing, combined with legislative history suggesting an interpretation that would offer courts and defendants greater certainty in sentencing, presents a convincing argument that the more certain interpretation is in line with congressional intent. Additionally, the Court often finds interpretations that are in line with the general purpose of a large enactment to be favored over those that are less so. See, for example, State Farm Fire and Casualty Co v United States, 137 S Ct 436, 444 (2016) (finding that a Senate Committee Report’s “recitation of the general purpose of the statute” favored the textual interpretation offered by one party over the other).
  • 250See 18 USC § 3583(i).
  • 251S Rep No 86-1696 at 3 (cited in note 107) (emphasis added).
  • 252S Rep No 98-225 at 3 (cited in note 22).
  • 253See Salerno, 481 US at 747. See also Bell v Wolfish, 441 US 520, 535 (1979).
  • 254Bell, 441 US at 535.
  • 255Id.
  • 256Palmer v Dugger, 833 F2d 253, 254 (11th Cir 1987).
  • 257Faye v Gray, 541 F2d 665, 667 (7th Cir 1976).
  • 258Palmer, 833 F2d at 254.
  • 259See id at 254–55; Faye, 541 F2d at 667.
  • 260See Part I.B.3.
  • 261See text accompanying notes 121–22.
  • 262See 18 USC § 3583(d) (providing an extensive list of both mandatory and discretionary conditions that may be placed on supervised release terms).
  • 263Bell, 441 US at 535.
  • 264See, for example, Hooper v California, 155 US 648, 657 (1895) (“The elementary rule is that every reasonable construction must be resorted to in order to save a statute from unconstitutionality.”).
  • 265Edward J. DeBartolo Corp v Florida Gulf Coast Building & Construction Trades Council, 485 US 568, 575 (1988).
  • 266See, for example, Leading Cases, 101 Harv L Rev 119, 175–76 (1987) (stating that the Court’s approach in Salerno “ignores the fact that detention can be both punitive and regulatory” and “silently reduced the presumption of innocence to nothing more than an allocation of the burden of proof at trial”).
  • 267See Reply Brief for Petitioner, United States v Salerno, Docket No 86-87, *2–5 (US filed Jan 12, 1987) (available on Westlaw at 1987 WL 880539).
  • 268New Hampshire v Maine, 532 US 742, 749 (2001).
  • 269S Rep No 86-1696 at 3 (cited in note 107). See also Part I.B.3.
  • 27074 Stat at 738.
  • 271See 18 USC § 3585(b).
  • 272See 18 USC § 3142(c)(2). This, however, does not mean that any financial condition of bail set at an amount the defendant cannot afford is prohibited. See Samuel R. Wiseman, Pretrial Detention and the Right to Be Monitored, 123 Yale L J 1344, 1396 n 234 (2014) (explaining that an unaffordable financial condition is permitted “because without the money, the risk of flight is too great”), citing United States v Jessup, 757 F2d 378, 389 (1st Cir 1985).
  • 273See Loupe, Comment, 77 La L Rev at 114–15, 138–39 (cited in note 40) (describing a Louisiana statute limiting recognizance bonds).
  • 274As noted in Part I.A, this controversy also implicates the relationship between federal supervised release terms and state pretrial detention. See note 63 and accompanying text.
  • 275See Probation and Parole in the United States, 2014 *1 (Bureau of Justice Statistics, Nov 2015), archived at (putting the supervised release population at an estimated 856,900).