United States v. Jose Avalos and Rudolfo Castrillon

541 F.2d 1100, 1976 U.S. App. LEXIS 6395

• Court: Court of Appeals for the Fifth Circuit
• Decided: November 4, 1976
• Docket: 20-90011
• Status: Published

Opinion

GOLDBERG, Circuit Judge:

In this case we delve into the calculus by which the length and breadth of a criminal defendant’s right to speedy trial is measured. Our exploration traces the path of the appellants and their prosecutors through two jurisdictions — a journey that contributed to the government’s delay in bringing the defendants to trial. While the tickets for the journey and the post-arrival events bear a patina of spuriousness, this does not guarantee a constitutional sanctuary at journey’s end for the defendants.

We must engage in the calculus of delay, balancing the competing interests of the government and the defendants. We must examine the attitudes and the conduct of government prosecutors and agents during the journey, through its unscheduled stops, and upon its end at the trial destination.

In our contemporary world, we measure the speed of light and sound by mathematical formulae. But the more mundane concept of speedy trial is not susceptible to formula or table. It is, rather, a composite of objective and subjective factors that defies the mathematicians and confounds the philosophers — hence the necessity of the tome of doubts, resolved, we trust, as the law commands and fairness dictates.

Were we convinced that the government’s unhurried and petty pace impeded the protection of constitutional principle, we would not hesitate to reverse. On the other hand, because we rule on a speedy trial challenge, however strong our distaste for the government’s misconduct we must focus equally on the length of delay and the prejudice caused thereby. We affirm.

The appellants, Rudolfo Castrillon and Jose Avalos, were arrested for conspiring to violate federal narcotics laws under an arrest warrant issued in the District of Columbia on October 26, 1972. The arrest warrants were subsequently dismissed. On August 16, 1973, the appellants were indicted along with six others in the Southern District of Florida for conspiring to violate 26 U.S.C. § 4705(a), 7237(b) and 21 U.S.C. § 174. The charges arose from the same conspiracy for which they had previously been arrested in the District of Columbia. The appellants’ trial in the Southern District began on January 22, 1974. A jury found them guilty on January 28, 1974. The appellants filed a motion for a new trial, which was granted on May 7, 1974. 1 On August 27, 1974, the appellants’ second trial began. Three days later the jury returned a guilty verdict against both defendants. Castrillon and Avalos bring this appeal.

The appellants argue, inter alia, that each has been denied his sixth amendment right to a speedy trial. One fact is plain. A warrant issued for the appellants’ arrest in October 1972, fully fifteen months before their first trial for the offenses alleged in the warrant. The story of those fifteen months is a tale of two cities, Miami and Washington, and of two prosecutions. The narrative appears as twisted and difficult to decipher as the motives of its principals.

1. Facts

In September 1972, the Bureau of Narcotics and Dangerous Drugs 2 received tips from informants implicating the appellants in a conspiracy. The investigation of the appellants was part of a general investiga tion of Cuban-Americans allegedly engaged in the illegal sale of narcotics in the District of Columbia. Agent Albert Puglia conducted the investigation, which identified about fifty possible defendants. The United States Attorney’s Office in the District of Columbia divided the investigation into two conspiracy prosecutions, each with different defendants. With this bifurcation our story branches; it is necessary to follow each branch to its conclusion.

On October 26, 1972, a grand jury in the District of Columbia returned an indictment charging twelve defendants with conspiracy to violate narcotics laws. 3 Neither Avalos nor Castrillon was among the suspects indicted. On March 29, 1973, the government moved to dismiss the indictment. The prosecutor stated that he took this action because the health of one key witness had deteriorated, jeopardizing the appearance of the witness at the June trial, and because the government had discovered that another important witness, Jose Zayas, was unreliable. Nine months later the government obtained another indictment, this time in the Southern District of Florida. The case was set for trial in May 1974. On motions to transfer the proceeding back to the District of Columbia pursuant to Fed.R.Crim.P. 21(b) and to dismiss for lack of a speedy trial, Chief Judge Fulton of the Southern District found that “in seeking the indictment of these defendants in the Southern District of Florida the government was ‘Court Shopping.’ ” He observed that the majority of the defendants resided in or near the District of Columbia, that most of the overt acts allegedly took place there, and that in the District of Columbia prosecution the government had opposed Judge Gesell’s pre-trial order requiring the prosecution to divulge substantially in advance of trial the names of its witnesses and their grand jury testimony. Chief Judge Fulton concluded that the Bureau brought the case to Florida “believing that government counsel and the Court would be more favorable to the government’s cause [there] than in the District of Columbia.” United States v. Lara, 172 U.S.App.D.C. 60, 520 F.2d 460, 462 (1975) (quoting Chief Judge Fulton). He ordered the case returned to the District of Columbia.

On May 16, 1974, in the District of Columbia, Judge Gesell granted the motion to dismiss for want of speedy trial as to those defendants named in the earlier indictment. The Court of Appeals affirmed, agreeing that “when the government shuttled the case to Florida it was deliberately seeking the supposed advantage of more favorable treatment in that jurisdiction.” United States v. Lara, supra, 520 F.2d at 465 (1975). The court held that because “the decision to dismiss in the District of Columbia’ and begin the case anew in Florida was dictated by tactics and not by necessity the government must be charged with these nine months of delay.” Id. at 465.

On the same day it obtained the ill-fated indictment that led to Lara, the government obtained arrest warrants for several other suspects. Included in this latter group were Castrillon, Avalos, and their future co-defendants in the Southern District. The warrants followed a complaint filed on October 25, 1972, by Agent Puglia in the United States District Court for the District of Columbia, charging the appellants with violations of 26 U.S.C. § 4705(a), 7237(b). Avalos was arrested and jailed in Puerto Rico. He was shortly removed to the District of Columbia, where he was again incarcerated. Castrillon was apparently never brought to the District of Columbia to face charges. 4 Neither was ever indicted in the District of Columbia.

At a preliminary hearing for one of the appellants’ future co-defendants, the presiding magistrate found that there was not probable cause to hold the accused. The government subsequently elected not to seek immediate indictments against all those, including the appellants, who had been arrested but not yet indicted. Avalos was released after being imprisoned in the District of Columbia for one week. Prosecutors took no action against Castrillon, who apparently had remained in Florida.

The government did not dismiss the arrest warrants, however, until April 1978. The government contends that from November 1972 through March 1973, the United States Attorney’s Office in Washington and Agent Puglia continued to interview witnesses, concentrating on the case involving those suspects who had been indicted.

Agent Puglia was, by all accounts, a busy man during the spring of 1973, conducting his own brand of shuttle diplomacy between the District of Columbia United States Attorney’s Office and two trial attorneys of the Department of Justice in Florida. Upon dismissing pending charges against the indicted defendants on March 29, 1973, the government decided to bring the entire investigation to the more salubrious climate of the Southern District of Florida. Pursuant to its southern strategy, the government finally dismissed the District of Columbia arrest warrants against the appellants in April.

On August 16, 1973, a grand jury returned an indictment in the Southern District of Florida against the appellants and six others. An October trial date was scheduled. By October, however, Jose Avalos and two co-defendants had not yet been located. The government requested a thirty to sixty day continuance. The trial was rescheduled for January 1974. During the January trial, at which the appellants were convicted, Jose Zayas testified against them. The appellants moved for a new trial, claiming that the District of Columbia prosecutor’s statement repudiating Zayas’ testimony should have been turned over to the defense under Brady v. Maryland, 373 U.S. 83, 83 S.Ct. 1194, 10 L.Ed.2d 215 (1963). The district court granted the appellants’ motion for a new trial. The second trial resulted in the convictions now before this Court.

The appellants contend principally that the fifteen month delay between their arrest and trial violates the sixth amendment’s guarantee of speedy trial. They allege that the sole reason for this delay was that cited in United States v. Lara, supra —the government’s intentional delay to gain a tactical advantage through “court-shopping.” 5 The appellants also allege that they were deprived of due process during the period of pre-indictment delay between the alleged completion of the conspiracy in 1969 and the August 16, 1973 indictment. In addition, the appellants challenge on constitutional grounds the composition of the grand jury and the conduct of the trial. We find that there was no due process or sixth amendment violation and that the additional claims of the appellants do not require reversal.

II. Speedy Trial

Two time periods are relevant to consideration of whether the appellants’ constitutional rights were violated by the government’s delays. We shall consider first the time period between the commission of the substantive offense and the appellants' arrest. The issue will be solely whether this delay violated the appellants’ rights to due process of law. We shall then consider the interval between the time the appellants were designated “accused” persons in the District of Columbia and the date of their trial in the Southern District of Florida. Sixth amendment rights attach during the latter period.

A. Pre-Accusation Delay

The sole constitutional limit on the delay between the commission of a substantive offense and the initial accusation based on that conduct is the due process clause. In United States v. Marion, 404 U.S. 307, 92 S.Ct. 455, 30 L.Ed.2d 468 (1971), the Supreme Court rejected the notion that the sixth amendment limits the time within which an initial charging paper must be filed against an unarrested subject. 6 This Circuit had held even before Marion that delay in indictment did not violate the sixth amendment or mandate dismissal under Fed.R.Crim.P. 48(b) 7 so long as the statute of limitations was followed. 8 See Kroll v. United States, 433 F.2d 1282, 1286 (5th Cir. 1970), cert. denied, 402 U.S. 944, 91 S.Ct. 1618, 29 L.Ed.2d 112 (1971) [four years between termination of conspiracy and indictment did not violate sixth amendment or Rule 48(b)]; United States v. Grayson, 416 F.2d 1073, 1076-77 (5th Cir. 1969), cert. denied, 396 U.S. 1059, 90 S.Ct. 754, 24 L.Ed.2d 753 (1970). Cf. United States v. Ewell, 383 U.S. 116, 122, 86 S.Ct. 773, 777, 15 L.Ed.2d 627, 631 (1966).

If there is unnecessary delay in presenting the charge to a grand jury or in filing an information against a defendant who has been held to answer to the district court, or if there is unnecessary delay in bringing a defendant to trial, the court may dismiss the indictment, information or complaint.

The pre-accusation delay would constitute a violation of the appellants’ due process rights if they could show (1) that they incurred substantial prejudice as a result of the government’s delays and (2) that the prosecution had intentionally employed the delay to gain a tactical advantage. See United States v. Marion, supra, 404 U.S. at 325, 92 S.Ct. at 466; United States v. Duke, 527 F.2d 386 (5th Cir. 1976); United States v. Butts, 524 F.2d 975 (5th Cir. 1975); United States v. Beckham, 505 F.2d 1316, 1319 (5th Cir.), cert. denied, 421 U.S. 950, 95 S.Ct. 1683, 44 L.Ed.2d 104 (1975); United States v. Judice, 457 F.2d 414 (5th Cir.), cert. denied, 409 U.S. 886, 93 S.Ct. 104, 34 L.Ed.2d 142 (1972).

The appellants fail to satisfy either requirement of the due process test. 9 First, they fail to show that they were substantially prejudiced by the delay. This Court has consistently held that speculative allegations, such as general allegations of loss of witnesses and failure of memories are insufficient to establish the requisite actual prejudice. See United States v. McGough, 510 F.2d 598, 604 (5th Cir. 1975) (death of six potential defense witnesses did not show actual prejudice); United States v. Butts, supra, 524 F.2d at 977; United States v. Zane, 489 F.2d 269, 270 (5th Cir. 1973); United States v. Beckham, supra, 505 F.2d at 1319. But see United States v. Barket, 530 F.2d 189, 196 (8th Cir. 1976) (where there was forty-seven month delay between transaction and indictment, government had burden of demonstrating that missing witnesses did not possess exculpatory evidence). The appellants have alleged only in the most general and conclusory fashion that the government’s delay resulted in faded memories and a loss of potential defense witnesses. 10

The appellants also fail to satisfy the second prong of the due process test. There is no evidence to suggest that the .delay between the criminal act charged and appellants’ arrest was a tactical move by the government. In United States v. Butts, supra, 524 F.2d at 977, we said that “the government had every right to delay seeking an indictment until it had admissible evidence sufficient for conviction.” The appellants assert only that the government knew about the conspiracy before October 1972 and could have initiated the prosecution sooner. The same could be said of almost every complex conspiracy case; this scarcely shows the requisite tactical delay. In sum, the government’s pre-accusation delay did not violate the appellants’ constitutional rights.

B. Post-Accusation Delay

Dillingham v. United States, supra, establishes that arrest constitutes the initiation of a prosecution for purposes of applying the speedy trial test of Barker v. Wingo, 407 U.S. 514, 92 S.Ct. 2182, 33 L.Ed.2d 101 (1972), regardless of whether a formal indictment has been returned. 11 See United States v. Marion, supra, 404 U.S. at 459, 92 S.Ct. 455; United States v. Duke, supra, 527 F.2d at 388. The government designated the appellants accused persons when the original warrants were issued for their arrest and they were arrested. 12 That Castrillon was never incarcerated and Avalos was imprisoned for a relatively short period does not expunge the public accusation. 13 What is crucial is that the appel lants were arrested in the District of Columbia for the same activities for which they were subsequently prosecuted in the Southern District of Florida. In Gravitt v. United States, supra, we said that “[f]or purposes of determining when the right to speedy trial attaches, the basis for arrest is critical.” 14 Because the activities forming the basis of the conviction under review are those for which the appellants were originally arrested, we hold that the right to a speed trial attached on the date of the initial arrests. 15

Having marked the inception of the speedy trial period, we must next locate its terminus. It might be suggested that the applicable period should run until the appellants’ second trial, in August 1974, rather than their first trial, in January 1974. To be sure, the government created the necessity of a second trial by virtue of its failure to proffer Brady material to the defense at the first trial. Were the speedy trial period uniformly extended from arrest through retrial, however, then in many cases in which a conviction was reversed the government would be barred from retrying the defendants because the resulting delay would violate the sixth amendment.

This Court has held the government responsible for conduct that led to a mistrial and attendant delay. In United States v. Joyner, 494 F.2d 501 (5th Cir.), cert. denied, 419 U.S. 995, 95 S.Ct. 308, 42 L.Ed.2d 268 (1974), a mistrial was granted over the objection of defendant Joyner, who wished to commence trial immediately, where the defendants had been led, chained and handcuffed, into the courtroom in the presence of the jury. We said that the resulting postponement was chargeable to the government. 494 F.2d at 505. Similarly, in United States v. Pena, 527 F.2d 1356, 1364 (5th Cir.), cert. denied, - U.S. -, 96 S.Ct. 3168, 49 L.Ed.2d 1185 (1976), we counted the duration between a mistrial and defendant’s second trial as part of the speedy trial period. 16

In this case unlike Joyner, or Pena, the first trial was completed. The defense presented its entire case, and the jury reached a verdict. Moreover, even if we were to count the delay between the first and second trials as part of the speedy trial period, this delay would not contribute to any violation of the sixth amendment. There is no allegation that the government initiated the first trial as a dilatory tactic. There was, on the facts of this case, no prejudice to the appellants from the delay between the first and second trial. No witnesses for the defense at the first trial were unable to testify at the second. Any fading memories should have been sufficiently crystallized by the first trial so that no incremental prejudice was caused by the delay between trials. In any event, the appellants themselves measure the speedy trial period as fifteen months, and therefore do not contest the matter.

Thus, for purposes of determining whether the appellants were denied their sixth amendment right to a speedy trial, we shall consider only the fifteen-month duration between the government’s accusation in October 1972 and the first trial in January 1974.

C. Applying the Speedy Trial Test

Barker v. Wingo, supra, 407 U.S. at 530-33, 92 S.Ct. at 2192-93, established a four-pronged test to be applied on an ad hoe basis in all cases in which a speedy trial claim is raised. The four factors to be considered are the length of delay, the reason for the delay, the defendant’s assertion of his right, and prejudice to the defendant. None of the four factors is either a necessary or sufficient condition to finding a denial of the right to speedy trial. The test is more suggestive than exhaustive, and this Court has found implicit in Barker other factors to be thrown into the balance, including the number and complexity of legal and factual issues involved in a case and the availability of evidence. 17

The Supreme Court has described the right to speedy trial as “amorphous” and “slippery.” Barker v. Wingo, supra, 407 U.S. at 522, 92 S.Ct. at 2188. It is impossible precisely to determine when the right has been violated, when the government has gone too far. The right to speedy trial is unlike other constitutional rights because deprivation of the right may work to the accused’s advantage. The fading of witnesses’ memories with time may disadvantage the government far more than the accused because the government carries the burden of proof. When, as in this case, the prejudice to the appellants is unclear, the time of delay is not per se oppressive, and the reasons for delay are a matter of sharp disagreement, 18 our task is particularly difficult.

1. Length of Delay

The length of the delay serves in part merely as a “triggering mechá nism.” Barker v. Wingo, supra, 407 U.S. at 530, 92 S.Ct. at 2192. Having determined that the length of delay is fifteen months, the threshold question is whether this delay requires further consideration of the speedy trial claim. We are cognizant that a greater delay may be tolerated in prosecuting a complex conspiracy than in prosecuting an ordinary street crime. 19 See id., 407 U.S. at 531, 92 S.Ct. at 2192; see note 17, supra. Nevertheless, we shall consider the complex nature of this case as a potential reason for the delay; in this case a fifteen month delay is sufficient seriously to consider the reasons for delay and other elements of the appellants’ speedy trial claim. See Gravitt v. United States, supra, 523 F.2d at 1214 (sixteen month delay required evidentiary hearing on speedy trial claim); Arrant v. Wainwright, 468 F.2d 677 (5th Cir. 1972) (almost two-year delay was “clearly sufficient” to require serious consideration of speedy trial claim).

2. Reasons for Delay

In considering speedy trial claims, 20 we have recognized three classes of reasons for delay in bringing a defendant to trial. See Turner v. Estelle, 515 F.2d 853 (5th Cir. 1975), cert. denied, 424 U.S. 955, 96 S.Ct. 1431, 47 L.Ed.2d 361 (1976). We shall attempt to identify the reasons that account for the post-accusation delay in the case at bar, the weight to be accorded each class of reasons under applicable precedent, and, to the extent possible, the portion of the delay attributable to each class of reasons.

(a) Deliberate delay

“A deliberate attempt to delay the trial in order to hamper the defense should be weighted heavily against the government.” Barker v. Wingo, supra, 407 U.S. at 531, 92 S.Ct. at 2192. In Arrant v. Wainwright, supra, 468 F.2d at 680-81, we said that the state’s fifteen month delay in requesting a trial date would be weighed heavily against the state when the prosecutor could justify the delay only by averring that he did not want to see the defendant acquitted of murder. An eyewitness to the killing had recanted her previous statement incriminating the defendant and had indicated she would offer exculpatory testimony, whereupon she was charged with perjury and held on $50,000 bond. Testimony established that when she affirmed her earlier statement the state dropped charges against her and offered to help her find a job. The delay in the defendant’s trial made possible these machinations by the state.

Arguing that the prosecution was guilty of deliberate delay in the instant case, the appellants observe correctly that in United States v. Lara, supra the district court judge in Florida found the government guilty of “Court-Shopping,” and the district court judge in the District of Columbia dismissed the indictments because the defendants had been denied speedy trial. The appellants argue that the same motives that infected the government’s conduct in Lara taint the prosecution in this case. The Assistant United States Attorney for the District of Columbia, William Brodsky, conceded that the original Lara indictment in the District of Columbia was dismissed in part because the government believed it could no longer sponsor the testimony' of Jose Zayas, who testified against the appellants at their first trial. The appellants claim that the government sought prosecutors in Florida who would be willing to sponsor Zayas in the case at bar and in Lara, and that the delay in the appellants’ trial is attributable to the government’s court-shopping.

After careful consideration of the record, it is apparent that one of the government’s motives in dismissing the District of Columbia complaint and seeking a Florida indictment was to find prosecutors who would use Zayas’ testimony against the appellants. The simple fact is that the Florida prosecutors did use Zayas’ testimony. The government’s conduct is to be condemned, and had it caused a longer delay or prejudiced the appellants, we should not hesitate to reverse their convictions.

In the case at bar, however, the government moved swiftly once it had dismissed the District of Columbia arrest warrants. That alone would distinguish this case from Lara. In Lara, the defendants had already been scheduled for trial on June 11, 1972, before the dismissal of the District of Columbia indictment on March 29, 1973. A critical fact for the District of Columbia Circuit in Lara was that nine months had elapsed between the dismissal of the first indictment and the entry of the Florida indictment on December 17, 1973. 21 By contrast, in the case at bar, the appellants were neither indicted nor assured of a trial date in the District of Columbia. More important, after the dismissal of the District of Columbia arrest warrants in April 1973, evidence was presented to the Florida grand jury in July and the indictment returned in August 1973, only four months after dismissal of the warrants. 22 An October trial date was scheduled, but by October only three of the eight defendants had been arraigned in Miami. Jose Avalos, for example, had not been located. The government requested a thirty-sixty day continuance. The trial was moved to January 1974.

In sum, the government’s transferring the prosecution to Florida likely cost the appellants no more than about four months. 23 Although it is impossible to apportion fault for the delay with mathematical precision, it appears that attempts to gain a tactical advantage, if that is indeed what they were, account for only a small portion of the fifteen month interval between the appellants’ arrest and trial. 24

This was not a case in which the government delayed to harass the appellants or intended specifically to impair their defense, as by delaying to prevent a key witness from testifying for the defense. 25 Although deserving of censure, the government’s conduct in this case was not so much calculated to injure the defense as it was to strengthen the prosecution’s case. Although the result was, of course, injurious to the defense insofar as Zayas’ testimony proved damaging, and thus whether the government specifically intended to help the prosecution or hurt the defense might appear irrelevant as a practical matter, we .should doubtless be far less tolerant had the prosecution tampered with the appellants’ defense. See note 25, supra ; note 26, infra. In considering solely the reasons for delay, the primary focus is not prejudicial effect but culpability.

We conclude that some portion of the delay was attributable to tactics, and is therefore to weigh heavily against the government. The delay directly caused by tactics was about four months. The government’s misconduct, although reprehensible, was not so opprobrious that the length of delay caused thereby becomes irrelevant. 26 That the government’s tactical pursuit of a hospitable forum for its presentation of an unreliable witness directly delayed appellants’ trial by four months is not independently sufficient to overturn the appellants’ convictions on a speedy trial challenge.

(b) Negligent delay

The second class of reasons for a post-accusation delay involves more “neutral” reasons such as negligence or overcrowded courts. A delay attributable to reasons of this class “should be weighted less heavily but nevertheless should be considered since the ultimate responsibility for such circumstances must rest with the government rather than the defendant.” Barker v. Wingo, supra, 407 U.S. at 531, 92 S.Ct. at 2192. See United States v. Rodriguez, 510 F.2d 1 (5th Cir. 1975); United States v. Shepherd, 511 F.2d 119, 122 (5th Cir. 1975). Some portion of the fifteen month delay, particularly the interval between accusation and the dismissal of the District of Columbia arrest warrants, is attributable to such neutral reasons. The government was focusing its efforts during this period on the Lara defendants, not on these appellants.

(c) Justified delay

Finally, some portion of the fifteen month delay is attributable to the third class of reasons. “[A] valid reason, such as a missing witness, should serve to justify appropriate delay.” Barker v. Wingo, supra, 407 U.S. at 531, 92 S.Ct. at 2192. See United States v. Atkins, 528 F.2d 1352 (1976); Turner v. Estelle, supra, 515 F.2d at 857.

The fifteen month delay was overdetermined — the result of many causes, the product of a mixture of motives that even the prosecutors themselves could not have sorted out. Some indeterminate fraction of the delay resulted from the government conduct that was entirely justified. For much of the fifteen month period, for example, the government continued to gather evidence against the appellants. In June 1973, the investigation bore fruit; the government elicited information from an unindicted co-conspirator that led directly to the government’s charge of one of the overt acts contained in the indictment. In United States v. Rosson, 441 F.2d 242 (5th Cir.) cert. denied, 404 U.S. 843, 92 S.Ct. 140, 30 L.Ed.2d 78 (1971) 27 we held that a reasonable delay that enabled the government to accumulate the evidence necessary to make out a prima facie case against an accused did not prejudice the accused by impairing his ability to defend himself. In Turner v. Estelle, supra, 515 F.2d at 856-57, we said that such a reason fell into the third class of justifications and weighed heavily against finding a denial of speedy trial.

It is of critical importance that at least some of the fifteen month delay was justified by the government’s discovering new evidence to augment its case. Moreover, even the entire fifteen month period was not unusually oppressive in the context of a complex interstate conspiracy, for which relatively greater delays are to be expected. 28 Finally, the government cannot be charged with the delay, from October 1973 to January 1974, which is instead largely attributable to the appellants. The trial date was moved from October after the government’s presumably bona fide efforts to locate Avalos had failed. Castrillon cannot complain of this latter delay because he did not ask to be tried separately, but chose instead to wait until Avalos had been found. See United States v. Garner, 529 F.2d 962, 967 (6th Cir. 1976). During this period, the defense asked for a continuance. In addition, the court was kept busy by a flood of pre-trial motions raised by the defense. See United States v. Scallion, 533 F.2d 903 (5th Cir. 1976).

The government should be charged neither with that part of the delay for which it had a valid investigative or prosecutorial justification nor for that part of the delay resulting from its inability to locate some of the co-defendants by the original October trial date.

In sum, that portion of the fifteen month interval between the appellants’ arrest and trial for which the government is responsible is in part attributable to apparent government misconduct and in part attributable to the prosecutor’s concentrating on another case. Much of the fifteen month delay does not weigh at all against the government. We conclude that although the reasons for the delay make out, on balance, a good case for denial of speedy trial, the case is by no means compelling and will fail unless supported by the remaining elements of the Barker test.

3. Assertion of the Right to Speedy Trial

The third factor to be balanced is whether the appellants asserted their right to speedy trial. Barker v. Wingo, supra 407 U.S. at 531, 92 S.Ct. at 2192. Although the appellants did assert their rights, they did not do so in timely fashion. The appellants do not allege that they requested speedy trial in the District of Columbia prior to the dismissal of the arrest warrants in April 1973. According to the record, appellant Castrillon first raised the question of speedy trial in his motion to dismiss on September 21, 1973, almost eleven months after his arrest and only four months before his first trial. Jose Avalos first raised the question of speedy trial in adopting Castrillon’s motion to dismiss on January 3,1974. 29 Both appellants complained primarily of the pre-accusation delay, rather than post-accusation delay. Neither demanded an immediate trial. See United States v. Palmer, 537 F.2d 1287 (5th Cir. 1976). As we said in Palmer, to the extent that promptness in asserting the right is important, then appellants’ silence during the entire pre-indictment period and some of the post-indictment period works against them “because it suggests that any hardships [they] suffered were either minimal or caused by other factors.” Id. at 1288. Moreover, as we have seen, the government was granted a continuance and the first trial was rescheduled from October 1973 to January 1974 because some defendants, including Avalos, were not present. The record does not indicate that Castrillon interposed any objection to the postponement. In sum, the record does not show that either before or after indictment the appellants aggressively asserted that desire for speedy trial which they now urge has been thwarted. That fact will militate against a claim that speedy trial was denied. See Barker v. Wingo, supra, 407 U.S. at 534-35, 92 S.Ct. at 2194. 30

4. Prejudice to the Appellants

The fourth element to be balanced is the prejudice to the appellants that resulted from the delay. In order to appraise the prejudice caused by the fifteen month delay in the case at bar, we must look to the interests sought to be protected by the speedy trial guarantee. The Supreme Court has identified three such interests: (i) to prevent oppressive pre-trial incarceration; (ii) to minimize anxiety and concern of the accused; and (iii) to limit the possibility that the defense will be impaired. Barker v. Wingo, supra, 407 U.S. at 533, 92 S.Ct. at 2193.

The government did not abuse the first of these interests in this case. Castrillon was never incarcerated, and Avalos was imprisoned only for a week in the District of Columbia. The appellants argue that the government failed to protect the second interest, but submit only vague complaints regarding the anxiety they experienced during the fifteen month delay. Anxiety of the sort “present to some degree in virtually every case” does not amount to actual prejudice. See United States v. Shepherd, supra, 511 F.2d at 123; cf. Arrant v. Wainwright, supra, 468 F.2d at 682. Were the appellants shown to have suffered particular hardships or was this a capital case, where the anxiety would justifiably be more severe, we would accord the appellants’ allegations greater weight.

Finally, the appellants fail to allege with particularity impairment of their defense. They rely on vague assertions of faded memory and lost witnesses without connecting this loss to any material fact in issue. Where the appellants allege no more than the general inability to recall a transaction with desired specificity that occurs in any trial regardless of delay, and where such faded memories do not appear substantially to relate to any material fact in issue, no actual impairment of their defense for purposes of a speedy trial challenge is shown. See United States v. Shepherd, supra, 511 F.2d at 124. See also United States v. Pena, supra, 527 F.2d at 1364; Turner v. Estelle, supra, 515 F.2d at 860; United States v. Rodriguez, supra, 510 F.3d at 3; United States v. Lane, supra, 465 F.2d at 412. 31 We conclude that the record demonstrates no actual prejudice.

Careful application of the sixth amendment’s guarantee of a speedy trial, however, requires distinguishing three levels of prejudice that, in conjunction with various constellations of other elements of the Barker test, may require reversal of a conviction.

First, a showing of actual prejudice to the conduct of the defense will weigh heavily against the government and may lead to dismissing the indictment even when the three remaining factors are not weighted heavily in favor of the accused. When a showing of actual prejudice is conjoined with a lengthy delay by the government that is unexplained or deliberate, then a defendant states a compelling case for denial of speedy trial. See Arrant v. Wainwright, supra, 468 F.2d at 683.

Second, where the government’s lengthy delay is unexcused or purposeful, in the sense of a deliberate delay to gain tactical advantage, the government’s delay is prima facie prejudicial. The government will have the burden of demonstrating that the defendant has not been prejudiced by the delay. See Murray v. Wainwright, supra, 450 F.2d at 471.

Third, when the three other elements of Barker’s weighing and balancing test are heavily weighted in favor of the accused, the accused need demonstrate no prejudice at all. 32 “Prejudice — either actual or presumed — becomes totally irrelevant.” Hoskins v. Wainwright, supra, 485 F.2d at 1192. See Prince v. State of Alabama, supra, 507 F.2d at 706-07.

We have seen that on the facts of this case the other three factors are not heavily weighted in favor of the accused. The fifteen month delay, though sufficient to trigger serious consideration of the appellants’ claim, is not per se oppressive or clearly excessive in the prosecution of a complex interstate conspiracy. This was not a case in which the government specifically intended to impair the appellants’ defense or prejudice the accused by dilatory tactics. The government apparently did seek a tactical advantage through “court-shopping,” and some delay resulted. The resulting delay, which counts heavily against the government, amounted only to about four months. Some of the remaining delay is excused and not properly chargeable to the government.

This was not a case where the “government’s lengthy, inexplicable delay, in the face of vigorous demands for an immediate trial, is so offensive to the Sixth Amendment’s guarantee of a speedy trial that” no prejudice need be shown. Hoskins v. Wainwright, supra, 485 F.2d at 1189 n. 3 (eight year delay). Nor was this a case in which the prima facie prejudicial delay resulting from governmental misconduct or negligence exceeded the delays normally incurred in prosecuting the kind of offense charged where we might sustain the appellants’ speedy trial challenge in the absence of a showing of actual prejudice. See Murray v. Wainwright, supra, 450 F.2d 470-72 (six years delay). The appellants failed to demonstrate that the delay resulted in any actual prejudice. We conclude that on the facts of this case the appellants have not been denied their sixth amendment rights to speedy trial.

III. Other Issues

A. Scope of Cross-Examination

The appellants argue that the trial judge erred by prohibiting the appellants from asking two government witnesses for their present addresses. The appellants rely on Alford v. United States, 282 U.S. 687, 51 S.Ct. 218, 75 L.Ed. 624 (1931), and Smith v. Illinois, 390 U.S. 129, 88 S.Ct. 748, 19 L.Ed.2d 956 (1968). We examined these cases in United States v. Alston, 460 F.2d 48 (5th Cir. 1972). We said that Alford and Smith require the admission of background information, such as a witness’s address, that is essential to identify the witness with his environment, “to place the witness in his proper setting and put the weight of his testimony and his credibility to a test . .” United States v. Alston, supra, 460 F.2d at 51, quoting Alford v. United States, supra, 282 U.S. at 692, 51 S.Ct. at 219. We concluded that the Supreme Court had established no hard and fast rule regarding questions about present addresses. We specifically excepted from any requirement situations in which the physical safety of the witness or his family might be endangered by the disclosure. Subsequently, in United States v. Crockett, 506 F.2d 759, 762 (5th Cir.), cert. denied, 423 U.S. 824, 96 S.Ct. 37, 46 L.Ed.2d 40 (1975), we held that where the government had reason to believe that two government witnesses’ lives were in danger and the witnesses were under the protection of a United States Marshal, the court did not err in prohibiting the defense from asking on cross-examination for the witnesses’ present addresses. We added that despite the restriction on cross-examination the defendants had been able to identify the witnesses with their environment.

This case falls within the same exception. Both government witnesses had been placed in the Department of Justice Witness Protection Program and one had recently been re-located as a precautionary measure. Both testified that they were in danger and that the government was giving them subsistence allowances to enable them to stay off the street. Moreover, as in Crockett, the appellants fail to demonstrate that prejudice resulted from the court’s ruling. Both witnesses talked freely about their residences during the period of the conspiracy. The defense was able to identify them with their environment. We conclude that the trial court did not err by prohibiting the appellants from asking the witnesses for their present addresses.

B. Continuance

Jose Avalos contends that the trial judge erred in denying his motion for a continuance. A motion for a continuance is addressed to the sound discretion of the trial court. The district court’s ruling will not be disturbed on appeal unless there is a showing of an abuse of that discretion. See, e. g., United States v. Uptain, 531 F.2d 1281, 1285 (5th Cir. 1976); United States v. Gidley, 527 F.2d 1345 (5th Cir. 1976); United States v. Sahley, 526 F.2d 913 (5th Cir. 1976). The issue must be decided on a case by ease basis in light of the circumstances presented.

On August 29, 1974, at the conclusions of the government’s case, Avalos said that he would present one witness who was to arrive that afternoon. On the following day, Avalos advised the court that the witness had not yet arrived but that he thought the witness would arrive the next day. The court observed that there was no certainty the witness would arrive the next day and denied the appellant’s motion. Counsel for Avalos stated that the witness had been served with a subpoena the day before trial and left the country the same day, aware of the impending trial. The witness would have testified that a government witness had a poor reputation for veracity. The defense had already used the testimony of an Assistant United States Attorney to impeach the credibility of the government witness. We cannot say that the trial court’s denial of the continuance constituted an abuse of discretion.

C. Sufficiency of the Evidence

Castrillon contends that it was error not to direct a verdict of acquittal. This argument simply attacks the sufficiency of the evidence used to connect him to the conspiracy. Castrillon claims that the evidence proved only his association with the conspiracy, not that he was a co-conspirator. That Castrillon did not know each of the other co-conspirators and was not aware of each part of the unlawful plan is entirely compatible with the jury’s conclusion that he was nevertheless a co-conspirator. See United States v. James, 528 F.2d 999 (5th Cir. 1976); United States v. Rodriguez, 509 F.2d 1342, 1348 (5th Cir. 1975). For purposes of an attack on the sufficiency of the evidence, we must adopt the view most favorable to the government. The verdict of the jury must be sustained if there is substantial evidence to support it. Glasser v. United States, 315 U.S. 60, 80, 62 S.Ct. 457, 86 L.Ed. 680 (1941). A careful review of the record persuades us that the jury had ample evidence from which to infer Castrillon’s guilt beyond a reasonable doubt.

D. Exclusion of resident aliens from grand jury

Finally, the appellants contend that the exclusion of resident aliens from the grand jury and trial venires denied them equal protection of the laws. The Supreme Court has recently rejected this contention. In Perkins v. Smith, 370 F.Supp. 134 (D.Md. 1974), aff’d mem., - U.S. -, 96 S.Ct. 2616, 49 L.Ed.2d 368 (1976), the Court affirmed the proposition that federal and state laws excluding resident aliens from grand and petit jury service do not violate due process or equal protection.

IV. Conclusion

In sum, the prosecutors’ zeal for a conviction, which led the defendants through two jurisdictions and contributed to delaying their trial, did not so deeply taint these convictions that we must reverse. In the weighing and balancing test of Barker v. Wingo and its progeny, the difficulty is that the scales often do not reveal a clear result. The difficulty goes much deeper, however, for not only are the weights we place on the scale fashioned of a mixture of fact and supposition, but all too often the measures of these weights themselves are shifting and unstable.

In coming to a conclusion of affirmance, we do so in full knowledge that in this case the line is thin and the area gray. But we are convinced that the government has not crossed the line and entered into the black area of purposeful or oppressive frustration of the appellants’ defense.

The judgment of the district court is

AFFIRMED.

1 . Two of the original eight defendants had obtained severance for trial, another entered a plea of guilty, one was a fugitive, one was acquitted, and another, Carlton Bryant, was found guilty after the first trial and did not join appellants’ motion for a new trial. Bryant instead directly appealed his conviction. This Court summarily affirmed. United States v. Bryant, 505 F.2d 1302 (5th Cir. 1974) (R. 21).

2 . The Bureau was a previous incarnation of the Drug Enforcement Administration.

3 . On January 31, 1973, the U.S. Attorney procured a superseding indictment naming the twelve original defendants and one additional defendant.

4 . Agent Puglia’s testimony and the government’s memorandum opposing the appellants’ motion to dismiss repeatedly refer to CastrilIon’s arrest. The record on appeal does not document his arrest, although it does show Castrillon named in the District of Columbia arrest warrant. There is no evidence that Castrillon was ever incarcerated. We shall proceed on the assumption that he was arrested but not incarcerated.

5 . Although we are not bound by the District of Columbia Circuit’s separate action on somewhat different facts in Lara, we must nevertheless evaluate the interval of fifteen months between the appellants’ arrest and trial in light of the reasons for this delay. See text at notes 20-26, infra.

6 . See generally Amsterdam, Speedy Criminal Trial: Rights and Remedies, 27 Stan.L.Rev. 525, 528 (1975).

7 . Rule 48 provides:

8 . The Supreme Court has since made clear, however, that the sixth amendment does limit pre-indictment delay once the defendant has been accused. Dillingham v. United States, 423 U.S. 64, 96 S.Ct. 303, 46 L.Ed.2d 205 (1975); United States v. Marion, supra, 404 U.S. at 323, 92 S.Ct. at 464-65, 30 L.Ed.2d at 480. This will become important in considering the merits of the appellant’s speedy trial claim. For present purposes, Dillingham and Marion enable us to delimit the period from the occurrence of the substantive offenses to the appellants’ arrests as one in which no speedy trial right attaches. The provisions of the sixth amendment do not require the government to discover, investigate, or accuse any person within a certain period of time. Marion, supra, 404 U.S. at 313, 92 S.Ct. at 459, 30 L.Ed.2d at 474.

9 . We observe only that a showing of both tactical delay and substantial actual prejudice would be sufficient to establish a violation, and that both are absent here. See United States v. Croucher, 532 F.2d 1042, 1044 (5th Cir. 1976). The origin of this two-prong test is United States v. Marion, supra, 404 U.S. at 324, 92 S.Ct. 455, where the Court announced merely that the government conceded that satisfaction of both requirements would be sufficient. There is no Supreme Court authority squarely holding that satisfaction of both elements of the test is necessary to' find a due process violation. The Court did intimate that a showing of tactical delay might often be required, observing that “[a]ctuai prejudice to the defense of a criminal case may result from the shortest and most necessary delay; and no one suggests that every delay-caused detriment to a defendant’s case should abort a criminal prosecution.” Id. 404 U.S. at 324—25, 92 S.Ct. at 465.

United States v. Beckham, supra, 505 F.2d at 1319, suggests that both requirements must be met in order to establish a due process violation. In subsequent cases we have, in dicta, interpreted Marion as supporting that proposition. See, e. g., United States v. Duke, supra, 527 F.2d at 388. Nevertheless, there remains substantial doubt whether, in a case in which actual pre-accusation prejudice was overwhelming, the government’s purposeful delay would have to be shown; or, alternatively, where the government’s misconduct was blatant, whether the defendant would still bear the burden of showing actual prejudice. In United States v. Barket, 530 F.2d 189, 194 (8th Cir. 1976), the Eighth Circuit observed:

“In previous cases, this and other circuits have not had occasion to determine whether Marion is to be read conjunctively or disjunctively because the circumstances before them did not include sufficient prejudice to the defense from pre-indictment delay to merit relief even if joined with a showing of serious governmental ‘gamemanship.’ ”

The Barket court found that negligent though inadvertent governmental conduct, when joined with prejudice to the appellant’s ability to present a defense, was sufficient to constitute a due process violation.

10 . We need not delineate the parameters of “substantial” prejudice. We decide only that no substantial prejudice was shown on the facts of this case.

11 . Until Dillingham, the rule in this Circuit had been that with respect to post-arrest, pre-indictment delay, a showing of actual prejudice was required to make out a denial of speedy trial. See United States v. Palmer, 502 F.2d 1233 (5th Cir. 1974), rev’d sub nom., Dillingham v. United States, supra, 423 U.S. 64, 96 S.Ct. 303, 46 L.Ed.2d 205 (1975) (per curiam) on remand sub nom., United States v. Palmer, 537 F.2d 1287 (5th Cir. 1976) [slip no. 73-1717, August 30, 1976]. Even before Dillingham, we had expressed reservations about the rule. In Gravitt v. United States, 523 F.2d 1211, 1215 n. 7 (5th Cir. 1975), rehearing denied, 526 F.2d 378 (5th Cir. 1976) (per curiam), we said: “To recognize that the speedy trial right has attached but to refuse to apply the Barker v. Wingo . . ., balancing test appears to be inconsistent with the teachings of United States v. Marion . . . and Barker.” [citations omitted]

12 . In Gravitt v. United States, supra, 523 F.2d at 1215, we held that the appellant’s right to a speedy trial attached prior to the return of the first indictment when he was in state custody, federal authorities knew where to locate him, a formal complaint was filed, and an arrest warrant issued.

13 . It might be argued that because the warrant for appellants’ arrest was dismissed in the District of Columbia in April 1973, the speedy trial period should begin with appellants’ subsequent indictment in Florida. We reject this argument here. In United States v. Beckham, supra, 505 F.2d at 1319, the defendant was arrested in Texas on charges of tampering with the mail in 1971, the complaint was dismissed in favor of a homicide prosecution in Pennsylvania, the mail tampering case was reopened when the Pennsylvania charges were later dropped, and defendant was indicted in 1973 and convicted in 1974. We suggested without deciding that he might have no speedy trial claim but might be restricted to a due process attack. Whatever the proper result in that case, where the incremental detriment to the defendant of mail tampering charges paled in relation to that of the pending homicide prosecution, here the arrest warrants were dismissed not because of unrelated charges pending in another court, but in order to prosecute the same charges in another court. To fail to measure the speedy trial period from initial arrest would allow the government to circumvent the speedy trial requirement by successively dismissing and reinstituting a complaint or indictment for the same offense. In United States v. McKim, 509 F.2d 769 (5th Cir. 1976), we measured the speedy trial period from the first indictment, in January 1973, to the defendant’s trial on a third indictment for the same charge in January 1974, where the first and second indictments had been dismissed; but see United States v. Davis, 487 F.2d 112 (5th Cir. 1973), cert. denied, 415 U.S. 981, 94 S.Ct. 1573, 39 L.Ed.2d 878 (1974), where the applicable speedy trial period was measured from the defendant’s second indictment on the same conspiracy charge.

14 . 523 F.2d at 1215 n. 6 [italics in original]. In Gravitt and United States v. Cabral, 475 F.2d 715 (1st Cir. 1973), the issue was whether the speedy trial period began when the defendant was arrested by a state officer. In Cabral, where a state police officer arrested the defendant for possession of a sawed-off shotgun, the speedy trial period for the prosecution of a federal weapons charge began at the time of initial arrest, notwithstanding the government’s claim that the weapons arrest was a mere temporary device to restrain the defendant until the state could arrest him for possession of stolen property. In Gravitt, we said that where the defendant had been arrested on state robbery charges, not on charges stemming from the federal firearms conviction under review, the speedy trial period did not begin on the date of arrest.

15 . The rationale for beginning the speedy trial period on the date of arrest is summarized by the Supreme Court in United States v. Marion, supra, 404 U.S., at 320-21, 92 S.Ct. at 463:

To legally arrest and detain, the Government must assert probable cause to believe the arrestee has committed a crime. Arrest is a public act that may seriously interfere with the defendant’s liberty, whether he is free on bail or not, and that may disrupt his employment, drain his financial resources, curtail his associations, subject him to public obloquy, and create anxiety in him, his family and his friends.

16 . In Pena, where an allegedly material witness who had not testified at the first trial had died during the interval between the first and second trials, we found there was no speedy trial violation despite a three year delay between arrest and second trial.

17 . Compare United States v. Dyson, 469 F.2d 735 (5th Cir. 1972) (in prosecution for draft evasion, twenty-two month delay following receipt of indictment was “excessive”) with United States v. Lane, 465 F.2d 408 (5th Cir. 1972) (in complex conspiracy case, post indictment delay of two and a half years was not excessive). See also United States v. Perez, 489 F.2d 51 (5th Cir. 1973), cert. denied, 417 U.S. 945, 94 S.Ct. 3067, 41 L.Ed.2d 664 (1974).

18 . Although the district court judge made no specific findings of fact relative to the reasons for delay, we do have the benefit of a full hearing on appellant Castrillon’s motion to dismiss under Fed.R.Crim.P. 48(b). In that hearing, held on November 2, 1973, Castrillon and the government offered conflicting versions of the scenario leading to the Florida indictment. The issue of the reason for the delay is briefed by both the appellants and the government. This distinguishes the instant case from United States v. Dyson, supra, 469 F.2d at 742, where we remanded “to permit the Government to make a showing to be followed by specific findings of fact” regarding the reason for the delay in that case, where the delay was otherwise unexplained.

19 . We have in the past routinely applied the Barker balancing test even in a complex conspiracy case in which the speedy trial period was thought to be nine or ten months. See United States v. Davis, supra, 487 F.2d at 116— 17. It is not clear how the Supreme Court can suppose the length of delay to be a “triggering mechanism,” implying that if the delay is too short the inquiry ceases, while at the same time insisting that none of the four factors is a necessary condition to finding a violation. See Barker v. Wingo, supra, 407 U.S. at 533, 92 S.Ct. at 2182.

20 . An excused or improper delay might lead to reversing the appellants’ convictions under either the due process clause or the speedy trial provision of the sixth amendment. If the government “intentionally delayed to gain some tactical advantage” over the appellants or “to harass them,” and the appellants could show that such conduct actually prejudiced the conduct of their defense, then we should find that the appellants were denied due process of law. See United States v. Marion, supra, 404 U.S. at 325, 92 S.Ct. at 466. We have already rejected the due process attack relative to the pre-accusation period. See text and notes at notes 6-10, supra. In the post-accusation period, the more stringent speedy trial standard, which does not require actual prejudice, applies as well; if we find no denial of speedy trial, a fortiori there is no violation of due process.

21 . “Since we hold that the decision to dismiss in the District of Columbia and begin the case anew in Florida was dictated by tactics and not by necessity the government must be charged with these nine months delay.” United States v. Lara, supra, 520 F.2d at 465. See id. at 466-67 (Wilkey, J., concurring).

22 . It might be suggested that the government’s plans for court-shopping delayed prosecution of the appellants even prior to the dismissal of the warrants in April 1973. This appears unlikely, since the Assistant United States Attorney for the District of Columbia did not repudiate Zayas’ testimony until March 1973, and the government had no reason to suppose that the prosecutor would not support his testimony before that. The government contends that it did not move for indictment of the appellants in the District of Columbia before April 1973 simply because the case remained incomplete and the government was concentrating its efforts on the defendants already indicted and set for trial in June. The reason for this segment of the delay does not excuse the government, which must be held responsible, but it is not the kind of intentional delay for tactical reasons that is to be weighed heavily against the government. Under Barker v. Wingo, supra, 407 U.S. at 531, 92 S.Ct. at 2192, neutral factors causing delay, such as negligence or overcrowded dockets, are to be weighed less heavily against the government than an attempt by the government to prejudice the defense by delay.

23 . Even this computation assumes that the appellants could have been indicted in April 1973 and that the interval between indictment and trial in the District of Columbia would have been no greater than it was in the Southern District of Florida.

24 . Compare Arrant v. Wainwright, supra, 468 F.2d at 681. In that case we found that the state’s “unsatisfactory” explanation for a fifteen month segment of a two year delay weighed heavily against it; considering the clear prejudice to the defendant, we reversed for denial of speedy trial. In Murray v. Wainwright, 450 F.2d 465, 471 (5th Cir. 1971), we said that “where the delay is not only excessive but the result of unexcused inaction or misconduct by the Government, it is prima facie prejudicial.” The unexcused delay in that case was six years, and we shifted to the state the burden of showing that the defendant was not prejudiced. We have recognized that “there must be some coalescence of the other three factors in a movant’s favor at which prejudice — either actual or presumed — becomes totally irrelevant.” Hoskins v. Wainwright, 485 F.2d 1186, 1192 (5th Cir. 1973) (unexplained eight-and-one-half year delay); see Prince v. Alabama, 507 F.2d 693 (5th Cir. 1975) (unsatisfactory reason for eight year delay); cf. Moore v. Arizona, 414 U.S. 25, 94 S.Ct. 188, 38 L.Ed.2d 183 (1973).

25 . Cf. Arrant v. Wainwright, supra, 468 F.2d, at 683:

“The allegation is clearly that appellant was denied the testimony of an admitted eyewitness to the killing because of the affirmative and highly questionable actions taken by the state attorney during the period that appellant’s trial was delayed.”

26 . In United States v. McKim, 509 F.2d 769 (5th Cir. 1975), we faced a similar problem. The defendant had been arrested for marijuana possession but escaped from prison. His conviction for marijuana possession was subsequently reversed on appeal. The government indicted McKim on the escape charge for the first time in January 1973, but dismissed both that and a second indictment without the defendant’s consent. Only after McKim’s possession conviction was reversed did the government finally decide to prosecute the escape charge. McKim was indicted in November 1973 and tried in January 1974. We held that the one year delay did not deny the defendant’s right to speedy trial where no prejudice was shown. Although we distinguished the “troublesome” fact of the government’s conduct from a “deliberate attempt to delay the trial in order to hamper the defense,” we admonished the prosecutor that “that is the most favorable thing that can be said of it.” Id. at 773. In the case at bar, the government’s conduct was not aimed at delaying in order to hamper the defense; that is perhaps the only favorable thing that might be said of it.

27 . The appellant in Rosson argued that the government did not have a prima facie case against him at the time of his arrest and that the government used an eleven month delay between arrest and trial to make plea bargains with witnesses whose testimony it used to convict him. We stressed that the speedy trial guarantee concerns “the accused’s ability to present the best defense he can muster,” not “whether the delay enables the prosecution to assemble a case.” 441 F.2d at 246.

28 . See note 17, supra.

29 . On December 27, 1973, Avalos’s attorney moved for a continuance of the trial then set for January 7, citing the press of other business and the extensive preparation necessary on behalf of-.Avalos. Counsel had been appointed to represent Avalos on December 6, 1973.

30 . The failure to move for immediate trial following indictment cannot count as heavily against Avalos because his counsel was not appointed until December 1973. See Turner v. Estelle, supra, 515 F.2d at 858 n. 9. Even on appeal, however, both appellants concentrate on pre-accusation delay rather than any prejudice caused by post-accusation delay. At best, therefore, their assertion of the right to speedy trial does not weigh in their favor. At worst, their current arguments suggest that the post-accusation delay was not particularly burdensome to either.

31 . Compare Dickey v. Florida, 398 U.S. 30, 90 S.Ct. 1564, 26 L.Ed.2d 26 (1970) (two witnesses who would have established an alibi for defendant were unavailable due to disappearance and death, and the police officer who recorded the witnesses’ account of the robbery lost his notes).

32 . See note 24, supra. The rationale for dispensing with the prejudice requirement entirely in such cases is that of deterrence: “the prosecution should not be permitted to engage in inexcusable misconduct on the hope that the defendant will not be able to make out a case of prejudice. Where such misconduct has occurred, the state cannot complain that the legitimate interests of its criminal justice system, being pursued in good faith, are being sacrificed because of an honest mistake in a case in which no ultimate harm has been done.” Turner v. Estelle, supra, 515 F.2d at 858.