The United States appeals from the district court's dismissal with prejudice, under the Speedy Trial Act ("STA"), 18 U.S.C. Secs. 3161-3174, of its superseding indictment charging defendant Larry Lee Taylor with conspiracy to possess cocaine and possession with intent to distribute. The dismissal was granted based upon the government's violation of the STA's 70-day indictment-to-trial provision, 18 U.S.C. Sec. 3161(c)(1).

The government contends the 70-day STA time "clock" should start over when a fugitive is apprehended after failing to appear for trial. The government also contends that the district court improperly computed the delays excludable under the STA. Finally, the government maintains that the district court abused its discretion in dismissing the indictment with prejudice. We affirm.

* BACKGROUND

Larry Lee Taylor was indicted on July 25, 1984, for conspiracy to possess cocaine with intent to distribute in violation of 21 U.S.C. Sec. 846, and for actual possession of cocaine with intent to distribute in violation of 21 U.S.C. Sec. 841(a)(1), (b)(1)(A), and 18 U.S.C. Sec. 2. He was scheduled for trial on these criminal charges before the United States District Court for the Western District of Washington on November 19, 1984. Taylor failed to appear for trial and a fugitive bench warrant was issued. Taylor was apprehended in California on February 5, 1985, by officers of the San Mateo County sheriff's department.

After his arrest, Taylor was held in San Mateo County jail on both the federal bench warrant and a state bench warrant issued following Taylor's failure to appear for trial on a petty theft charge. The federal government obtained a superseding indictment on April 24, 1985.1

Taylor was transferred from San Mateo County custody to federal custody on February 7, 1985, pursuant to a writ of habeas corpus ad testificandum issued by the United States District Court for the Northern District of California to obtain his testimony in another federal narcotics case, United States v. Seigert. He testified in the Seigert trial on February 21, and was held for possible recall in Seigert through February 22.

On February 28, the charges pending against Taylor in San Mateo County were dismissed, and the United States Marshal Service ("USMS") was notified on March 1 that local holds were released. On March 6, Taylor made an initial appearance on the federal fugitive warrant before a magistrate in the Northern District of California. On April 3, the magistrate signed an order directing that defendant be transported to the Western District of Washington.

On April 8, Taylor was transferred from San Francisco County Jail to Sutter County Jail while the USMS waited to assemble other prisoners for transport to Oregon and Washington rather than traveling with defendant alone. On April 17, Taylor was transported to Portland, Oregon, but the following day the United States District Court for the Northern District of California issued a second writ of habeas corpus ad testificandum ordering defendant returned to California for the retrial of Seigert. Taylor was returned to California on April 23, retrial began around May 7, and on May 17 he was transported to the Western District of Washington. On April 24, a grand jury in the Western District had returned a superseding indictment, adding a charge of failure to appear to the original narcotics charges.

After Taylor's return, the United States District Court for the Western District of Washington held that, since only one day had remained on the STA clock when trial was scheduled on November 19, 1984, and since the clock did not begin anew when defendant was arrested on February 5, the court had to examine the time which had elapsed between his disappearance on November 19, 1984, and the issuance of the superseding indictment on April 24, 1985, to determine which delays were excludable under 18 U.S.C. Sec. 3161(h). The court concluded that fifteen days of the delay were not excludable. Since the STA clock had expired fourteen days before Taylor was brought to trial, the district court dismissed the narcotics charges under the superseding indictment.

We review de novo the district court's interpretation of the provisions of the Speedy Trial Act. United States v. Gallardo, 773 F.2d 1496, 1501 (9th Cir.1985); United States v. Henderson, 746 F.2d 619, 622 (9th Cir.1984), aff'd, --- U.S. ----, 106 S. Ct. 1871, 90 L. Ed. 2d 299 (1986).

The STA, 18 U.S.C. Sec. 3161(c)(1), "operates like a statute of limitations." United States v. Mehrmanesh, 652 F.2d 766, 769 (9th Cir.1980). Pursuant to the statute, a defendant must be brought to trial within 70 days from the later of (1) the filing date of the information or indictment, or (2) the date of his initial appearance before a judicial officer in the charging district. The STA provides numerous exclusions from this 70-day period. See 18 U.S.C. Sec. 3161(h). But if the defendant is not brought to trial within the 70-day period plus the period allowed under the exclusions, the court must dismiss the indictment on motion of the defendant. 18 U.S.C. Sec. 3162(a)(2).

The district court concluded that the time which elapsed between Taylor's failure to appear for trial on November 19, 1984, and his apprehension on February 5, 1985, was excludable under 18 U.S.C. Sec. 3161(h)(3)(A), (B). This section provides:

(h) The following periods of delay shall be excluded ... in computing the time within which the trial of any such offense must commence:

....

3(A) Any period of delay resulting from the absence or unavailability of the defendant or an essential witness.

(B) For purposes of subparagraph (A) of this paragraph, a defendant or an essential witness shall be considered absent when his whereabouts are unknown and, in addition, he is attempting to avoid apprehension or prosecution or his whereabouts cannot be determined by due diligence. For purposes of such subparagraph, a defendant or an essential witness shall be considered unavailable whenever his whereabouts are known but his presence for trial cannot be obtained by due diligence or he resists appearing at or being returned for trial.

As applied under the facts of this case, section 3161(h)(3) clearly tolled the STA time "clock" for that period during which Taylor was a fugitive. However, upon his apprehension, the clock simply resumed from the point at which it was stopped by defendant's absence. In this case, there was one day remaining before Taylor was to be tried. According to the district court, fifteen days after Taylor's arrest were identified as unexcludable under STA. As a consequence, the government failed to comply with the statutory time period by fourteen days.

The United States contends that section 3161(h)(3) does not control under the circumstances of this case. The government instead argues that the clock should have been restarted on Taylor's apprehension, thereby giving the government an additional 70 days within which to bring Taylor to trial upon the superseding indictment.

Although the government failed to cite any authority for this proposition, we have discovered the following dicta in a footnote of an Eleventh Circuit decision:

Appellant assumes, as do we, that where a defendant fails to appear for trial and is recaptured a year later and placed in federal custody, the Government has 70 days in which to try him for the offense for which he was originally charged. Requiring that the period during which a defendant is a fugitive be excluded from the original 70-day calculation would be unfair to the government, for if defendant became a fugitive 69 days after his initial appearance before a judicial officer, upon his recapture one year later, the Government would have only one day to try him for the original offense. We reject, however, the Government's argument that upon becoming a fugitive, a defendant waives his right to speedy trial upon recapture. Requiring that the 70-day period begin anew upon a defendant's recapture is the most reasonable result.

United States v. Studnicka, 777 F.2d 652, 657 n. 16 (11th Cir.1985).

However, all other courts which have considered the problem presented when a defendant had been at large for some period after the STA time clock had started, with the filing of an indictment or an initial appearance, have held that the delay occasioned by the defendant's absence should simply be excluded under section 3161(h)(3).2 None of these decisions suggests that the clock should be restarted upon the defendant's apprehension.

These holdings comport with the plain language of the STA. Since section 3161(h)(3) expressly provides a scheme for considering the effect of a defendant's absence or unavailability, we should not try to improve upon the statutory scheme by implying a provision restarting the clock upon apprehension of an absent defendant. Indeed, holding that the clock is to be restarted whenever a defendant is absent or unavailable would render section 3161(h)(3) meaningless. There could be no purpose in the exclusion of delays instigated by a defendant's absence or unavailability if the clock would instead be restarted upon his reappearance. We will not accept an interpretation of a statute which renders any part of the statutory scheme superfluous. People of California v. Tahoe Regional Planning Agency, 766 F.2d 1308, 1314 (9th Cir.), amended, 775 F.2d 998 (9th Cir.1985).

Moreover, the legislative history of the STA indicates that Congress was aware of the potential problems in quickly bringing to trial a defendant who became a fugitive when the time clock was just about to expire. For example, in 1971, then Assistant Attorney General William H. Rehnquist suggested that the STA legislation include a special provision allowing additional time after a fugitive defendant has been apprehended:3

Further, if a defendant is available for 58 days prior to trial, but then becomes a fugitive for two years, under section 3161(c)(3), upon his rearrest, the Government would only have two days to bring him to trial. This is obviously impossible since the evidence could not be reassembled on such short notice. We therefore believe that the terms "absence" and "unavailability" should be defined and that specific provisions should be made for situations where the defendant becomes a fugitive.

Deputy Attorney General Joseph T. Sneed also warned the Congress:4

A defendant could "skip" bail on the 59th day of the time period and once apprehended, the Government would have 1 day within which to reassemble the evidence and to try him. This results in the anomalous situation of an escapee being given priority as to a trial date over those defendants who have abided by the conditions of their bail.

To address this perceived flaw in the legislation, the Justice Department proposed an amendment providing that a defendant who failed to appear for trial would be deemed arraigned only upon the date of his subsequent appearance before the court after his apprehension.5 In effect, the amendment would have restarted the STA time clock when the fugitive defendant was brought before the court where the charge was pending.6 The proposed amendment was not adopted.

Despite the fact that these concerns were raised, and alternative approaches suggested, Congress decided to address the problem of fugitive defendants through section 3161(h)(3) and excluded only that the period during which the defendant was missing.

As the district court ruled in this case, "The STA specifically provides that delay resulting from the unavailability of the defendant constitutes excludable time, Sec. 3161(h)(3)(A), not that defendant's return to a district after flight and recapture starts the running of an entirely new 70-day period. If Congress had intended the latter result, it would have said so."

The United States argues that the clock should nevertheless begin anew when a fugitive defendant is apprehended because Congress intended that "if a climactic and unpredictable event occurs that disrupts the customary flow of a prosecution, the defendant is to be tried within seventy days of that event." The United States attempts to analogize a defendant's flight from trial to sua sponte dismissal by the trial court, grant of a mistrial or new trial, and withdrawal of a plea of guilty or nolo contendere. The STA period is restarted with the occurrence of these latter events; the government contends it should also be restarted when a defendant fails to appear.

We reject this analogy because the STA specifically provides that the clock is restarted for withdrawal of a plea, 18 U.S.C. Sec. 3161(i),7 and declaration of mistrial or new trial, 18 U.S.C. Sec. 3161(e).8 Moreover, we recently held that a sua sponte dismissal by the trial court constitutes "a charge ... dismissed or otherwise dropped" under section 3161(d), which provides that the clock be restarted under such circumstances. United States v. Feldman, 788 F.2d 544, 549 (9th Cir.1986). By contrast, the statute does not allow restarting the clock when the defendant is unavailable, 18 U.S.C. Sec. 3161(h)(3)(A), it only excludes such periods from the 70-day calculation.

Our holding today does not mean that a defendant will benefit by absconding from the trial court's jurisdiction just before the STA clock runs or trial is scheduled to begin. There remain substantial criminal sanctions which attach to a defendant's failure to appear for trial. See 18 U.S.C. Sec. 3146. Furthermore, if there is insufficient time remaining for effective trial preparation upon the fugitive defendant's apprehension, the district court retains the discretion to grant a continuance to serve the "ends of justice" under 18 U.S.C. Sec. 3161(h)(8)(A). See United States v. Kamer, 781 F.2d 1380, 1389-90 (9th Cir.), cert. denied, --- U.S. ----, 107 S. Ct. 80, 93 L. Ed. 2d 35 (1986); United States v. Gallardo, 73 F.2d at 1505-06. In this case, the district court found that the STA clock had expired due to government delays before the defendant had even been returned to the Western District of Washington. Consequently, no continuance could have been requested or granted.

Finally, although we hold that the plain language of the statute precludes an order to restart the clock upon the fugitive defendant's apprehension, the trial court may certainly take into account the defendant's culpable absence in deciding whether any resulting dismissal for violation of the STA should be with or without prejudice.

We now turn to the calculation of the number of nonexcludable days between the date of Taylor's arrest in California and the date upon which the United States District Court for the Western District of Washington granted the dismissal. We review de novo the district court's method of computing excludable days under the STA, and we review the factual findings underlying the STA determination under the "clear error" standard. United States v. Gallardo, 773 F.2d at 1501; United States v. Henderson, 746 F.2d at 622.

As explained above, the defendant concedes that the period of time between his flight on November 19, 1984, and his arrest on February 5, 1985, was excludable under section 3161(h)(3)(A). The parties also concede that the short delay while Taylor was in state custody on February 6 and 7 was excludable; the delay attributable to removal proceedings between March 6 and April 3 was excludable under 18 U.S.C. Sec. 3161(h)(1)(G); and the period during which Taylor was held for testimony in the Northern District of California, between February 7 and 22, and again between April 18 and 24, was excludable. Although the district court did not consider it, apparently it was assumed that the period after April 24 was excludable because of the superseding indictment.

Two time periods are disputed, the eleven days between February 23 and March 5, and the fourteen days between April 4 and 17. Both are periods during which Taylor was in the custody of the United States Marshal Service. The district court concluded that fifteen of these days counted against the 70-day time clock.

Between February 23 and March 5, Taylor was still in custody in San Francisco following his first testimony at the United States v. Seigert trial in the Northern District of California. The government contends that the six days between February 23 and February 28 should be excluded as processing time during which Taylor was effectively in state custody on a charge of petty theft. The district court properly rejected this argument. Although the San Francisco sheriff had been ordered to return Taylor to state custody in San Mateo, the USMS retained custody of Taylor in San Francisco. Consequently, we cannot regard these six days as tantamount to Taylor's having been in state custody on the pending petty theft charge. Moreover, this argument hardly accounts for the remaining five day delay between March 1, the day after the San Mateo County charges against Taylor were dropped, and March 5, the day before he was finally brought before a federal magistrate in the Northern District of California on the federal bench warrant.9

The second disputed period of time began April 4, after the magistrate in California ordered Taylor's removal to the Western District of Washington, and ended April 17, the day before he was ordered returned from Portland, Oregon to testify again in the Northern District of California. The government contends this entire period should be excluded, arguing that when a defendant absconds and is apprehended in a distant jurisdiction, the USMS may consider matters of economy in arranging transportation back to the charging jurisdiction. In this instance, the USMS delayed Taylor's transportation to the Western District of Washington while it collected a larger number of prisoners to transport together.

The government's position is not supported by the statute. On the matter of transportation of defendants between jurisdictions, section 3161(h)(1)(H) of the STA is straightforward:

(h) The following periods of delay shall be excluded in computing the time within which ... the trial of any such offense must commence:

(1) Any period of delay resulting from other proceedings concerning the defendant, including but not limited to--

....

(H) delay resulting from transportation of any defendant from another district, ... except that any time consumed in excess of ten days from the date an order of removal or an order directing such transportation, and the defendant's arrival at the destination shall be presumed to be unreasonable....

See also United States v. Greene, 783 F.2d 1364, 1368 (9th Cir.), cert. denied, --- U.S. ----, 106 S. Ct. 2923, 91 L. Ed. 2d 551 (1986).

Under 18 U.S.C. Sec. 3161(h)(1)(H), the ten days from April 4 through April 13 were excludable. The remaining four days must be presumed unreasonable. The government failed to rebut the presumption that these additional days taken to transport Taylor to the Western District of Washington were not reasonable. Ten days is surely ample time within which to remove an individual from northern California to western Washington. The legislative history indicates that delays to accommodate the USMS, in its desire to effect economical transportation of prisoners in larger groups, are not excludable under the Act.10

Finally, the government contends that the entire period following Taylor's apprehension should be excluded if the government exercised "due diligence" in procuring his return to the charging jurisdiction. This "due diligence" requirement applies only in the context of an "unavailable defendant" under 18 U.S.C. Sec. 3161(h)(3)(A), (B). Where the defendant's whereabouts are known, the government must exercise "due diligence" to secure his appearance at trial. If his appearance cannot be secured even through "due diligence," the defendant will be regarded as "unavailable" and any resulting delay is excludable from the STA time clock. 18 U.S.C. Sec. 3161(h)(3)(A), (B). Consequently, the delays occurring while Taylor was in state custody, while he was being held to testify in the Northern District of California, and while he was resisting removal from California to the Western District of Washington were excludable because Taylor was effectively "unavailable" during that period. The exclusion of these periods, however, is not at issue in this case.

During the two disputed periods, totaling fifteen days, Taylor was not "unavailable;" he was in the custody of the USMS, and within the full control of the federal government. The government simply failed to bring him promptly before a federal magistrate to initiate removal proceedings and to transport him to the charging jurisdiction within ten days after the removal order and that failure resulted in the delays which count against the 70-day time limit.

We affirm the district court's conclusion that fifteen days of non-excludable time elapsed after Taylor's apprehension. The 70-day clock was thus exceeded by fourteen days.

The Speedy Trial Act provides that dismissal of the indictment upon motion of the defendant is mandatory when the 70-day statutory period is exceeded. 18 U.S.C. Sec. 3162(a)(2). The dismissal may be with or without prejudice. In making this determination, the trial court is to consider, among others, the following factors: "the seriousness of the offense; the facts and circumstances of the case which led to the dismissal; and the impact of a reprosecution on the administration of [the STA] and on the administration of justice." Id.

We review the district court's decision to dismiss Taylor's indictment with prejudice for an abuse of discretion. See United States v. Frey, 735 F.2d 350, 353 (9th Cir.1984).

The district court recognized that the drug violations with which Taylor was charged were serious. The fourteen day delay beyond the STA 70-day period, although not wholly insubstantial, was not so great as to mandate dismissal with prejudice. Although there is no indication that the delay would have prejudiced Taylor in preparing for trial, he did suffer prejudice in that he was incarcerated during the entire period.

In ordering dismissal with prejudice, the district court focused primarily upon the government's failure to bring Taylor promptly before a magistrate during the eleven days he was in federal custody following the Seigert trial, and the unreasonable delay by the USMS in transporting him to the Western District of Washington after the removal order was issued. The court decried the "government's lackadaisical behavior in this case," and the fact that the government "placed more value on accommodating the convenience of the USMS than in complying with the plain language of the STA."11

The court also held that "the administration of the STA and of justice would be seriously impaired if the court were not to respond sternly" to the violation. Under the circumstances, the district court found dismissal with prejudice was mandated because "if the government's behavior in this case were to be tacitly condoned by dismissing the indictment without prejudice, then the STA would become a hollow guarantee."

The purpose of the district court's order was to send a strong message to the government that the STA must be observed, despite the government's apparent antipathy toward a recaptured fugitive. Under the peculiar circumstances of this case, we see no need to disturb that ruling on appeal. The district court acted within the bounds of its discretion.

AFFIRMED.

POOLE, Circuit Judge, concurring in part and dissenting in part.

Concededly, the defendant was not brought to trial until 14 days beyond the time prescribed by the Speedy Trial Act, considering the period in which delay was excludable under 18 U.S.C. Sec. 3161(h). Where a defendant is not brought to trial within 70 days from the date the indictment is returned or (in case of information) from the date of an initial appearance before a magistrate in the district of the prosecution, dismissal is mandatory. But the court has discretion, and is obligated, to determine whether the dismissal should be with or without prejudice. The district court dismissed the indictment (and with it the entire case against the defendant) with prejudice. I believe this was entirely uncalled for and constituted an abuse of allowable discretion.

Taylor was scheduled for trial on very serious charges of conspiracy to possess cocaine (21 U.S.C. Sec. 846) and for possession of cocaine with intent to distribute (21 U.S.C. Sec. 841(a)(1)), and for aiding and abetting (18 U.S.C. Sec. 2). He was to have been tried in Seattle, Western District of Washington, on November 19, 1984. He did not show up for trial and a fugitive warrant was issued. He successfully avoided capture for 78 days, until February 5, 1985, when he was taken into custody on state charges by the Sheriff of San Mateo County, California. Two days later the Sheriff delivered Taylor to the United States Marshal pursuant to an outstanding writ of habeas corpus ad testificandum which had been issued out of the United States District Court, Northern District of California, where his testimony was required in another narcotics prosecution, United States v. Seigert. Under that authority he was detained in federal custody through February 22, 1985. Thereafter, on February 28, the state charges upon which the Sheriff had arrested Taylor were dismissed, and on March 1 the Federal Marshal was informed that the "holds" on Taylor were dismissed. As of that date Taylor was a fugitive in custody in the Northern District of California charged in a complaint under the Federal Fugitive Act with having fled from the Western District of Washington, Seattle. On April 3, a United States Magistrate entered an order directing the Marshal to transport Taylor back to Seattle. That Friday (April 5) began the Easter weekend and on Monday, April 8, Taylor was transferred from jail in San Francisco to Sutter County jail in Northern California. He remained there until April 17 while the Marshal assembled other prisoners for transport to the districts of Oregon and Washington. On April 17 Taylor arrived in Portland, but on the next day, April 18, the district court in San Francisco issued a second writ of habeas corpus ad testificandum requiring Taylor to be brought back to San Francisco for the second trial of defendant Seigert. He was returned to San Francisco on April 23 and was held there during the trial which began on May 7. On May 17 Taylor was returned to the Western District of Washington.

The district court concluded that the period of detention between February 23 and March 5, 1985 constituted 11 days of federal custody following Taylor's appearance in the Seigert trial, and rejected the government's contention that he had only been temporarily in federal custody for the 6 day period between February 23 and February 28 because he was to have been returned to state custody.

It is true that Taylor was not "effectively in state custody" during this time. A state court judge cannot "order" the United States Marshal to deliver a prisoner to the state court. However, all the San Mateo County authorities had to do in order to exercise state custody was to come to the San Francisco county jail with the required papers and take custody of the prisoner after notifying the United States Marshal, who would then "sign off." The record does not indicate why San Mateo did not physically transport Taylor across the county line from San Francisco to San Mateo. Under the circumstances, however, informed as it was of the state holds and that its custody of Taylor could be interrupted at any moment, the government cannot be found "lackadaisical" see Maj. Op. at 1383 n. 9, or irresponsible in not immediately having taking advantage of these few days of Taylor's presence to begin proceedings against him.

Furthermore, when on March 1 the Marshal was notified that the state petty theft charges against Taylor had been dismissed and the "local holds" should be released, there was another outstanding federal charge (violation of the Federal Fugitive Act) pending against Taylor. On March 6 he came before a federal magistrate for arraignment on that charge and for removal proceedings under Rule 40 of the Federal Rules of Criminal Procedure. The trial judge and the majority find the period between March 1 and March 5 an unjustifiable delay necessitating dismissal with prejudice. See Maj. Op. at 1383. This approach is likewise unduly harsh. March 1 was a Friday. We do not know at what time of day the Service was notified of the release of local holds on Taylor, but if it was in late afternoon, the ability to process and move Taylor before the following Monday, March 4, would be doubtful. For some reason not apparent in the record, Taylor could not be brought before the charging magistrate until Wednesday the 6th. If Taylor appeared on the morning of that day, the majority's rationale for dismissal with prejudice in this case would have been founded upon a mere 24-hour delay. The district court does not exercise the discretion of the statute in coming to such a Draconian result.

The other critical period of time was that which elapsed between the date of April 3 when the Magistrate signed the papers transferring Taylor back to Washington, and April 18 when he was ordered to return from Portland back to San Francisco. The majority has found that most of this delay occurred while the Marshal waited to collect a larger number of prisoners for simultaneous transport in order to effect economy of expenses.

It seems clear to me that none of the delay shown in this case--although admittedly non-excludable under the statute--was of such studied, deliberate, and callous nature as to justify dismissal with prejudice. The Marshal was not indifferent to the duty to move prisoners so as to invite the harsh sanction of dismissal with prejudice, barring forever a public trial of the defendant on very serious violations of the narcotics laws. Congress intended to give to district judges some substantial leeway in deciding how to treat delays in bringing a defendant to trial. This case presents the incongruous situation where a defendant can be the instrument of his own deliverance. Taylor fled the day before his scheduled trial. He created his own 78-day "excludable time" by his own will, traveling from Seattle to California where he became the subject of criminal charges in two jurisdictions 800 miles away from the place of trial. It is ironic that the statutory scheme which would have assured his orderly trial in November 1984, is resorted to, five months later, as the reason for "springing" him to freedom and conferring upon him complete absolution from further prosecution. The delay we review in this case--a few days--rises to no level of constitutional wrong. The majority says that in opening the door to freedom for this defendant, the trial judge sought "to send a strong message." The context of that message reflects badly upon our notions of sound, evenheaded administration of justice. Consequently, I write this dissent.