When the Clock Runs

When the Clock Runs

The first Barker factor asks how long the delay lasted, and that question has no answer until a court knows when the counting begins, which periods are counted, and when the counting stops. All three answers come from one premise in the constitutional text. The Sixth Amendment provides that "the accused shall enjoy the right to a speedy trial," so the protection belongs to a person whom the government has formally designated as an accused.¹ That status matters because it produces the harms the guarantee addresses: pretrial confinement, the anxiety of a public charge left unresolved, and the weakening of the defense as memories fade and witnesses disappear.² A person who has not yet been arrested or charged has not taken on the burdens of a pending public accusation, so time before accusation falls outside the Sixth Amendment and is policed by statutes of limitation and the Due Process Clause of the Fifth Amendment instead.³

The clock therefore starts with the first official act that designates a person as an accused, ordinarily the arrest or the filing of charges, whichever comes first, and delay is counted from that point to trial.⁴ The rest of this section tests that premise against four situations in which the answer is less obvious: the time before accusation, charges that are dismissed or left suspended, a defendant held in another jurisdiction, and the period after guilt has been established. Each of those rules is measured against the baseline starting point, so the baseline comes first.

Attachment at Arrest or Charge

A court cannot count delay until it knows when the speedy-trial right began to protect the defendant. The right attaches, meaning it begins to apply, with the "first official act" designating the person as an accused, which ordinarily is the arrest or the filing of charges, whichever comes first.⁵ Delay is then measured from that date to trial.⁶ The rule takes the earlier of the two events because either one alone brings on the harms the guarantee addresses. An arrest confines the person or restricts liberty on release and marks the person publicly as a suspect. A formal charge, even without any arrest, hangs a public accusation over the person's life while the defense loses witnesses and memories as time passes. Whichever event comes first starts those burdens, so whichever comes first starts the clock.

The date of the crime plays no part in this calculation. A defendant who commits a burglary on January 1 but is not arrested or charged until June 1 has a speedy-trial clock that begins on June 1, and the five months between the offense and the accusation do not count toward the length of delay.⁷ If trial opens the next June 1, the measured delay is twelve months rather than seventeen. Those earlier months are not left unprotected, because other doctrines govern delay before accusation, a subject the next section takes up.

A sealed indictment refines the baseline. Under Rule 6(e)(4), the magistrate judge to whom an indictment is returned may direct that it be kept secret until the defendant is in custody or has been released pending trial, and the clerk must then seal it, with no one permitted to disclose its existence except as needed to issue or execute a warrant or summons.⁸ While the seal holds, the person is neither confined nor publicly accused, so none of the harms that give the speedy-trial right its content has yet begun. The clock therefore starts on the date of unsealing, since no incarceration or public accusation exists before then.⁹

The start date also does not depend on what the defendant knew. In Doggett v. United States the defendant was indicted but did not learn of the charge for years, because the government made little effort to find him.¹⁰ The right had attached with the accusation, so the long delay counted, and the roughly six years traceable to government neglect were extreme enough to presume prejudice to the defense absent a government showing to the contrary.¹¹ The sealed indictment and Doggett can look alike, since the defendant in each was unaware of the charge. What separates them is the government's role. A sealing order is a deliberate, court-authorized decision to keep the accusation secret, so no public accusation or confinement has begun, whereas in Doggett the government simply let a pending charge sit. Whether the defendant's ignorance of the charge matters at all comes back under the assertion and prejudice factors, and for now it is enough that it does not move the start date.

Pre-Accusation Delay

Every speedy-trial claim is measured from the attachment date, which leaves open what the Constitution does about the months before it. The Sixth Amendment gives no protection for the stretch between the crime and the arrest or charge. In United States v. Marion the Court read the Amendment's opening phrase, "In all criminal prosecutions," together with the history and purposes of the Speedy Trial Clause, and concluded that the guarantee does not apply to the pre-accusation period.¹³ The reasoning starts from the text. A prosecution has not begun until the government arrests or charges someone, so a person who is still under investigation is not yet bearing the confinement, the public accusation, and the eroding defense that the clause was written to relieve.¹⁴

A second reason supports the boundary, and it concerns prosecutors rather than defendants. Investigations take time, because witnesses must be interviewed, records gathered, and suspects identified, and prosecutors need broad discretion over when the evidence is strong enough to justify a charge. If every month of investigation counted against the government under the Sixth Amendment, a prosecutor would feel pressure to file early just to stop the clock. The result would be charges built on thin evidence against people whom further investigation might have cleared. Starting the clock at the accusation lets the government finish its work before the guarantee begins to run.

The absence of a Sixth Amendment right does not leave the pre-accusation period unprotected. In Marion and in United States v. Lovasco, the Court held that claims of undue pre-charge delay are governed by the Due Process Clause of the Fifth Amendment.¹⁵ That protection is narrow. The defendant must offer clear proof of specific and significant prejudice to the defense, almost always an impaired ability to present the case, as when a witness has died or memories have faded, and the court then examines the reasons for the delay while giving the prosecutor's charging discretion wide deference.¹⁶ Some statements of the test put the second showing as bad faith, meaning the prosecutor delayed for an improper reason. The practical consequence is the same either way: delay that results from good-faith investigation does not violate due process.¹⁷ Each demand answers a different concern. The prejudice requirement ties the claim to actual damage to the defense, so the passage of time alone never suffices. The reasons inquiry protects the legitimate investigative wait that the Sixth Amendment boundary was drawn to preserve. Because both showings are required, successful claims are extremely rare.¹⁸

The only other check on pre-accusation delay is the statute of limitations for the offense. A limitations period is a legislative deadline that runs from the commission of the crime, and a prosecution begun after it expires is barred whatever the reason for the wait or the condition of the evidence. Limitations periods differ by offense and by jurisdiction, so they supply an outer limit on how long the government may take, while due process polices the unusual case that falls inside that limit. Together these two checks cover the time before accusation, and the Sixth Amendment covers the time after it.

The same two years therefore produce opposite analyses, and the only variable is whether the person had been arrested or charged. For delay after attachment, the court weighs the length of the delay, the reason for it, the defendant's assertion of the right, and prejudice, with no single factor required.¹⁹ For delay before attachment, the defendant must meet the demanding due process showing. The usual error is to run Barker on the months before arrest or charge, because those months never enter the length-of-delay count that opens the Barker inquiry.²⁰

Dismissals and Suspended Charges

The baseline rule counts delay from attachment to trial, yet a prosecution does not always move in a straight line. A charge may be dismissed and later refiled, or left formally alive while nobody tries it, and each situation raises one question: during the pause, was the person still an accused? The first answer comes from United States v. MacDonald, which excluded the time between a dismissal and the renewal of charges, because the individual was not an accused during that gap.²² The rule applies where the charges are dismissed and the defendant is released.

The reason traces back to the premise that the Sixth Amendment protects an accused. Once the charges are gone and the defendant is free, the pressures the guarantee exists to relieve have lifted: there is no confinement, no public charge waiting for resolution, and no identified case to prepare against. The gap therefore resembles the pre-accusation period, which sits outside the Sixth Amendment and is policed by limitations statutes and the Fifth Amendment Due Process Clause.²³ The rule also removes an incentive problem. If the gap counted against the government, a prosecutor unsure of the evidence would have a reason to leave the charge pending rather than dismiss it, and the defendant would then live under an open accusation instead of a closed one. Excluding the gap keeps dismissal available as a genuine end to the accusation.

A second excluded period follows from the same logic. In United States v. Loud Hawk, the time during which the prosecution appealed a dismissal, while the defendants were neither incarcerated nor subject to bail, did not count, and the Court treated an interlocutory appeal (a review sought before the case is finally decided) as ordinarily a valid reason for delay.²⁴ Both exclusions rest on something concrete that ended or explained the pause: a dismissal that released the defendant in the first, a legitimate appeal in the second. That limit matters because a state can also pause a case without ending it.

The counterpoint comes from Klopfer v. North Carolina, which held unconstitutional a state practice that, in effect, suspended a prosecution with automatic leave to reinstate it.²⁵ The device was the nolle prosequi with leave, in which the prosecutor announces that the state will not proceed on the charge for now (the Latin means "will not prosecute") while keeping permission to revive it whenever the state chooses. The charge stays on the books, so the defendant cannot force a trial or clear the record, and the anxiety and the slow erosion of the defense continue with no trial in sight. If calling the case dismissed could stop the clock while all of those pressures remained, a state could hold an accusation over a person for years and claim that no time had elapsed. The dismissal in MacDonald ends the accusation, whereas leave to reinstate preserves it.

The two versions differ in one respect only, which is whether the accusation survived the pause. That is the sorting question for any case that stops and starts. If the charges were dismissed and the defendant released, the gap is excluded, and if the state shelved the charge while keeping power to revive it at will, the accusation survived and the gap cannot be written off. Students lose this issue when they read a nolle prosequi with leave to reinstate as an ordinary dismissal.²⁶

Detainers and Other Custody

The attachment and exclusion rules so far assume a defendant who is free or held only on the charge at issue. Many accused persons are already serving time on something else when a second jurisdiction charges them, and custody elsewhere does not suspend the Sixth Amendment. The right attaches when a detainer, a notice filed with the institution holding a prisoner to announce that another jurisdiction has a charge pending, is lodged against an accused who is in custody on other charges, and the delay is then counted like any other.²⁸ What remains is to say what the charging jurisdiction must do while its defendant sits in someone else's prison.

In Smith v. Hooey the Court held that prisoners enjoy the Sixth Amendment speedy-trial right too, and that once a prisoner demands trial the charging jurisdiction has a constitutional duty to make a "diligent, good-faith effort" to bring the prisoner before the court.²⁹ The duty exists because each interest the right protects is, if anything, more exposed for a prisoner. A pending charge can cost the prisoner a sentence that would have run partly concurrent with the one being served, and a detainer can lengthen the present imprisonment and worsen its conditions. The anxiety of an open accusation also undercuts rehabilitation, and a prisoner cannot go out and find witnesses or gather evidence while the passage of time wears the defense away.³⁰

Dickey v. Florida supplies the application. When a defendant demands a speedy trial, the state must respect that demand even though the defendant is held in another jurisdiction. The state may lack authority to compel that jurisdiction to hand the defendant over, but failing for a long period even to request cooperation may violate the right.³¹ The decision came before Barker, yet its reasoning maps onto the four Barker factors: a seven-year delay, no tenable reason for never seeking custody of the defendant from a federal prison, diligent and repeated efforts by the defendant to obtain a prompt trial, and prejudice in the form of two dead witnesses, a third who was unavailable, and lost police records.³²

The standard asks for effort rather than success, and each half of that has its own logic. The charging jurisdiction cannot command another sovereign's prison, so a rule demanding the prisoner's actual appearance would charge the state with a refusal it cannot control. A rule demanding nothing would be worse, because a request costs little and the state could otherwise leave the charge untouched for years while witnesses die and records disappear. The standard sits between those two. The lack of power to compel is not a valid reason for failing to ask, and the government bears the burden of showing a reason for the delay other than a failure to make a diligent, good-faith effort.³³ If it cannot make that showing, the failure counts against it, though less heavily than an actual intent to hamper the defense would.³⁴ The duty to ask also has a boundary: no effort is needed while the other jurisdiction holds the defendant pending or during trial and sentencing, because a return would not occur then, and courts are somewhat less demanding when the defendant is held in a foreign country, where cooperation is less likely.³⁵

The failure this rule targets is easy to picture: a prisoner held elsewhere demands trial on a pending charge, and for years the prosecutor never asks the holding jurisdiction to produce the prisoner. That inaction weighs heavily against the state and may establish a violation.³⁶ The matched pair below illustrates the reasonable-efforts standard rather than reporting a decided case.

The only variable in that pair is whether the state asked, which is why the other jurisdiction's refusal does not decide the claim. Federal prosecutions build the same idea into a procedure. Under the Speedy Trial Act of 1974, a prosecutor who knows that a person charged with a federal offense is serving a prison term must promptly either seek the prisoner's presence for trial or file a detainer with the custodian and ask that the prisoner be told of the right to demand trial, and a demand then obliges the prosecutor to promptly seek the prisoner's presence.³⁷ The demand matters in the constitutional analysis as well, and courts account for the prisoner's situation when judging it. A prisoner may not know of the charge when no detainer has been filed, and a prisoner is unlikely to have a lawyer, so a missing or informal demand is not held as harshly against the prisoner as it would be against a free defendant.³⁸

Sentencing Delay

The speedy-trial window opened by an arrest or charge closes when guilt is determined. The constitutional right protects the accused from arrest through trial, and it does not apply to sentencing, meaning the stage that begins once a defendant has been found guilty at trial or has pleaded guilty.⁴⁰ In Betterman v. Montana the Court set that boundary, so a defendant who has been convicted or has pleaded guilty cannot rest a Sixth Amendment speedy-trial claim on delay at the sentencing phase.⁴¹ The practical consequence is that the delay measured under the Barker factors runs from attachment to the verdict or plea, and the months that follow never enter the count. Earlier sections showed what starts the clock and which pauses stop it, and this one marks where the clock ends.

The reason lies in what the guarantee is built to accomplish. Because it secures a speedy trial, its purpose is to move the determination of guilt along, and the three interests the Court identified in Barker all describe the position of someone waiting for that determination: oppressive pretrial incarceration, the anxiety of living under pending charges, and impairment of the defense through lost witnesses or faded memories.⁴² A guilty verdict or a guilty plea ends that wait. The defendant is no longer an accused waiting to learn whether guilt will be established, and no contested question of guilt remains for a delay to distort, so the rationale behind the guarantee no longer fits. This reasoning explains only why the Sixth Amendment speedy-trial right stops where it does. It says nothing about whether some other rule governs a long wait for sentencing, and this chapter does not take that question up.

Delay before trial is the contrasting case. Barker v. Wingo is the seminal decision on Sixth Amendment speedy-trial claims, and its four-factor balancing test, which weighs the length of the delay, the reason for it, the defendant's assertion of the right, and prejudice, governs the wait between arrest and trial.⁴³ The same length of wait can therefore fall on either side of the line, depending on where in the prosecution it occurs.

The only variable in that pair is whether guilt had been found or admitted when the wait occurred. The usual error is to treat a long wait for sentencing as speedy-trial delay, when the right stops at the guilty verdict or plea and the Barker inquiry never begins for that period.⁴⁴