United States v. Ailemen , 165 F.R.D. 571 ( 1996 )

REPORT AND RECOMMENDATION RE MOTION BY DEFENDANT PIUS AILEMEN FOR PRETRIAL RELEASE ON GROUNDS OF DUE PROCESS

BRAZIL, United States Magistrate Judge.

I. INTRODUCTION

Defendant Pius Alemen was arrested twenty-six months ago on serious drug charges. He has been in jail since, detained without bail. As we explain below, it is virtually certain that at least nine more months will pass before Ailemen’s trial can be commenced and completed. Alemen contends that continued confinement would violate his due process rights.

II. PRELIMINARY FINDINGS OF FACT

A Background

1. This is the second time the government has brought drug charges against Pius Alemen. He was first indicted on drug-dealing counts in 1989.¹ In a 1990 trial on those charges a jury acquitted him of the alleged drug offenses, but convicted him on a passport fraud count. See 3/24/95 Report and Recommendation re Motion by Defendant Ailemen to Dismiss on Grounds of Double Jeopardy (RRDJ) at 2-3.

2. The government resumed investigating Alemen’s activities some time after his acquittal. On July 29, 1993, the government received court authorization to conduct a wiretap to aid its investigation. The government thereafter applied for and received five additional wiretap authorizations. To support each authorization after the first, the government filed affidavits which at least partially relied on wire communications that had been intercepted pursuant to earlier authorizations. See RRDJ at 3-8. The wiretap lasted almost five months and intercepted nearly 10,000 telephone conversations, many of which were in foreign languages. See 4/14/95 letter from Omni Interpreting; 3/20/95 order by Magistrate Judge Wayne Brazil (WB); 12/9/94 affidavit by Gail Shifman (GS).

3. Alemen was arrested in mid-December 1993. See transcript of 12/16/93 detention hearing before Magistrate Judge Patrick Attridge. He was indicted on drug-dealing charges; a July 11, 1994 superseding indictment charged him and fourteen co-defendants with forty-two drug-related counts. See 7/11/94 Superseding Indictment.² Ale-men has been detained without bail since his arrest. See 1/25/94 WB hearing minutes; 12/16/93 detention hearing transcript.

4. It can be reasonably expected that a substantial period will pass between arrest and trial in a complex drug conspiracy case such as this one, because of the large number of defendants charged in the indictment, the large number of charges against the defendants, the volume of evidence in the case, and the presence of a number of complicated legal issues. However, the length of time that has passed since Alemen’s arrest and that is expected to pass before he can be tried has been significantly extended by two main sets of circumstances. First, it is necessary to translate the wiretap tapes before Alemen’s trial can take place and before suppression hearings on most of the wiretap authorizations can be held. Most of the translation work still needs to be done. Second, all the counts against Alemen except one were dismissed by the district court pursuant to a decision that trying Alemen on those counts would violate the Double Jeopardy Clause of the Fifth Amendment. This decision has been appealed by the government — and at the trial court level the government’s ease against Ailemen has been stayed pending resolution of the appeal. Another factor contributing to the length of the pretrial period, albeit less significantly, was slowness by the prosecution in turning over certain discovery.

B. The Wiretap Tapes³

5. Though some parts of the conversations that were wiretapped are in English, most of these communications were made in one or another of four different Nigerian languages. See 3/20/95 WB order; 12/13/94 letter from Loxy Amoni; 12/9/94 GS affidavit. The parties agree that all the wiretaps need to be translated into English and transcribed before this case can be tried. See 2/13/95 order by Judge Vaughn Walker (VW).⁴

6. Counsel for Pius Ailemen conducted a substantial search for persons or entities able and willing to undertake a translation project of this magnitude on a short timetable. That search included submitting samples of taped conversations to potentially interested parties. By late Spring of 1994 Ailemen’s lawyer had secured a bid from a company that appeared competent. Armed with that bid, on May 16, 1994, Ailemen filed a request for authorization to expend Criminal Justice Act (CJA) funds for translation of the wiretap tapes. 12/9/94 GS affidavit.

7. The application that Ailemen initially submitted asked for $334,913 in CJA funds and proposed assigning the translation project to Omni Interpreting, a California-based company. After reviewing the application for CJA funds, the district court forwarded it to the Federal Public Defender for suggestions about ways to reduce the cost of translation. The Federal Defender forwarded the application to the Administrative Office of the United States Courts for suggestions on cost minimization. Ailemen’s counsel worked with the Administrative Office in an attempt to locate lower bids for the translation project. The efforts to reduce the cost of translation failed, and the Administrative Office stated in August that the amount of the original request was appropriate and that CJA funds would be available upon approval of the court.

8. Having received no response to his initial request, on August 9, 1994, Ailemen renewed his request for CJA funds for the translation project. On December 12, 1994, he renewed his request again. The district court authorized the expenditure of the CJA funds on December 27, 1994. The amount the court approved for the translation project was the same amount originally requested, and the project was assigned to Omni Interpreting, the company which was originally proposed.

9. While Omni had a translation team ready to do the project in May 1994, at least some of the members of that team were no longer available in early 1995, when Omni learned that it had been awarded the contract. The fact that Omni had to assemble an at least partly new team of interpreters further delayed commencement of the translation process.

10. On March 16, 1995, this court set a schedule for the translation and transcription of the tapes, dividing the project into six blocks, each corresponding to approximately one sixth of the taped conversations. 3/20/95 WB order. Under current estimates, it appears that about 15,000 minutes of wiretap tapes need to be translated and transcribed.⁵ Translation of the entirety of the tapes was originally expected to take about six months. See 4/14/95 Omni letter; 3/16/95 WB hearing transcript at 62.⁶

11. On May 10, 1995, and again on May 31, 1995, this court ordered translation and transcription of the tapes to be stopped — in substantial measure because of the district court’s rulings on and the government’s subsequent appeal of the double jeopardy issues. 6/1/95 WB order; 5/10/95 WB minute order. To date, Omni has partially translated two of the tape blocks — but has not commenced substantive work on any of the remaining four blocks. Completing the translation and transcription of the tapes is expected to take an additional five months.⁷

C. The Double Jeopardy Issue

12. In September 1994, the Court of Appeals for the Ninth Circuit issued an important and largely unpredicted ruling that appeared to offer many defendants opportunities to challenge the prosecutions against them under the Double Jeopardy Clause of the United States Constitution — United States v. $405,089.23 U.S. Currency, 33 F.3d 1210 (9th Cir.1994), opinion amended and reh’g denied, 56 F.3d 41 (9th Cir.1995), cert. granted sub nom. United States v. Ursery, — U.S. —, 116 S.Ct. 762, 133 L.Ed.2d 707 (1996). $405,089.23 allowed certain defendants to argue that the charges against them should be dismissed on the grounds that they were previously punished for the charged offenses through forfeiture proceedings where they lost property allegedly used in the commission of the offenses. On October 24, 1994, Dele Ailemen, a co-defendant and the defendant’s brother, filed a motion to dismiss the counts against him on grounds of double jeopardy based on $405,089.23. 10/24/94 Motion to Dismiss Indictment with Prejudice for Violation of Double Jeopardy Clause. On February 7, 1995, Pius Ailemen filed a similar motion. 2/7/95 Motion to Dismiss Superseding Indictment Based upon Double Jeopardy.

13. The district court referred these motions, along with all other pending motions, to this court on February 13, 1995. 2/13/95 VW order. On March 24, 1995, this court recommended that Pius Ailemen’s double jeopardy motion be granted and that all counts against him be dismissed. RRDJ at 56. On May 9, 1995, the district court accepted this court’s recommendation in part, dismissing all counts against Ailemen except count 41, a money laundering count. See 5/9/95 VW hearing minutes.

14. On May 24, 1995, the prosecution appealed the district court’s double jeopardy rulings (except, of course, the decision permitting the government to proceed on the money laundering count). 5/24/95 Notice of Appeal. On the same day, the district court stayed (pending disposition of the appeals of the double jeopardy rulings) all trial court proceedings on the one count (money laundering) that remained within its jurisdiction. 5/24/95 VW hearing minutes. Ailemen did not object to this stay. See 5/24/95 VW hearing transcript at 17-18. On June 1, 1995, Ailemen cross-appealed the district court’s decision not to dismiss the money laundering count. 6/1/95 Notice of Appeal.

15. On August 4, 1995, the Court of Appeals for the Ninth Circuit issued its opinion in United States v. Cretacci, 62 F.3d 307 (9th Cir.1995), holding that a defendant could have no double jeopardy claim if he had failed to file a claim in an earlier forfeiture proceeding against property in which he claimed an interest. Pius Ailemen never filed a claim in the proceedings through which the property he claimed was forfeited. RRDJ at 14. Thus, since early August of 1995 it has appeared quite likely that the district court’s decision dismissing most of the charges against Ailemen will be reversed. However, as of this writing, the appeal of the district court’s double jeopardy decision in this case has not been scheduled for oral argument or assigned to a panel.⁸ If oral argument takes place, it will not occur before April 1996. See tape No. CR 96-1 of 1/30/96 WB hearing.⁹

D. Discovery

16. Ailemen alleges that the discovery necessary for him to prepare a suppression motion attacking the initial wiretap authorization was not turned over by the government until July 1994, nearly seven months after his arrest. See 12/19/95 Supplemental Memorandum in Support of Motion for PreTrial Release at 6,18. The government does not dispute this assertion. While it is difficult to determine from the record exactly what discovery was turned over when, and to what extent the government may have improperly delayed turning over discovery, we note the following events.

17. On February 24, 1994, Ailemen orally moved for discovery. On that same date the district court ordered the government to produce a list of the codewords allegedly used by the defendants to refer to illicit substances, as well as a list of pertinent taped phone calls. Minutes of 2/24/94 VW hearing. On April 14, 1994, Ailemen filed several discovery motions. See 4/14/94 Motion for Disclosure of Informant Information; 4/14/94 Motion for Production of Electronically Stored Information; 4/14/94 Motion for Production of Discovery and for Disclosure of Exculpatory Evidence. On May 13,1994, the government filed motions opposing some of Ailemen’s discovery requests. See 5/13/94 Opposition to Motion for Disclosure of Informant Information; 5/13/94 Response re Motion for Production of Discovery, re Motion for Disclosure of Exculpatory Evidence; 5/13/94 Response re Motion for Production of Electronically Stored Information.

18. At a May 26, 1994 hearing, the district court expressed its displeasure with the pace at which the government was turning over discovery and with the government’s failure to turn over certain discovery materials necessary for Ailemen’s wiretap suppression motion. The court suggested that the government was engaging in “gamesmanship.” See 5/26/94 VW hearing transcript at 151-58; see also id. at 159-75. Based on the aforementioned events, it appears that for several months the government resisted turning over some discoverable materials that the defense needed to prepare for the wiretap suppression motion.¹⁰

E. Pretrial Release Motion

19. Ailemen filed a motion for pretrial release on May 23, 1995. 5/23/95 Motion for Pretrial Release. He argued that release was mandated both by the Bail Reform Act of 1984, 18 U.S.C. §§ 3141-56 (1988), and by the Due Process Clause of the Fifth Amendment. See 1/24/95 Memorandum in Support of Motion for Pre-Trial Release.

20. The court’s Pretrial Services Agency prepared a report about Pius Ailemen on June 28, 1995. This report stated that Ailemen was bom in Nigeria, but had lived in the San Francisco Bay Area since 1981. Ailemen travelled to Nigeria once after 1981, in 1988, for three months. Ailemen has a brother and a half-sister who live in Los Angeles. His parents and six of his siblings or half-siblings live in Nigeria. 6/28/95 Pretrial Services Report.

The report also disclosed that Ailemen has never established legal residency in the United States, and that INS proceedings against him have been initiated but put on hold pending resolution of the criminal charges he faces. Not including the instant ease, Ailemen’s criminal history consists solely of his 1990 passport fraud conviction, which arose out of his 1989 arrest, principally on drug charges. His Nigerian passport was seized during the 1989-90 criminal proceedings. Id.

The Pretrial Services report concluded that while Ailemen poses some risk of flight, that risk could be minimized by the imposition of appropriate conditions. The report noted that Ailemen has never been convicted of a violent offense, and that any danger he poses to the community could also be minimized by appropriate conditions. The report recommended that Ailemen be released on a $100,000 personal recognizance bond with certain conditions. Id.

21. On July 28,1995, considering only his motion under the Bail Reform Act, and not ruling on his Due Process Clause arguments, I ordered Ailemen released to a half-way house under strict conditions, including a $300,000 personal recognizance bond cosigned by five employed individuals. See tape No. CR 95-28 of 7/28/95 WB hearing.¹¹

22. The government moved for reconsideration of this order before the district court. 7/31/95 Motion for Reconsideration of Magistrate’s Pretrial Release Order. The district court reversed this court’s release decision, ordering that Ailemen’s pretrial detention continue. Minutes of 7/31/95 VW hearing. However, the district court stated that it would reexamine Ailemen’s pretrial release motion if Ailemen could present a surety or sureties who would post a significant amount of property which would be subject to immediate forfeiture if Ailemen were to violate any pretrial release conditions. Transcript of 7/31/95 VW hearing at 20.

23. On September 21, 1995, and again on December 19,1995, Ailemen filed supplemental memoranda which further developed his argument that due process mandates an end to his pretrial detention. 12/19/95 Supplemental Memorandum in Support of Motion for Pre-Trial Release; 9/21/95 Supplemental Memorandum of Points and Authorities in Support of Motion for Pre-Trial Release.

Last fall, when I expected a ruling very shortly from the Ninth Circuit on the double jeopardy appeals, I indicated that, if pressed, I probably would recommend that the district court deny Ailemen’s due process motion. See tape No. CR 95-50 of 9/22/95 WB hearing. I asked for additional information about which players in the litigation drama were chargeable with responsibility for the passage of so much time before the trial of this matter. Since then, months have passed. I have re-examined the authorities. For reasons set forth at length below, I have concluded that what was a close case in September of 1995, when I thought the Court of Appeals would return this prosecution promptly to the trial track, is no longer a close case. I am recommending that the District Court hold that it would violate Pius Ailemen’s rights under the Due Process Clause to continue his pretrial detention in jail instead of in a half-way house.

The next order of business is to examine in detail the law on which the due process motion must be resolved.

III. LAW

It is clear that long pretrial detentions, at least in some circumstances, can violate the Due Process Clause of the Fifth Amendment of the Constitution of the United States. See, e.g., United States v. Millan, 4 F.3d 1038, 1043 (2d Cir.1993); United States v. Gelfuso, 838 F.2d 358, 359-60 (9th Cir.1988); United States v. Gonzales Claudio, 806 F.2d 334, 339 (2d Cir.1986), cert. dismissed, 479 U.S. 978, 107 S.Ct. 562, 93 L.Ed.2d 568 (1986); United States v. Accetturo, 783 F.2d 382, 388 (3d Cir.1986). What is appreciably less clear, however, is the structure of the analytical process courts should use to determine, in a particular case, whether continued detention of a given defendant would violate that clause. To reduce the likelihood that we will err in our effort to identify that structure, we will begin by examining the sources of the due process limits on pretrial detention.

The Due Process Clause states that “[n]o person shall ... be deprived of life, liberty, or property without due process of law.” U.S. Const, amend. V. The Due Process Clause has long been understood as having both a procedural component, which prohibits deprivations of life, liberty, or property without adequate procedural safeguards, and a substantive component, which prohibits the government from abridging certain rights under any circumstances. See, e.g., Zinermon v. Burch, 494 U.S. 113, 125, 110 S.Ct. 975, 983, 108 L.Ed.2d 100 (1990).

The constitutional ban on excessive pretrial detention can be understood as a right guaranteed by substantive due process. Substantive due process protects “ ‘personal immunities which’ ... are ‘so rooted in the traditions and conscience of our people as to be ranked as fundamental,’ ... or are ‘implicit in the concept of ordered liberty.’” Rochin v. California, 342 U.S. 165, 169, 72 S.Ct. 205, 208, 96 L.Ed. 183 (1952) (quoting Snyder v. Massachusetts, 291 U.S. 97, 105, 54 S.Ct. 330, 332, 78 L.Ed. 674 (1934); Palko v. Connecticut, 302 U.S. 319, 325, 58 S.Ct. 149, 152, 82 L.Ed. 288 (1937)).

Identifying those personal immunities which are “implicit in the concept of ordered liberty” or are “so rooted in the traditions and conscience of our people as to be ranked as fundamental,” is not always an analytically straightforward undertaking. It is clear, however, that such rights are not limited to those that are expressly enumerated in the first eight amendments to the Constitution; were this not the case, the concept of substantive due process would be wholly unnecessary surplusage (as would the Ninth Amendment). See, e.g., Moore v. East Cleveland, 431 U.S. 494, 501-04, 97 S.Ct. 1932, 1936-38, 52 L.Ed.2d 531 (1977).

One way to remain rooted in the text of the Constitution as we attempt to identify those un enumerated rights that the substance of due process protects, or, stated differently, one way to guard against the possibility that identifying such rights could degenerate into undisciplined judicial subjectivity, is to insist on a clear and close connection between enumerated fundamental rights and those unenumerated immunities that also warrant protection directly by the Constitution itself. Thus, immunities that are “implicit in the concept of ordered liberty” may include unenumerated rights that it is necessary to protect in order to assure that one or more of the fundamental rights that are enumerated in the Constitution will remain viable and meaningful. The enumerated fundamental rights that are squarely implicated by Pius Ailemen’s invocation of substantive due process include the Sixth Amendment’s provision that “[i]n all criminal prosecutions, the accused shall enjoy the right to a speedy ... trial” and the Eighth Amendment’s prohibition against “[ejxcessive bail”. Ailemen’s motion also squarely implicates another concept that, while not enumerated in an amendment to the Constitution, nonetheless is “so rooted in the traditions and conscience of our people as to be ranked as fundamental.” That concept, which has served as the cornerstone of our system of criminal justice, is that an accused must be presumed innocent until proven guilty. See, e.g., United States v. Gatto, 750 F.Supp. 664, 672 (D.N.J.1990); United States v. Gallo, 653 F.Supp. 320, 341 (E.D.N.Y.1986).

What we must recognize as we proceed with our analysis in this matter is that these indisputably fundamental rights — the right to a speedy trial, the protection against excessive bail, the presumption of innocence — all would be empty pronouncements, full of sound and fury but signifying nothing, for a defendant who could be imprisoned indefinitely while awaiting trial.

The issue here is not, of course, whether some pretrial detention, by itself, violates substantive due process. The United States Supreme Court put that question to rest in United States v. Salerno, 481 U.S. 739, 746, 107 S.Ct. 2095, 2101, 95 L.Ed.2d 697 (1987), holding that limited pretrial detention does not violate due process so long as it is “regulatory” and “not penal.” In Salerno, the Supreme Court rejected a facial challenge to the constitutionality of a Bail Reform Act provision permitting pretrial detention based on potential dangerousness to society, 18 U.S.C. § 3142, ruling that some detention, so premised, would not, by itself, constitute “impermissible punishment” instead of “permissible regulation.” Id. at 747, 107 S.Ct. at 2101.

Significantly, the Salerno court expressly stated that it was intimating “no view as to the point at which detention in a particular case might become excessively prolonged, and therefore punitive, in relation to Congress’ regulatory goal.” 481 U.S. at 747 n. 4, 107 S.Ct. at 2102 n. 4. Thus the high court acknowledged that there could be a point, in a particular case, where the length of pretrial confinement could become excessive “in relation to Congress’ regulatory goal.” Such excessive confinement would be deemed punitive and would offend due process constraints. Thus Salerno provided an analytical framework for shaping our inquiry at the most general or abstract level. Under this seminal opinion, we are to identify the regulatory goals that pretrial detention can be said to serve (in a particular case), determine how long the pretrial detention will be, then make a judgment about whether the length of that detention is excessive in relation to those regulatory goals.

While the Salerno opinion did not say so explicitly, we infer from subsequent court of appeals decisions that what the courts are called upon to do in these settings is to conduct a balancing analysis in which the following components are on the scales. First, on the government’s side of the scale, the courts (1) identify the regulatory goal or goals that detention of the particular defendant can be said to serve, and (2) assess the magnitude or extent of the contribution toward achieving those goals that the detention in question appears likely to make. Second, on the defendant’s side of the scale, the court determines how long the pretrial detention will last. Finally, the court makes a judgment about whether the length of that detention outweighs (is “excessively prolonged ... in relation to”¹²) the interests that have been identified and measured on the government’s side of the scales.

This sounds fairly straightforward until one tries to do it. The rub, of course, is that the Supreme Court has given us essentially no guidance about how, or through what criteria, to conduct that ultimate balancing-comparison. How do we value or ascribe weight to different lengths of confinement? Does that process vary with or depend at all on characteristics or circumstances of the individual defendant (e.g., should it matter whether the defendant is old, ill, or the kind of person who suffers more in confinement than most others)? More significant, when we compare apples (the government’s regulatory interests) with oranges (the length of pretrial detention), how do we ascribe relative value to each? How do we determine their relative weight? And is the balancing analytically open-minded, or does it start with the scales pre-weighted in some measure (how much?) in favor of the government (whose detention law, Salerno and other courts have held, is clearly supportable, in the abstract, by important governmental interests)?

Because the Salerno opinion did not purport to answer these difficult questions, and because our task involves, centrally, identifying norms that are deeply rooted in our socio-political traditions, we must look for guidance in the many decisions where courts of appeals or district courts have tried to solve these problems. We look to these opinions, in part, as sources of evidence about what “the traditions and conscience of our people” have to say about the relative value of the interests we must compare. In short, we are looking for messages from the heart of our society — messages about intangibles, about values and interests that cannot be measured physically, but only philosophically or politically. And because this is fundamentally a search for our society’s conscience, for truly collective norms, it is improper for me simply to ascribe the weights to the competing interests that I personally think the interests deserve. To be lawful, these assessments cannot be personal; they must be societal.

Before turning to our examination of the pertinent cases, we should pause to characterize at a general level the doctrinal landscape through which we will be travelling. As we will demonstrate below, we believe that it is possible to find, in most of the pertinent eases, a pattern — both of analysis and outcome. It is that pattern, something akin to a center of legal gravity, on which we rely most in forming our recommendations about how to resolve Pius Ailemen’s motion. We think the law compels us to rely on this pattern — because, again, we are directed (in substantive due process analysis) to look for central societal norms.

But we hasten to acknowledge that there are a fair number of reported opinions that do not appear to fit within the patterns we identify. There are some that would suggest different ways of thinking about the ultimate issue before us, and some that would ascribe different weights to the interests here in competition. In other words, we have not been able to generate a coherent system that would reconcile all the judicial pronouncements in these kinds of cases. This is not surprising. We are dealing, after all, with issues and norms that are more resistant than most to definitional precision and to linear reasoning, and that are more susceptible than most to influence by personal values. Despite the undeniable presence of the cases that fail to fit, however, we believe that there is a pattern at the center of this area of doctrine — and it is that pattern at the center on which our analysis turns.

•What are the “regulatory interests” that the cases have identified that could serve as justifications for pretrial detention? The interest that Salerno identified consists of “preventing danger to the community” (protecting society from dangerous persons). Id. at 747, 107 S.Ct. at 2101. Interests that other courts have identified include assuring that the defendant will not flee before trial and preventing the defendant from jeopardizing the trial process through acts such as threats against witnesses. See, e.g., Millan, 4 F.3d at 1043; United States v. Melendez-Carrion (Melendez-Carrion I), 790 F.2d 984, 1002 (2d Cir.1986).

Clearly, each of these interests is important. In the abstract, their importance would deliver considerable weight to the government’s side of the scales. As we explain in detail below, however, it is not simply the abstract significance of these kinds of interests that the courts use when they measure the weight on the government’s side of the scale. Rather, the courts also attempt to analyze, separately for each regulatory goal, how much the continued detention of the particular defendant who is pressing the due process challenge is likely to contribute to achieving the regulatory objective that is under consideration. Since some defendants pose no threat to one or even two of the three acknowledged goals, detaining such defendants can contribute nothing to preserving those unthreatened interests. Other defendants may pose extraordinarily clear and immediate threats to all three regulatory objectives — so that their continued detention would contribute greatly to all of those ends.

It also is extremely important to emphasize here a point that seems to have been overlooked in some of the reported opinions. When courts assess the magnitude of the threat that an individual defendant poses to the government’s regulatory interests, we think that the proper focus is not on how big that threat would be if the defendant were released on np conditions, but, instead, the focus should be is on how big that threat would be if the defendant were released on stringent conditions aimed at reducing as much as possible the likelihood of harm to the threatened regulatory interests. In other words, the issue is not how much threat the defendant would pose if he were as free as any law-abiding citizen, but how much threat he would pose if he were released on the most restrictive available conditions (e.g., 24-hour confinement to a half-way house, electronic monitoring, a high bail guaranteed by people close to the defendant, random searches without probable cause, unannounced substance abuse testing, etc.). It is only by focusing on the actual conditions of release, and what those conditions would contribute to reducing the threat of harm to the government’s regulatory interests, that courts can accurately assess how much continued imprisonment would contribute to achieving the government’s regulatory goals.¹³

Reflecting on these considerations, it is clear that there is a huge range of circumstances that different defendants can present vis-a-vis the government’s regulatory goals. That range of circumstances makes it essential, as most of the authorities repeatedly emphasize, that due process analysis proceed on a case-by-case basis. See, e.g., Gelfuso, 838 F.2d at 359. So there is no bright-line limit, applicable in all settings, on the length of detention pending trial. See, e.g., Gonzales Claudio, 806 F.2d at 340; United States v. Melendez-Carrion (Melendez-Carrion II), 820 F.2d 56, 59 (2d Cir.1987). But see Accetturo, 783 F.2d at 396 (Sloviter, J., dissenting in part) (suggesting that supervisory power of courts be used to establish fixed time limits on pretrial detention).

The weakness of any case-by-case balancing approach, however, is that it can vest excessive discretion in judges and lead to inconsistent decision-making.¹⁴ It is because we want to reduce that risk as much as possible, and because our charge is to find central societal norms, that we have conducts ed the detailed analysis of the relevant authorities that we present below. Along the way, we try to identify which factors carry greater weight under various circumstances.

At the outset, we point out that most courts, when responding to due process motions of this kind, have tended to focus principally on three factors: (1) the non-speculative length of expected confinement; (2) the extent to which the government (the prosecution and/or the court system) bears responsibility for pretrial delay; and (3) the strength of the evidence indicating a risk of flight, a threat to the trial process, and/or a danger to the community. See, e.g., Millan, 4 F.3d at 1038; Melendez-Carrion II, 820 F.2d at 59; Gonzales Claudio, 806 F.2d at 340; Accetturo, 783 F.2d at 388; Gallo, 653 F.Supp. at 338.¹⁵ In addition, some courts, in some circumstances, have compared the length of the likely sentence faced by the defendant to the length of the pretrial detention (in general, the closer the length of pretrial detention gets to the probable sentence, the more likely the courts are to find a due process violation). See United States v. Shareef, 907 F.Supp. 1481, 1484 (D.Kan.1995); United States v. Lofranco, 620 F.Supp. 1324, 1325 (S.D.N.Y.1985), appeal dismissed, 783 F.2d 38 (2d Cir.1986).

There is one court of appeals that has broken away from the balancing approach by focusing on and ascribing (in some circumstances) dispositive weight to only one of the factors described in the preceding paragraph. In United States v. Infelise, 934 F.2d 103, 104-05 (7th Cir.1991), the Seventh Circuit suggested that no defendant, regardless of how long he had been confined, could even begin to mount a due process challenge to his pretrial detention unless he could show, as a threshold requirement, that government action had in fact unnecessarily delayed the trial.

If judge and prosecutor are doing all they reasonably can be expected to do to move the ease along, and the statutory criteria for pretrial detention are satisfied, then we do not think a defendant should be allowed to maintain a constitutional challenge to that detention. To get to first base, therefore, he must show that either the prosecution or the court has unnecessarily delayed in bringing the case to trial.

According to the Seventh Circuit, “the formulation ... suggested explains the cases.” 934 F.2d at 105.

Perhaps because we have the advantage of a substantially larger pool of cases (all those decided since the spring of 1991), we do not share the view that the Seventh Circuit’s “formulation ... explains the cases.” As our forthcoming detailed description of the pertinent authorities makes clear, the analysis that the vast majority of courts conduct reaches well beyond this one factor. Nor do we understand how such a single factor bar to due process claims could be squared with the Supreme Court’s position in Salerno that the focus of the due process inquiry is to be on whether the length of the pretrial detention has become “excessively prolonged ... in relation to Congress’ regulatory goal.” See 481 U.S. at 747 n. 4, 107 S.Ct. at 2102 n. 4 (emphasis added). The high court clearly demands that we engage in an analysis that compares competing considerations, not that isolates one factor and assigns it dispositive power. Even absent a showing that the government unnecessarily delayed getting the case to trial, surely we would be required to at least consider a due process challenge by a defendant whose pretrial detention had lasted for years, whose maximum sentence could include no additional time in custody, and whose detention under the Bail Reform Act was a close question.¹⁶ It is telling that no reported opinion outside the Seventh Circuit has adopted a view that it is impossible, as a matter of law, to mount a due process challenge to the length of pretrial confinement without first showing that the government’s conduct unnecessarily delayed commencement of trial.¹⁷

The one very brief pronouncement on these matters from the Court of Appeals for the Ninth Circuit accepted the balancing, case-specific approach that was then still being developed in the Second Circuit. In Gelfuso, the Court of Appeal for this Circuit declared:

Like the Second Circuit, we find that the due process limit on the length of pretrial detention requires assessment on a case-by-case basis. United States v. Gonzales-Claudio, 806 F.2d at 340. We consider the length of confinement in conjunction with the extent to which the prosecution bears responsibility for the delay that has ensued. Id.

838 F.2d at 359 (emphasis added).

It is clear from the passage just quoted, and from Gonzales Claudio, 806 F.2d at 340, the Second Circuit opinion on which the quoted passage relies, that the Court of Appeals for this Circuit has not concluded that no due process claim can exist unless the government has caused unnecessary delays. I also infer that the Gelfuso court did not intend to imply, in this summary statement, that only the two factors specifically mentioned could be considered in due process analysis. Clearly the Gelfuso panel endorsed the approach taken in Gonzales-Claudio, and in that opinion the Second Circuit expressly endorsed the view that courts also should consider the strength of the evidence justifying the detention decision under the Bail Reform Act. See 806 F.2d at 340. Thus the Seventh Circuit’s “formulation” in Infelise does not accurately represent the law of this Circuit — which, instead, runs with the mainstream by insisting on a real balancing analysis.

We turn now to our detailed review of the pertinent authorities. The pattern we find in them (we concede, again, that there are decisions that do not fit) can be summarized as follows. It appears that the cases fall into three groups, which are separated from one another primarily (but not precisely) by the length of the expected pretrial detention. In the first group, the length of detention already served, or of non-speculative expected detention, is relatively short, between about six months and a year. In this setting, many courts refuse even to entertain due process challenges, concluding that the issue is not ripe. Generally, these courts announce this position without conducting a full analysis of the three main factors used in the balancing approach.

By contrast, when expected detention is of medium length, between just under a year and about two years, courts usually examine, in some detail, all three of the main factors in the balancing approach. In this group of opinions, however, there appears to be some (soft) correlation between the findings about the presence or absence of unnecessary delay that is chargeable to the government and the outcomes of the due process analyses. In many (but not all) of the cases that fall into this middle group, the courts seem inclined to find a violation of the due process clause if they find that the government was responsible for substantial delay, but when the courts find that the government is not chargeable with any significant portion of the delay they are less likely to grant the defendant’s motion.

The third and final group is the most important for our purposes. It consists of cases where the expected pretrial detention is very long, more than two years. In this category, the courts, with one notable exception, have upheld the defendant’s detention only if the government did not bear responsibility for any significant portion of the delay and there were special circumstances indicating that the risk of harm to the government’s interests (preventing flight, danger to the community, or threat to the judicial process) was extraordinarily acute.

We turn now to our detailed examination of the eases in the groupings just described.

Several courts have rejected due process challenges to pretrial detention, without undertaking a full analysis of the three main factors discussed above, on the grounds that the length of detention already served, or the non-speculative length of future detention,’ was too short to make the due process inquiry ripe. For example, in Accetturo, the defendants’ trial was set to start six months after detention began. 783 F.2d at 383, 387. The court ruled that the defendants’ “demand for stronger due process protections for detainees awaiting distant trials is ... better met farther down the procedural road.” Id. at 388.

Likewise, in United States v. Tortora, 922 F.2d 880, 889 (1st Cir.1990), the defendant had been confined for more than half a year as of December 1990, his trial was not expected to start until early 1991, and the trial was expected to last up to eight months. The court held that “[a]t this stage of the proceedings, [the defendant’s] incarceration ha[d] not been so protracted as to support a due process claim.” Id.

See also United States v. Portes, 786 F.2d 758, 768 (7th Cir.1985) (due process challenge held “premature” where defendant had been detained six months at time of decision); United States v. Colombo, 777 F.2d 96, 96-97, 100 (2d Cir.1985) (due process determination “premature” where defendant detained seven months at time of decision and expected to be detained thirteen months to two years before completion of trial); United States v. DiGiacomo, 746 F.Supp. 1176, 1182 (D.Mass.1990) (due process violation “has not yet occurred” in case where defendants already incarcerated for four months and trial not expected to be completed for another year).

Turning to the second group of cases, where expected detention was of medium length, between about nine months and two years, we find courts usually discussing all three of the main factors of the balancing analysis — length of delay, responsibility for unnecessary delay, and flighVdangerousness. In some eases, they have discussed additional factors, such as the relationship of the length .of pretrial detention to the length of the prison sentence the defendant would face if convicted. Though these cases usually do not state that the presence or absence of unnecessary delay (attributable to the prosecution or the judiciary) plays any special analytical role, as we noted, above, there appears to be some correlation between that factor and the outcomes of the cases. This may reflect a feeling in some courts that they can be fairly confident that detention is being pressed primarily to achieve legitimate regulatory goals only as long as it appears that the prosecution is not unnecessarily protracting the pretrial period. The Ninth Circuit’s decision in Gelfuso stands as an example of a case approving expected detention of medium length in the context of finding no governmental responsibility for delay. In that matter the defendants’ trial was scheduled to begin nine or ten months after their detention began. 838 F.2d at 859. The court found that the government was not responsible for any appreciable portion of the pretrial delay, as the government had denied the defendants access to evidence for only a short period while the tapes of intercepted conversations were being copied. Id. The court ruled, without much developed consideration, that in these circumstances due process permitted continued detention. Id.

Another example is United States v. Zannino, 798 F.2d 544 (1st Cir.1986). In that case, trial was delayed primarily because the defendant, an alleged mafia leader, had suffered a heart attack. Time passed while a determination was being made concerning whether he was fit to stand trial. Id. at 545-48. Even though his trial would not begin until at least sixteen months after his detention had begun, the court found that he was “palpably dangerous” to society, noted that the government had done all it could to bring the case to trial expeditiously, and asserted that a defendant should not go free simply because he happened to suffer a physical disability. These considerations led the court to hold that continued detention did not violate the defendant’s rights under the Due Process Clause. Id. at 547-49.¹⁸

See also United States v. Orena, 986 F.2d 628, 630-32 (2d Cir.1993) (continued detention permitted where defendant detained nine months until start of trial, government not responsible for delay, defense counsel declined offer to accelerate trial date, and defendant’s violent mafia activities presented great danger to society); Infelise, 934 F.2d at 105 (continued detention approved where defendants expected to be detained two years until end of trial, as delay was caused by time defense counsel took to prepare defense and neither prosecution nor court system were implicated in delay); United States v. Quartermaine, 913 F.2d 910, 916-18 (11th Cir.1990) (pretrial detention continued where defendants would be detained eight to ten months before start of trial, evidence of risk of flight and danger to society strong, and no mention made of any government action causing delay of trial); United States v. Jackson, 823 F.2d 4, 7-8 (2d Cir.1987) (challenge to pretrial detention denied where detention would last at least eight months until trial, trial postponed at defendant’s request, government did not bear significant responsibility for delay, and “very strong showing of risk of flight” made); United States v. Berrios-Berrios, 791 F.2d 246, 253-54 (2d Cir.1986) (detention allowed to continue where defendant posing risk of flight detained eight months and delay exacerbated by defendants’ many motions), cert. dismissed, 479 U.S. 978, 107 S.Ct. 562, 93 L.Ed.2d 568 (1986); Shareef 907 F.Supp. at 1485-87 (continued detention of two defendants presenting significant risk of flight and substantial danger to community approved, where defendants detained nine months at time of decision, and delay caused by legitimate appeal by prosecution; appeal did not constitute government responsibility for delay); United States v. Gotti, 776 F.Supp. 666, 667, 672 (E.D.N.Y.1991) (due process challenge to detention rejected where mafia defendant posing strong danger to community detained ten months at time of decision and delay not attributable to government).

In another subgroup of cases in this set— where expected detention was of medium length, between about nine months and two years — the courts have concluded that the government was responsible for at least a non-trivial portion of the pretrial delay. In these matters the courts generally (though not uniformly) have ordered the defendants released — finding that to continue to detain them would violate their due process rights. Again, however, en route to these conclusions, the courts usually discuss and assess all of the factors involved in the balancing analysis — and none of these opinions explicit ly confirms the presence or absence of unnecessary delay to be the determinative factor.

One interesting case from this subgroup is United States v. Vastola, 652 F.Supp. 1446 (D.N.J.1987), which was decided a few months before the Supreme Court issued its opinion in Salerno. In Vastola the defendants whose due process motion was the subject of the court’s attention were expected to be in jail eighteen months without a determination of guilt or innocence (the two petitioners were among 21 defendants who were charged in a 117-count indictment). The court found that the multiple causes of the trial being delayed included, among others, the fact that the government had been slow in providing tapes and transcripts of wiretap recordings to defense counsel and had failed to facilitate review of the tapes by providing an index to their contents. Id. at 1448. In part because the government bore responsibility for some of the delay, in part because the court believed that it could impose conditions on the defendants’ release that would reduce appreciably the likelihood that they would pose a danger to the community (there is no indication that the court felt that release would threaten the government’s interests in preventing flight or harm to the judicial process), and “most importantly [because of] the heightened infringement of defendants’ liberty interest resulting from the enlargement of time,” the Court held that continued detention would offend due process. Id. at 1447. While the Vastóla court’s characterization of the analytical role played by the length of detention¹⁹ might have been modified after Salerno, there is no reason to believe that the court would have reached a different outcome, especially given its views about how the threats to the government’s interests could be reduced by imposing conditions on pretrial release.

For other cases in this middle group where courts attended with more than passing analytical interest to the role the government played in delaying the march of the case toward trial, see United States v. Theron, 782 F.2d 1510, 1512, 1516 (10th Cir.1986) (where defendant expected to be detained ten to fourteen months until end of trial, delay caused by complexity of co-defendants’ cases, and court denied defendant’s request for severance and immediate trial, court rules that “serious constitutional questions” would arise unless it construed Speedy Trial Act, 18 U.S.C. §§'3161-74 (1988), as requiring release or trial within thirty days); United States v. Chen, 820 F.Supp. 1205, 1209-11 (N.D.Cal.1992) (court orders release under Bail Act and suggests continued detention would present due process problem, where defendants already detained for one year, future expected detention indefinite but lengthy, delay caused by government appeal of suppression of a videotape which court found “add[ed] little to the government’s case,” defendants posed no serious danger to community, and release conditions sufficiently minimized risk of flight). But see United States v. Majeed, 807 F.Supp. 357, 358-59 (S.D.N.Y.1992) (court permits continued detention where defendant expected to be confined for ten months including trial, even though government responsible for some of delay because it filed superseding indictment and requested extension of time to turn over discovery in order to protect witness identities, as defendant’s intimidation of witnesses posed threat to trial process, but court states that any further delay will lead to revocation of detention order), aff'd, 48 F.3d 1213 (2d Cir.1994).²⁰

There are cases in this subset, however, where courts have ruled that due process required release even without a finding that the prosecution or courts had caused any unnecessary delay. See Shareef, 907 F.Supp. at 1484-85 (due process requires release of defendant detained nine months where case delayed because of necessary appeal by prosecution, defendant likely to serve only eighteen to twenty-four months if convicted, defendant had strong ties to community, and only danger defendant posed to community was economic; two co-defendants facing longer potential sentences and posing stronger risk of flight not released); Gallo, 653 F.Supp. at 343-45 (where defendant who had apparently made threats against witnesses was expected to be detained for eight months before beginning of long trial, court rules due process will not permit detention to continue, finding that while pretrial delay was not attributable to any fault in the government’s motives or conduct, the government nonetheless bore some responsibility for delay because of volume of evidence involved and legal complexity of case); Lofranco, 620 F.Supp. at 1325-26 (focusing mainly on length of delay, court rules continued detention would violate due process where defendant expected to be jailed for nine and one half months without a trial, despite fact that defendant posed risk of flight and danger to trial process and community, and even though no mention made of any government responsibility for delay).

See also Melendez-Carrion I, 790 F.2d 984 (controlling opinion in panel split three ways rules unconstitutional pretrial detention lasting eight months and justified by potential danger to society); United States v. Frisone, 795 F.2d 1, 2 (2d Cir.1986) (under Melendez-Carrion I, pre-trial detention of defendant for twelve months solely on grounds of dangerousness unconstitutional); United States v. Hall, 651 F.Supp. 13, 16 (N.D.N.Y.1985) (release ordered under Bail Act where defendant detained six months with pretrial motions still pending; court states that continued detention would raise serious constitutional questions).²¹

For purposes of deciding the instant motion, the most important group of cases consists of those where expected detention was two years or more. With one notable exception, Millar, 4 F.3d 1038, continued detention has been approved in such cases only where the prosecution or court was not responsible for any significant unnecessary delay and the threat to the government’s regulatory interests that would have been posed by release of the defendant would have been extraordinarily acute, even, presumably, if stringent conditions on the release had been imposed. This pattern is well illustrated by examining three cases which dealt with the same criminal act, a robbery of a Wells Fargo depot by an alleged terrorist group dedicated to Puerto Rican independence — Gonzales Claudio, 806 F.2d 334; Melendez-Carrion II, 820 F.2d 56; and United States v. Ojeda Rios, 846 F.2d 167 (2d Cir.1988).²²

In Gonzales Claudio, the court ruled that due process required the release of some of the defendants charged in the Wells Fargo robbery. 806 F.2d at 343. The expected length of detention through the end of trial was twenty-six months. Id. at 341. The court found that the prosecution bore significant responsibility for the pretrial delay, because it failed to expeditiously provide translation of wiretap tapes, delayed in disclosing videotape evidence to the defense, and had not turned over all discoverable materials. Id. at 342. Thus, even though circumstances such as the defendants’ ties to an alleged terrorist organization created a significant risk that the defendants would flee, continued detention was not permitted. Id. at 343.

In Melendez-Carrion II, on the other hand, the court refused to permit release of two other defendants allegedly involved in the same Wells Fargo robbery. 820 F.2d at 61. The two had been expected to be detained thirty-two months before a trial could be concluded. Id. at 60. After Gonzales Claudio was decided by the court of appeals, the district court in which the case was pending made additional findings of fact. It concluded (in findings the court of appeals later accepted) that the government was not responsible for any significant portion of the delay. Id. The trial court also found that the two defendants who were petitioning for release had a record of prior flight and did not have strong ties to the community, and the court of appeals accepted these findings. Id. at 61. Given the clarity of the threat to the government’s interests and the fact that no delay was chargeable to the government, the court of appeals ruled that continued detention would not violate due process. Id.

Ojeda Rios, 846 F.2d 167, dealt with the last defendant charged in the robbery still detained before trial, one of the two who had failed to win release in Melendez-Carrion II. At this point, trial was not expected to begin until thirty-six months after detention began and was expected to take “many months.” Id. at 169. While the court stated that most of the responsibility for the delay lay with zealous defense counsel, the court noted that the government had been reluctant to agree to a severance for the defendant still detained. Id. Though there did not appear to be any changes in the circumstances bearing on the defendant’s risk of flight, the court ruled that due process required the defendant to be released under strict conditions. Id.

United States v. El-Gabrowny, 35 F.3d 63 (2d Cir.1994), and Gatto, 750 F.Supp. 664, are two other cases demonstrating that once expected detention becomes very long, it is generally approved only if the government has not caused unnecessary delay and there are special circumstances indicating an especially strong risk of flight and/or danger to the community. In El-Gabrowny, the defendants were expected to be detained twenty-seven months before the conclusion of trial. 35 F.3d at 65. The court did not find that the government had unnecessarily delayed the trial, noting that although the government had successfully opposed the defendant’s application for severance, the government’s position that it should not have to try the case twice and expose evidence was reasonable. Id, The defendant in El-Gdbrowny was charged in the 1993 World Trade Center bombing, had assaulted and obstructed police officers when arrested, and five fraudulent foreign passports had been seized from him. Id. at 64. Given the infamous crime the defendant was accused of and the circumstances surrounding it, the court’s finding that the defendant posed a strong risk of flight and a very serious danger to society is easily understandable. See id, at 65. The El-Gabrowny court ruled that continued detention would not violate due process. Id,

As in El-Gabrowny, the defendants in Gatto were highly dangerous — they were charged with murder, other violent acts, and mafia-related racketeering crimes. See 750 F.Supp. at 674. The court in Gatto found that “the government ha[d] played a role in the delay” of the trial, by filing a motion for an anonymous jury days before a scheduled trial date, by failing to provide certain discovery to defendants before the scheduled trial date, and by notifying the defendants on the eve of the scheduled trial of additional evidence which the prosecution had long possessed. Id. at 671. However, the court also found that the defendants had delayed the trial, and stated that the unnecessary delay factor did not weigh in favor of one side or the other. Id. at 675. The court placed significant emphasis on the fact that much of the evidence of the most serious crimes with which the defendants were charged was stale, and the fact that a co-defendant in the case had complied with the conditions of release proposed for the defendants. Id. at 674-76. The court ruled that the expected detention of defendants of twenty-one to thirty months until the end of trial would violate due process and that the defendants had to be released under strict conditions. Id.

The only case upholding expected detention lasting over two years while finding the government to blame for some of the delay is Millan, 4 F.3d 1038. In Millan, the defendants were expected to be confined for thirty to thirty-one months without a determination of guilt or innocence. Id. at 1044. The appellate court accepted a finding that the prosecution was responsible for a significant period of delay: a mistrial had to be declared as a result of the government’s failure to notify the court that some of the agents who had investigated the defendants were themselves under investigation for misconduct. Id. at 1045. The court noted, however, that the error was not intentional, but the product of poor communication within the prosecutor’s office, suggesting that the lack of intentional misconduct reduced the extent to which the delay weighed in favor of the defendants. Id. The court also found that the defendants posed a high risk of flight, because of their travels abroad, their ties to foreign communities, their financial resources, the seriousness of the charges against them, and the strong evidence of their guilt. Id. at 1046.

Perhaps most significantly, the Millan court also emphasized that the defendants would pose an extreme danger to the safety of the community if they were released. See id. at 1048. One of the defendants had been convicted of homicide, and the other had ordered the commission of numerous violent acts and had threatened witnesses and their families. Id. at 1047. Given this constellation of considerations, the court ruled that continued detention was not barred by due process. Id. at 1048.

In one sense Millan might be viewed simply as an outlier case, as it is the only opinion rejecting a due process challenge to detention that was expected to exceed two years when the court also concluded that conduct by the government unnecessarily caused an appreciable portion of the delay. On the other hand, Millan can be understood as conforming, at least loosely, to the broad balancing model we have developed. The court clearly was interested in ascribing relative weight to competing considerations. For example, while acknowledging the government’s responsibility for unnecessary delay, the court seemed to reduce the weight of that factor by suggesting that the source of the error was ‘innocent’ negligence rather than intentional misconduct. Even more telling, the court openly ascribed very great weight to the magnitude of the threat that release would pose to the government’s interest in protecting the safety of the community. In fact, the court went so far as to suggest that “the constitutional limits on a detention period based on dangerousness to the community may be looser than the limits on a detention period based solely on risk of flight.” Id. at 1048 (quoting Orena, 986 F.2d at 631).

In making this suggestion, the Millan court noted that “[i]n the former case [danger to the community], release risks injury to others, while in the latter case [risk of flight], release risks only the loss of a conviction.” Id. (quoting Orena, 986 F.2d at 631). While it is not clear that there are solid doctrinal underpinnings for this notion,²³ it is clear that the Millan court (and the Orena court, which Millan quoted) believed that an important part of the judiciary’s responsibility in conducting this kind of due process analysis was to ascribe relative weight to the legitimate (regulatory) governmental interests that would be threatened by pretrial release — and to let that relative weight (as fixed by the court) have its full play in the balancing of competing interests that ensued.

It also is clear that the Millan court concluded that this factor (threat of harm to the government’s interest in the safety of the community) weighed extremely heavily in favor of continued detention. So Millan can be understood as the product of a balancing analysis where the court of appeal concluded that even though the length of detention was quite substantial, and even though the government was responsible for some unnecessary delay, the weight on the other side of the scale was so great that due process was not offended. While at least some other courts probably would ascribe different weights to the competing considerations, and come to a different conclusion on the ultimate issue, the structure of the analysis employed by the Millan court falls squarely within the predominant mold.

Moreover, even if Millan is out of the mainstream in terms of relative weights ascribed to the factors in the analysis, we think it will be useful, subsequently, to compare the facts in that case to the facts in the instant matter. Because Millan arguably represents one end on a spectrum of relative judicial caution about these matters, we will feel more comfort in our recommendation if it seems supportable even within the kind of framework Millan represents.

IV. APPLICATION OF LAW TO FACTS: FINAL FINDINGS OF FACT AND CONCLUSIONS OF LAW

In this section we will proceed through the kind of balancing analysis that we believe the authorities compel. We will focus primarily on the three factors most discussed in the other cases: (1) the length of the expected delay, (2) the extent to which the prosecution and/or the courts (trial and appellate, in our case) have been responsible for any unnecessary delays, and (3) the magnitude of the risk to the government’s regulatory interests in preventing the defendant’s flight and in protecting the safety of the community and the integrity of the trial process that would be posed if Pius Ailemen were released on the strictest available conditions.

A. Length of Expected Delay

Ailemen has already been detained for twenty-six months — with no trial date in sight. Our task at this juncture is to determine how long he is likely, on a non-speculative basis, to remain in custody before conclusion of his trial. In many cases this would be a relatively straightforward inquiry, but in Mr. Ailemen’s case it is not.

We begin by pointing out that the district court does not have jurisdiction over this matter at this time. Jurisdiction currently lies in the Court of Appeals for the Ninth Circuit, where the district court’s rulings on the double jeopardy motions are under review. We simply do not know how long it will be before the Court of Appeals acts. We have been informed by the clerk’s office at the Court of Appeals that a panel of judges has not yet been selected for this matter and that oral argument has not yet been scheduled. If oral argument is heard, it is extremely unlikely that the appellate court will return this matter to the district court in less than three months.

It is possible, however, that the appellate court will decide the issues on appeal without oral argument. In that event, a ruling could be forthcoming literally at any time. That prospect, however, does not seem likely. If these appeals lent themselves to summary disposition, it is not clear why rulings could not have been made some time ago. The authority on which summary disposition of Pius Ailemen’s appeal seems most likely to be based, Cretacci, 62 F.3d 307, was decided on August 4,1995, more than six months ago. Since the appellate court has not availed itself of the opportunity to dispose of Mr. Ailemen’s appeal summarily in the intervening months, there is an increased likelihood that the Court has decided to resolve all the appeals in this case in the same time frame (in one proceeding). And it is not obvious that the appeals by Mr. Ailemen’s co-defendants lend themselves so readily to summary disposition; they do not appear to be controlled, for example, by Cretacci. See supra note 8. If the Ninth Circuit has decided not to unbundle these appeals, it is likely that even more than three months will pass before jurisdiction over this case is returned to this court.²⁴

It is possible, of course, that the outcome of the appeals could be termination of the case against Mr. Ailemen. But given the ruling in Cretacci, any such assumption would be highly speculative — and, as such, impermissible for purposes of ruling on this motion.

While the double jeopardy rulings are on appeal some parts of pretrial preparation necessarily remain on hold. But others are now being pressed forward, most notably the process of securing reliable translations (from the various Nigerian dialects into English) of the 15,000 minutes of wiretapped conversations. If we assume (as might well not be safe) that the case-development process is not slowed hereafter by the translating, an optimistic judgment about the pace at which the case could be prepared to commence trial would include the following. One month after jurisdiction is returned to this court I will hold the first evidentiary hearing on the defendants’ motions challenging the initial wiretap. Counsel advise that that hearing will consume about five court days (taking evidence). It is highly unlikely that I will be able to complete the process of making findings of fact and formulating recommended conclusions of law within less than three weeks thereafter.

While those findings and recommendations are being reviewed by the district court, I can proceed to the evidentiary hearing on the second wiretap application. That is the first such hearing in which the foreign language conversations are likely to play pivotal roles — but because of the difficulties generated by the extensive use of the foreign tongues, counsel also predict that five court days will be required for this second evidentiary hearing. Again I would be surprised if I could make the required findings and formulate the required recommendations in less than three weeks. Since I have no basis for assuming that any of these challenges to the wiretaps will result in dismissal of all counts against Mr. Ailemen, I must assume that I will be required to hold hearings on the motions the defendants will file attacking the third, fourth, fifth, and sixth wiretap authorizations in this matter. Under the most optimistic (to the government) of scenarios, conducting the evidentiary hearings on those attacks, and preparing the findings and proposed rulings, will consume at least another two months (and likely longer than that; four months may be appreciably more realistic).²⁵

Even if counsel for the defendants and for the government are working diligently on the other necessary aspects of pretrial preparation at the same time the wiretaps are under scrutiny, an additional two months, at least, will be necessary to get the case fully trial ready (after the final rulings on the wiretaps). Then, given the number of defendants (probably eight to ten after things shake out) and the number of counts in the superseding indictment (forty-two), and given the role in the prosecution and defense that the conversations in Nigerian dialects will play, we can be quite confident that the trial itself will consume no fewer than two months.

Under these non-speculative assumptions, all of which appear to be generous to the government and bordering on the naive, Pius Ailemen would remain in custody for an additional nine to ten months even if the Court of Appeals returned this case to the district court this afternoon. On the appreciably more realistic assumption that it takes the Court of Appeals at least three months to return this case to the trial court, it would be at least a year from now before Mr. Ailemen’s trial would be concluded. Since he already has served twenty-six months in custody, I hereby FIND that the non-speculative duration of his pretrial detention will be between thirty-five and thirty-eight months. And there is a very real prospect that this period will exceed forty months, perhaps by a good deal.

No case has upheld pretrial detention expected to last more than thirty-two months against a due process challenge. See supra § III; cf. Melendez-Carrion II, 820 F.2d at 60-61. Since the non-speculative expected length of Ailemen’s detention will be at least thirty-five months, the length of detention prong of the due process analysis weighs strongly in favor of releasing Ailemen.²⁶ It follows that the other two prongs — government responsibility for delay and threats to the government’s regulatory interests (flight/dangerousness) — will have to weigh strongly in favor of the government to justify continued detention.

B. Responsibility for Delay

There are three principal sources of delay in this case that are potentially chargeable to the government: (1) the prosecution’s failure to promptly turn over discovery that the defendants sought, (2) the time consumed by the judicial system in connection with securing reliable translations of the wiretapped conversations, and (3) the time consumed by the judicial system responding to the appeals of the trial court’s double jeopardy rulings.

Before turning to our examination of these sources of delay, we should address one threshold question: is delay that is caused by the courts, or by the judicial system, chargeable against the “government” in due process analysis of the kind we are undertaking here? No case addressing a due process challenge to pretrial detention has discussed in detail how court-caused delay should be treated. In Infelise, the court mentioned that success on a due process challenge to pretrial detention requires the defendant to “show that either the prosecution or the court has unnecessarily delayed in bringing the case to trial.” 934 F.2d at 104-05 (emphasis added). Another helpful case is Barker v. Wingo, 407 U.S. 514, 530, 92 S.Ct. 2182, 2192, 33 L.Ed.2d 101 (1972), where the Supreme Court enunciated a four-factor balancing test for determining whether a defendant’s Sixth Amendment speedy trial rights have been violated. One of these factors was the reason for the delay of the trial, which the Court discussed as follows:

A deliberate attempt to delay the trial in order to hamper the defense should be weighted heavily against the government. A more neutral reason such as negligence or overcrowded courts should be weighted less heavily but nevertheless should be considered since the ultimate responsibility for such circumstances must rest with the government rather than with the defendant.

Id. at 531, 92 S.Ct. at 2192 (emphasis added, footnote omitted).

Seeing no obvious reason why the Supreme Court’s treatment of court-caused delay in the speedy trial arena should not be at least instructive to courts considering due process challenges to pretrial detention, I will proceed on the assumption that unnecessary delays caused by the courts can weigh in the defendant’s favor in this analysis. But before focusing on court-caused delays, we will examine the one source of wasted pretrial time that appears chargeable, on the facts of this case, to the prosecution.

As noted earlier, it appears that, without substantial justification, the prosecution delayed for three to five months in producing certain discoverable information to the defendants. See supra Preliminary Findings of Fact ¶¶ 16-18. The prosecution has not challenged here Ailemen’s contention that his counsel needed this material to prepare his challenge to the first wiretap authorization. If that material had been turned over promptly, the evidentiary hearing and the ruling on the challenge to the first wiretap could have been completed by some time in the spring or early summer of 1994 — well before $105,089.23, 33 F.3d 1210, the Ninth Circuit decision that inspired the double jeopardy challenges to the entire case. Had that occurred, this case would be at least two months closer to being ready for trial than it is now.

The prosecution has attempted to blame the pretrial delay on the fact that the defendants have filed many pretrial motions. However, this does not excuse the government’s slowness in turning over discovery. Criminal defendants have the right to zealously defend themselves, and some of the motions the defendants filed in this case resulted in significant rulings in their favor. Moreover, the government’s duties to provide discovery are independent — they do not depend on the litigation behavior of defense counsel. In addition, the government has failed to make a persuasive showing that producing at least much of the requested discovery was anything but a straightforward task — requiring no extraordinary commitment of resources.

We also note that there is substantial authority for the view that the government’s obligation to expedite pretrial proceedings grows when the government has moved for detention pending trial. See, e.g., United States v. Salerno, 794 F.2d 64, 79 n. 2 (2d Cir.1986) (Feinberg, C.J., dissenting on other grounds) (“Pretrial detention under any circumstances is an extraordinary remedy with serious due process implications. If the government seeks the benefit of that remedy it must exert itself to accelerate the date for a trial.”), rev’d, 481 U.S. 739, 107 S.Ct. 2095, 95 L.Ed.2d 697 (1987);²⁷ Jackson, 823 F.2d at 8 (“when the government moves for pretrial detention it has an obligation to arrange for the trial as quickly as possible, using ‘extraordinary means’ if necessary”) (quoting Salerno, 794 F.2d at 79 n. 2); Gonzales Claudio, 806 F.2d at 342 n. 5 (“ ‘In cases involving many hours of taped intercepted material, the government may have to arrange for swift and reliable transcription, by extraordinary means if necessary, before moving for detention or immediately after obtaining it.’ ”) (quoting Salerno, 794 F.2d at 79 n. 2). Seeing no persuasive evidence that the government has taken special steps to expedite the pretrial process in this litigation, I see no reason not to conclude that at least two months of Ailemen’s pretrial detention is chargeable to delay unnecessarily caused by the prosecution’s handling of its discovery obligations.

More significant delays are attributable, however, to the way the judicial system handled the request by the defense for funds to support translating the wiretapped conversations. After undertaking a substantial search for candidates for the job, and after securing bid information, the defense filed a formal request for the necessary CJA funds in mid-May of 1994. Had that request been granted within a month or so, instead of at the end of that year, the teams of translators that had already been assembled could have proceeded immediately with their work, which they estimated they could have completed in segments (one wiretap at a time, seriatim) over a five to six month period.²⁸ With hard work, the defense lawyers could have prepared their challenges to the wiretaps along a parallel schedule, so that the hearings challenging the five authorizations that relied in part on translations could have been completed by the end of 1994 or early 1995.

Under such a schedule, this important part of the pretrial process could have been completed before the district court made rulings on the double jeopardy motions (Pius Ailemen’s motion to dismiss on grounds of double jeopardy was not even filed until February 1995). And if the challenges to all the wiretaps had been concluded before the trial court lost jurisdiction over most of the case as a result of the double jeopardy rulings, this matter would be closer to trial now by at least five to six months.

There were good reasons for the district court to raise questions about the proposed translation plan and its hefty price tag. Raising questions, and seeking assurances that no less-costly alternative was viable, were responsible steps, given the magnitude of the request. In the end, however, it was determined that the request as originally presented should be approved. That conclusion in essence vindicated the homework that defense counsel had done before making the original proposal. Thus, after all was said and done, the time consumed by the judicial system as it probed the original proposal, pondered the cost to society of the undertaking, and pressed for what turned out (unforeseeably) to be non-existent cheaper options, turned out to be wasted. This activity, which did not proceed at a rocket’s pace, unfortunately made no contribution to the disposition of the litigation. Because, in these senses, this delay was “unnecessary” (even though inspired by respect-worthy motives) it supports, perhaps without great weight by itself,²⁹ defendant’s contention that his continued detention would offend due process norms.

While the delay in translating the tapes is chargeable to the government (the judicial system), we cannot recommend the same conclusion about the prosecution’s decision to appeal the district court’s double-jeopardy-based dismissal of most of the charges against Pius Ailemen. The appeal by the prosecution was clearly necessary for its case — without it, the government would only be able to prosecute Ailemen on the money laundering count. And the appeal is far from frivolous — as a result of Cretacci, 62 F.3d 307, it is likely that the government will prevail. Under these circumstances, the prosecution’s decision to appeal cannot be charged against the government. Cf. Shareef, 907 F.Supp. at 1485 (refusing to charge to the government the delay caused by its legitimate appeal of trial court rulings).³⁰

On the other hand, the amount of time that it is taking the Ninth Circuit to decide the appeal of the double jeopardy rulings may weigh against the government to some extent. Ailemen’s case appears to be squarely controlled by Cretacci, 62 F.3d 307. And Cretacci’s holding — that an administrative forfeiture proceeding does not impose punishment for double jeopardy purposes if the defendant made no claim to the forfeited property — has been squarely endorsed by the Second, Third, Fifth, Sixth, and Seventh Circuits. See United States v. Idowu, 74 F.3d 387 (2d Cir.1996). No federal appellate decision has challenged or criticized Cretacci. Nonetheless, nine months have passed since the government filed its appeal, and a ruling from the Ninth Circuit does not appear imminent. See supra Preliminary Findings of Fact ¶¶ 14-15.

The Court of Appeals for the Ninth Circuit has a crowded docket, and there is no reason to believe that the judges of that court are not doing everything reasonably possible to resolve, essentially in the order filed, the many cases that are pending before them. Four considerations, however, make it arguable that some of the time being consumed while jurisdiction remains in the appellate court represents “unnecessary delay.” First, the appeals in this matter do not follow judgments after trial on the merits. Instead, they are, for all practical purposes, interlocutory — and while they remain pending, the progress of this litigation toward trial remains stalled. Second, this is a criminal case and is entitled to priority over most civil matters. Third, the defendant who presses this motion remains in pretrial custody while the appeals are under consideration — and he has been in custody for more than two years without a determination of guilt. Fourth, on August 4, 1995, almost 7 months ago, the Ninth Circuit resolved (in favor of the government) what appears to be the crucial issue on which the government’s appeal of the trial court rulings in favor of Pius Ailemen turns. See Cretacci, 62 F.3d 307. Given the ruling in Cretacci, and the fact that identical outcomes have been reached by all other federal appellate courts that have addressed the same issues, it is not at all clear why Pius Ailemen’s appeal could not be dispatched summarily.

It is possible, as we speculated earlier, that one reason the Court of Appeals has not ruled on the government’s challenges to the district court’s orders that relate to Pius Ailemen is that the appellate court is not addressing the matters that relate to him in isolation, but, as one might expect in the normal course, as part of one package in which it will resolve all the double jeopardy issues that have arisen in this one case, with respect to all the defendants. Since there are issues on appeal with respect to two of Pius Ailemen’s co-defendants that may well not be controlled by Cretacci, handling all these matters together would slow appellate disposition. See supra note 8.

The rub of course, is that Pius Ailemen is not a group, but an individual. The appeals run to him as an individual. And, most significantly for current purposes, he is the only individual in this case who remains in pretrial detention. Moreover, he is, by the government’s repeated averments, clearly the lead defendant in this prosecution. Thus, strong arguments can be made that, if it would expedite matters, it is appropriate to sever appellate consideration of the government’s appeal of the double jeopardy rulings that relate to Pius Ailemen from the double jeopardy rulings that relate to the two co-defendants.

I do not mean to suggest that sole or even primary responsibility for severing the appeal of the rulings about Pius Ailemen from the appeals of the rulings about his co-defendants rests with the Circuit Court. It may be that responsibility for suggesting such a severance rests more properly with the prosecution. For current purposes, it really doesn’t matter whether the prosecution or the courts should take this initiative or bear responsibility for the passage of the time since Cretacci was decided. What matters is that it is difficult to understand how one could characterize as “necessary” (to resolve the appellate issues with respect to Pius Ailemen) all the time that has passed since that apparently dispositive ruling was issued.

It also may be pertinent here to point out that the government already has severed its prosecution of two of the defendants who were named in the superseding indictment in this case — and the trials of those two defendants were completed some time ago.

Given all of these considerations, I cannot conclude that all of the nine months that have elapsed while the government’s appeal of the double jeopardy rulings has been pending represent necessary delay. Especially in light of Cretacci, decided almost seven months ago, at least three or four months of the time Pius Ailemen’s case has been stalled in the appellate court must be deemed “unnecessary.”

This is a finding, of course, not about culpability, but about whether, for purposes of this due process analysis, the extra time should be charged to the government. Like the delay that followed the request to fund the translations of the wiretaps, this is a “more neutral reason” for the passage of time and, for that reason, “should be weighted less heavily.” Cf. Barker, 407 U.S. at 531, 92 S.Ct. at 2192. But because the time charged here (the extra three to four months) cannot be deemed “necessary,” we cannot ignore it — “the ultimate responsibility for such circumstances must rest with the government rather than with the defendant.” Cf id.

In combination, the three sources of unnecessary delay that are attributable to the government (both the judicial system and the prosecution) have resulted in adding somewhere between eight and twelve months to Pius Ailemen’s pretrial detention.³¹ Even without any taint of culpable motives or conduct, that is a significant period. As such, it must add appreciable weight to the side of the scales that favors release on strict conditions. Because that side of the scales was already heavy (the expected length of pretrial detention being at least thirty-five to thirty-eight months), I will be compelled to find that continued detention would violate Ailemen’s rights under the due process clause unless the threat to the government’s competing regulatory interests that would be posed by his release on the strictest available conditions is extraordinarily severe.

C. The Magnitude of the Threat to the Government’s Regulatory Interests

As discussed earlier, there are three regulatory interests that pretrial detention could serve: (1) protecting the integrity of the trial process (by, for example, preventing a defendant from attempting to intimidate witnesses, jurors, or others involved in the prosecution), (2) preventing danger to the community, and (3) assuring that the defendant is present for trial and, if convicted, for sentencing (the risk of flight factor).

Ailemen’s continued detention cannot be supported by a potential threat to the judicial process, as the government has presented no evidence that Ailemen has threatened witnesses or that he intends to take any other action that could jeopardize the integrity of his trial.

Analysis of the magnitude of the threat to the government’s interest in preventing danger to the community is more complicated. We begin with a clear recognition that trafficking in illegal narcotics (heroin and cocaine, as here alleged), without more, causes serious harm to society. A drug trafficker is, by legal definition, a danger to the community. See 18 U.S.C. § 3142(e). Drug trafficking also creates additional real dangers— through crimes committed by addicts seeking to support their habits, through plain human suffering, and because people engaged in the sale of illegal substances sometimes commit or direct violent crimes in furtherance of their enterprises. In the case at bar, for example, the government points out that Pius Ailemen “was stabbed during a narcotics transaction” and that one of his assailants “had a gun at the time of his arrest.” 7/26/95 Response and Opposition to Motion for PreTrial Release at 5.

There appears to be substantial evidence supporting the government’s indictment of Pius Ailemen as a serious and sustained drug trafficker. In that very real sense, there is substantial evidence that he would pose a danger to the community if he were released without meaningful constraints.

The government has presented no evidence, however, that Ailemen has ever committed or attempted to commit acts of physical violence. While he apparently was the victim of violence in late 1993, the government has not claimed that he initiated that incident, or that in connection with it he had been carrying a weapon of any kind. Nor has the government offered evidence that he has used weapons in the past, or that he has directed crimes of violence by others. His only conviction is for passport fraud. In these senses, Pius Ailemen’s background is quite different, and in some material ways less threatening to the safety of society, than the backgrounds of many of the defendants who have been deemed extreme threats to society by other courts. Cf. El-Gabrowny, 35 F.3d at 64-65 (evidence indicated defendant was dangerous terrorist involved in World Trade Center bombing); Millan, 4 F.3d at 1047 (one defendant previously convicted of homicide, other defendant had threatened witnesses and their families and had ordered the commission of numerous violent acts); Melendez-Carrion I, 820 F.2d at 61 (evidence indicated defendants were leaders of terrorist group).

It also is relevant here that the government has not presented evidence that Ailemen himself uses drugs. As far as we are aware, he has no narcotics habit for the support of which he might be driven to commit other crimes. Thus it appears that the threat to the safety of the community that Ailemen poses is confined to the risk that he would traffic in drugs if he were released.

The critical point to be made in this prong of our analysis is that conditions can be imposed on Ailemen’s release that would dramatically reduce the likelihood that he would pose a danger to the community by trafficking in drugs or in any other way. He can be ordered to remain, twenty-four hours a day, in a highly regimented, well-supervised halfway house, an institution that has served this court for many years and that has deep experience in dealing with serious offenders. In the half-way house, Ailemen’s behavior can be monitored. His communication with persons outside the half-way house can be severely restricted. The court can order, for example, that he be permitted to make only a limited number of phone calls per day, and that each call (except those with his attorney) be monitored. The court also can order that his mail be inspected for weapons, drugs, or other contraband. In addition, the court can limit to a specified list the persons who may visit him, and can prohibit him from communicating with or seeing any of his co-defendants except in the presence of counsel. Within the half-way house, there can be strict limits on the items he may possess, and his room, his belongings, and his person can be subject to search at any time, on no notice and without cause. And he can be subjected to random testing for use of drugs or alcohol.

Under such conditions, it would be difficult, in the extreme, for Pius Ailemen to traffic in drugs or to orchestrate such trafficking by others. Such measures would reduce to a very low level the threat to the government’s interest in preventing danger to the community that otherwise might be posed if this defendant were not in custody. So reduced, this threat cannot weigh heavily in favor of continued detention.³²

In my view, the governmental interest most threatened by releasing Ailemen from custody is the interest in assuring his presence for trial and, if convicted, for sentencing. The half-way house is not as locked-down and not as secure as a jail or prison. A clever person (as Ailemen may be) with the will to escape from the half-way house very likely could. So while commitment to the half-way house on the strict conditions described above would reduce in some measure the risk of flight, it would not eliminate that risk. For that reason, we must examine the many factors and considerations that bear on the likelihood that Ailemen would attempt to flee from the half-way house. Only after such an examination will we be positioned to make a properly informed judgment about how much weight to ascribe to this factor in our due process analysis.

The pressure on Ailemen to try to flee will be considerable. If the Court of Appeals reinstates the drug trafficking charges against him, as seems likely, he faces possible penalties that could include life in prison.³³

How he views the likelihood that he will be convicted, however, is more difficult to assess. He was prosecuted by the same authorities on serious drug charges several years ago — but succeeded in gaining an acquittal.³⁴ So the pendency of charges, per se, may not affect him as much as it might others. And while there appears to be substantial evidence in support of at least some of the drug counts in the superseding indictment, proof of a great many of the counts, certainly the more ambitious and threatening (to Ailemen) of them, appears to turn in large measure on whether the prosecution can persuade the jury, beyond a reasonable doubt, that alleged codewords and phrases, recorded (sometimes imperfectly) from wiretapped conversations that were conducted largely in esoteric languages, clearly had culpable meanings. It appears to me that Ailemen understands (may even overestimate) the difficulties of proof presented by this evidence. Ailemen also has demonstrated a remarkable confidence and composure throughout the long pretrial period — possibly rooted in optimism that the government will not be able to overcome all of the many hurdles it will face when it tries to meet the demanding burden of proof that it must bear. These considerations are relevant because the incentives and pressures he will feel to flee are likely to vary, in part, with the level of confidence he feels about his chances of being acquitted, at least on the charges that carry the most severe sentences.

As the government points out, there are several additional considerations, some substantial, that support its contention that the risk of Pius Ailemen fleeing is high. It is not clear that he has ever had a lawful job in this country; certainly no employer has come forward to press for his return to work. While he has claimed at times in the past to have supported himself through an importing business, United States Customs has no record of defendant as an importer. And in a detention hearing in the District of Columbia after his arrest in December of 1993 he apparently stated that he was dependent for support on income sent to this country by relatives in Nigeria. He owns no property in the United States.

Ailemen also has demonstrated something less than rigid respect for his obligations to the court. He apparently does not deny, for example, that he made at least eleven unauthorized trips to Washington, D.C., while he was on probation as a result of his 1990 passport fraud conviction. The drug and money laundering activities of which he is charged also allegedly occurred while he was in that same probationary status.

The fact that the one conviction Ailemen has suffered is for passport fraud also supports the view that he represents a serious flight risk. Moreover, the government has presented evidence that he has used many aliases in the past and has travelled (or at least purchased tickets to travel) under false names. See 9/29/93 Affidavit of Special Agent Robert J. Silano. He is sophisticated and probably could muster the resources to pay for transportation overseas. On the other hand, Ailemen’s Nigerian passport was seized during the 1989-90 proceedings, and there is no evidence that he currently possesses any travel documents.

It also is significant that Pius Ailemen never has become a legal resident of the United States, and that the INS has initiated proceedings that might well result in his deportation. Those proceedings have been stayed, however, pending disposition of the current charges — and, as with the matters covered in the indictment against him, Ailemen projects an air of confidence about his prospects for success in that arena.

The government also points out that Ailemen has parents and siblings in Nigeria, where he was raised and clearly could fit in. Thus he has a place to go (unlike some defendants who appear before me). Again, however, it is not clear how strong the pull toward Nigeria would be. Ailemen has resided in the Bay Area continuously since 1981, with the exception of one three-month trip to Nigeria in 1988. And he has two siblings in California and a considerable number of friends in the Bay Area (a good many of whom have written to the court on his behalf). Some evidence of emotional ties here is reflected in the fact that five of his friends have signed a $300,000 personal recognizance bond to guarantee his court appearances and his compliance with strict conditions of pretrial release, including, of course, committing no additional crimes. In short, with the passage of time in this country, it appears (as his attorney has suggested) that Ailemen has come to prefer, actively, life and its promises here much more than life and its promises in Nigeria.

Ailemen asserts that he made no effort to avoid arrest in December 1993 even though he expected to be arrested; the government does not dispute this assertion. See 12/19/95 Supplemental Memorandum in Support of Motion for Pretrial Release, Appendix at 2; 7/24/95 Memorandum of Points and Authorities in Support of Motion for Pre-Trial Release at 5 n. 3. In fact, it appears that after he was injured (in December of 1993) in connection with the alleged drug transaction in Washington, D.C., he cheeked himself into a local hospital, under his own name, even though he had reason to believe that his arrest was imminent. Defense counsel argues that had he been seriously inclined to flee, he would have done so when he had the chance. Of course, it is not clear whether flight was a medically available option at that point — and Ailemen could not have known what specific charges he would end up facing (and how severe the penalties for them might be). There is a counterpoint here as well, however, as it appears (from statements attributed to the defendant in a part of the wiretapped conversations) that for a substantial period before he was arrested, Ailemen knew, or strongly suspected, that government agents were tapping his phones and actively investigating his activities. Despite that apparent knowledge, he did not flee.

Ailemen also has no known history of substance abuse, so there appears to be no threat that drugs or alcohol might impair his judgment if he were outside a custodial setting.

Finally, defense counsel points out that all of Ailemen’s co-defendants who are still awaiting trial in this case, including his brother Dele, have been released, and that the government has not presented any evidence that any of them have fled or violated their release conditions. See 12/19/95 Supplemental Memorandum in Support of Motion for Pretrial Release at 21-22, uncontradicted by prosecution; cf. Gatto, 750 F.Supp. at 675-76 (co-defendants’ compliance with release conditions weighs in favor of releasing defendant on similar conditions).

Having assessed all of the considerations described in the preceding paragraphs, I conclude that while release of Pius Ailemen would create a real risk of flight, the size of that risk can be reduced meaningfully through the imposition of a host of conditions. In addition to the twenty-four-hour-a-day commitment to the half-way house described earlier, and the other conditions there articulated (limited and monitored use of means of communication, restrictions on visitors, etc.), the court could insist, as I did earlier, that at least five persons co-sign a large bond ($300,000 is suggested) guaranteeing Ailemen’s compliance with all the conditions of his pretrial release. Further, the court could insist that these persons be in this country lawfully, have something meaningful to lose (e.g., garnished wages) if the defendant fails to honor his bond commitments, be emotionally connected to the defendant, and have no criminal record. The knowledge that fleeing would expose such people to severe economic consequences³⁵ would help counteract the pressure and temptations Ailemen undoubtedly will experience.

Several months ago I concluded that the stringent conditions just described would, in combination, reduce the risk of flight to a level that would satisfy the requirements of the Bail Reform Act, meaning that they would serve to “reasonably assure the appearance of the person as required.” See 18 U.S.C. § 3142(g). I freely concede that this was, and remains, a close question. That fact follows in. part from the nature of the judgment being made under the Act. The release/detention decision under the Bail Act is hardly an exact science. It is appreciably more inexact than trying to determine what events occurred in the past (as we do in trials). For here we are trying to make judgments about what specific individual persons, whom we certainly cannot claim to fully understand, will do in the future — how they will respond to circumstances and feelings that we have only quite limited powers to foresee. Charged with a duty by Congress, we do our best, but we know that the margins of error are ample.

The district court disagreed with my judgment about application of the Bail Reform Act to Pius Ailemen, concluding that he could not be released under the Act unless, at least, substantial property was posted on his behalf. I infer, from the district court’s willingness to reconsider releasing Ailemen under the Bail Act if sureties with sufficient property came forward, see supra Preliminary Findings of Fact ¶ 22, that the district judge did not conclude that this was one of the few extreme cases where it was obvious that no combination of conditions ever could be presented that would satisfy the Act’s requirements.

We also know that the range of defendants whom we would put on the detention side of the Bail Reform Act line is wide indeed: some belong there so clearly that application of the Act is essentially ministerial, while others come much closer to meeting requirements for release, at least on restrictive conditions. In other words, even if we confine the universe of defendants we are considering to those who cannot meet the requirements for release under the Act, the magnitudes of the threats to governmental interests that are posed by the defendants in this pool covers a very wide range.

That fact of life is one of the reasons why we cannot equate analysis under the Bail Reform Act with analysis under the Due Process Clause. The other reason, of course, is that the authorities who have construed the substance of due process in this setting require us to consider additional factors, including the length of the pretrial detention and the degree to which the government is chargeable with unnecessary delay in reaching a determination of guilt or innocence.

In the case at bar, I conclude that releasing Pius Ailemen on the restrictive conditions I recommend³⁶ would pose only moderate threats to the implicated governmental interests. In other words, I find that Ailemen’s case falls appreciably closer to the end of the spectrum nearest the line established by the Act than to the other end of that risk spectrum. This is not a case such as Melendez-Carrion II, where the defendants had a record of prior flight and the evidence indicated that they were leaders of a terrorist group. See 820 F.2d at 61. Nor is this case like Millan, where one defendant had been convicted of homicide, and the other defendant had threatened witnesses and their families and had ordered the commission of numerous violent acts. See 4 F.3d at 1047. And this case is unlike El-Gabrowny, where the defendant had resisted arrest and was charged with complicity in a terrorist act of immense proportions that had shocked America. See 35 F.3d at 64-65.

I further find that the magnitude of the risks to the government’s regulatory interests that Ailemen’s release would pose, on the conditions I would require, clearly is not sufficient to outweigh the considerations that cut decisively in his favor in the balancing analysis that we are commanded to perform under the Due Process Clause. The two competing interests are simply too heavy: (1) the extremely long period (thirty-five to thirty-eight months, at least) that he will spend in custody before his trial will be completed, and (2) the unnecessary extension of the pretrial (and up-to-now custodial) period of between eight and twelve months that is chargeable to the judicial system and the prosecution.

It is important at this .final juncture to remind ourselves that no case has concluded that expected pretrial detention lasting longer than thirty-two months would be consistent with due process. We also bear in mind that only one case (Millan, 4 F.3d 1038) has approved expected pretrial detention exceeding two years in a context where the government bore responsibility for some appreciable portion of the pretrial delay.

V. RECOMMENDATIONS

Given the norms that I have found in the cases, and my application of the balancing test to the specific facts of this matter, I RECOMMEND that the district court find that continued pretrial detention of Pius Ailemen would violate his rights under the Due Process Clause. I FURTHER RECOMMEND that he be released to a half-way house on the following special conditions:³⁷

1. At least five individuals who are employed or are willing to post a significant amount of property, who have no criminal record, who are lawfully in this country, and who are emotionally connected to the defendant must sign a $300,000 personal recognizance bond on Ailemen’s behalf.

2. Ailemen must reside at ECI, a halfway house in San Francisco.

3. Ailemen may not leave ECI for any reason other than to make court appearances; when he leaves for that purpose he must be escorted to and from court by ECI staff or by his attorney.

4. With the exception of calls to his attorney, Ailemen may not make more than three telephone calls per day.

5. Every telephone call he makes to anyone other than his attorney must be monitored by ECI staff.

6. With the exception of his brother, Dele Ailemen, Pius Ailemen may not communicate in any way, directly or indirectly, with any of his co-defendants, except in the presence of counsel.

7. Ailemen’s mail (incoming and outgoing) must be inspected by ECI staff for contraband.

8. Ailemen may not take any steps to acquire any travel documents.

9. Ailemen may not possess or use alcohol, controlled substances, firearms, or any other weapons or destructive devices.

10. Ailemen’s room, belongings and person shall be subject to search at any time, on no notice, and without cause.

11. Ailemen will be subject to random testing for drugs or alcohol, without notice and without cause.

12. Ailemen must comply with all ECI rules not already covered by the foregoing conditions.

IT IS SO REPORTED AND RECOMMENDED.

1 . The 1989 drug charges against Ailemen included conspiring to import heroin, conspiring to possess heroin with intent to distribute it, distributing heroin, and procuring interstate travel in furtherance of a business enterprise to unlawfully import and distribute heroin. RRDJ at 2.

2 . Pius Ailemen was charged in 39 of the 42 counts. The charges against him included conspiracy to distribute heroin and cocaine, engaging in a continuing criminal enterprise, distribution of heroin, use of the telephone to facilitate drug trafficking, travel in interstate commerce to facilitate drug trafficking, and money laundering. See 7/11/94 Superseding Indictment.

3 . Some documents concerning translation of the wiretap tapes are under seal. For this reason, I give either no record citations or incomplete record citations to certain statements about the translation process.

4 . One of the reasons that the tapes need to be translated in their entirety is that Ailemen plans to file motions to suppress the evidence gathered pursuant to each wiretap authorization. Each such authorization (after the first) was at least partially based on evidence gathered pursuant to previously authorized portions of the wiretap. Thus, in order for each suppression motion to be litigated, the wiretap evidence gathered before the authorization at issue in the suppression motion must be translated and transcribed. See 3/16/95 WB hearing transcript at 20-22.

In addition, the government alleges that the defendants spoke about drugs using codewords in many of the taped conversations. The defense argues that translating the entirety of the tapes is necessary to reveal conversations which could show that the alleged codewords were not references to illicit substances. See 12/12/94 affidavit of J. Frank McCabe.

5 . For reasons not entirely clear, the original estimate of the number of minutes that had to be translated was much higher — almost 29,000 minutes. See 4/14/95 Omni letter.

6 . The difficulty of translating the tapes is exacerbated by several factors. The nature of this case — especially the government's allegation that the defendants discussed drug transactions using codewords — requires that the translations be of the utmost accuracy. Some of the languages used on the tapes (and in which the codewords allegedly appeared) are among the most complicated Nigerian languages. 12/12/94 Omni letter. And the Nigerian languages involved are sufficiently different that in many instances different translators must be used for the separate languages. See 3/16/95 WB hearing transcript at 25.

7 . Omni’s estimate of the total amount of time required to translate the tapes has not gone down even though its estimate of the number of minutes on the tapes has. Apparently, the translation work has turned out to be considerably more difficult than Omni originally expected. Moreover, the passage of additional time between May of 1995, when the court ordered Omni to stop work on the project, and the beginning of this year, when work resumed in earnest, appears to have forced Omni to make some additional changes in the composition of the translation team. Yet another source of complication, and potential delay, is the apparent fact that many of the translators on whom Omni is forced to rely have substantial other work commitments — thus reducing the time they can devote to this one project.

Work on the translations also has been hampered by personnel and administrative difficulties within Omni. Even now the court cannot be fully confident that Omni will be positioned to complete this demanding task. If we are forced to search for another entity to complete the translation of the tapes, Ailemen’s trial is likdly to be delayed appreciably longer than the projections made later in this Report and Recommendation. See infra § 4.

8 . The time it may take the Ninth Circuit to decide the appeal of Ailemen’s case may be significantly lengthened by the fact that also at issue in the appeal are the double jeopardy based dismissals of charges against two of Ailemen’s co-defendants. See 5/24/95 Notice of Appeal. This court has found that one of these co-defendants made a claim to the forfeited property, and that while the other did not make a claim, he also did not receive adequate notice of the forfeiture proceeding. 4/13/95 Report and Recommendation re Defendant Robert A. Tinney’s Motion to Dismiss on Double Jeopardy Grounds at 8-9; 4/13/95 Report and Recommendation re Defendant Sidney Gladney’s Motion to Dismiss on Double Jeopardy Grounds at 14-16. It does not appear that the Ninth Circuit will be able to summarily dispose of the double jeopardy appeals of these other two defendants under Cretacci.

Furthermore, there is at least some possibility that the Ninth Circuit will not act on the double jeopardy appeals in this case until the United States Supreme Court issues a ruling in the appeal of $405,089.23, 56 F.3d 41. If the high court were to reverse $405,089.23's holding that civil forfeiture proceedings of the kind at issue in this case can serve as the basis for a double jeopardy claim, the Ninth Circuit would have to reinstate all the charges dismissed on double jeopardy grounds in this case. The respondents’ Supreme Court briefs in the appeal of $405,-089.23 are due on March 22, 1996. Ursery, — U.S. —, 116 S.Ct. 762. The appeal of $405,-089.23 will probably be argued in April and decided before the end of June.

9 . It is possible that the Ninth Circuit will decide the double jeopardy appeal without oral argument. See tape No. CR 96-1 of 1/30/96 WB hearing.

10 . The district court may have something to add to my discussion of these discovery matters based on its greater familiarity with the relevant proceedings.

11 . The conditions that I recommended be imposed on Ailemen's release, in addition to the $300,000 bond co-signed by five other persons, were the following:

(1) Confinement to a half-way house twenty-four hours a day, being permitted to leave only to attend court and only in the company of an escort;

(2) With the exception of calls to his attorney, Ailemen would be allowed only three telephone calls per day, which would be monitored;

(3) Ailemen would be permitted reasonable telephone communication with his attorney which would not be monitored;

(4) Ailemen would only speak with his co-defendants, except for his brother Dele Ailemen, in the presence of his attorney;

(5) Ailemen would not possess or use alcohol, controlled substances, firearms, or other weapons;

(6) Ailemen would not take steps to acquire any travel documents;

(7) Ailemen would comply with all other rules of the half-way house.

See tape No. CR 95-28 of 7/28/95 WB hearing.

12 . Salerno, 481 U.S. at 747 n. 4, 107 S.Ct. at 2102 n. 4.

13 . For an extended discussion of whether, in an appellate court's view, conditions imposed by a district court would in fact reduce appreciably the danger to the community that release of the defendant without conditions would pose, and whether the government should be required to provide the support for such conditions, see United States v. Orena, 986 F.2d 628, 632-33 (2d Cir.1993).

14 . Such inconsistency may fail to give prosecutors and courts sufficient incentives to move cases along when necessary to protect the due process rights of pretrial detainees.

15 . Some courts have mentioned additional factors, such as the seriousness of the charges against the defendant, the strength of the government’s case on the merits, the complexity of the case, the hardships caused by detention, the type of threat posed by the defendant, and the extent to which release conditions can minimize risk of flight and potential dangerousness. See United States v. Hare, 873 F.2d 796, 801 (5th Cir.1989); Accetturo, 783 F.2d at 388; Gallo, 653 F.Supp. at 338. Most of these additional considerations are related to the three main factors — length of delay, government responsibility, and flight/dangerousness — and can aid the analysis of whether each of the three main factors weighs in favor of the prosecution or the defense.

16 . As we point out below, determining whether a given defendant qualifies for release under the Bail Reform Act is hardly an exact science. See infra § IV. Not surprisingly, given the predictive character of the judgments (guesses?) involved, the number of cases in which reasonable judges disagree about whether the Act requires detention of the defendant is not small. Moreover, there is a wide range among the defendants who cannot qualify for release — some failing to meet the Act’s requirements by a far greater margin than others. The Infelise court’s approach seems not to take these kinds of considerations fully into account.

17 . There is one opinion, United States v. Gotti, 776 F.Supp. 666 (E.D.N.Y.1991) which quotes the passage from Infelise quoted by us. Gotti, 776 F.Supp. at 671 (quoting Infelise, 934 F.2d at 104-05). However, that opinion also cites cases which support the balancing approach and discusses all three of the factors generally used by courts that follow the balancing approach. See Gotti, 776 F.Supp. at 671-72.

18 . The Zannino court may have been somewhat skeptical about the defendant's claim that his physical condition made him unable to stand trial, for, after his heart attack, the defendant had allegedly continued to direct mafia activities from his hospital bed. 798 F.2d at 547.

19 . The Vastóla court suggested that the length of detention was "critical due to the 'crucial liberty interest at stake.'” Id. at 1448 (quoting United States v. Suppa, 799 F.2d 115, 120 (3d Cir.1986)).

20 . Courts in several cases in addition to Majeed have refused to order release while strongly suggesting that detention will not be allowed to continue if there are further pretrial delays. See Jackson, 823 F.2d at 8 (court approves expected detention of eight months until trial, but states that if trial is postponed further, it will "take any government responsibility for the delay very seriously”); Zannino, 798 F.2d at 549 (court approves "only a short, discrete, and finite period of additional detention" and states that if district court fails to set immediate trial date after imminent examination of defendant’s medical condition, "difficult constitutional issue” will resurface); Berrios-Berrios, 791 F.2d at 254 (while court refuses to order release of defendant already detained for eight months, court states that if trial does not begin for more than a year after detention commenced, "[u]nless it is clear that [the defendant’s] appearance cannot reasonably be assured short of detention, there would be a serious due process problem in continuing her custody”).

21 . None of the six cases ordering release without finding that the prosecution or court system caused any unnecessary delay where expected detention was of short to medium length have been expressly overruled. However, many of them based a finding that continued detention would violate due process primarily upon the length of detention. Since five of these six opinions were issued before the Supreme Court suggested in Salerno, 481 U.S. at 747 n. 4, 107 S.Ct. at 2102 n. 4, that analysis of the due process issue would require comparing the governmental interests served by the detention to the length of the pretrial custodial period, we cannot safely infer that the holdings in those five cases would be the same if they were being decided today. They remain of interest, however, because they constitute some evidence about the core societal values which we look for in resolving substantive due process issues.

22 . Berrios-Berrios, 791 F.2d 246, and Melendez-Carrion I, 790 F.2d 984, discussed earlier, also concerned defendants charged in connection with the Wells Fargo robbery.

23 . While the suggestion that dangerousness to the community might weigh heavier in the government’s favor than risk of flight in a due process balancing analysis seems to have intuitive appeal, it also appears to be at odds with at least some important doctrinal traditions — and those kinds of traditions help inform the substance of due process.

Prior to the enactment of the Bail Reform Act in 1984, pretrial detention without bail had historically been utilized only to assure the defendant’s appearance at trial or to stop the defendant from jeopardizing the trial process by acts such as threats against witnesses. See Melendez-Carrion I, 790 F.2d at 1002; see also United States v. Gotti, 794 F.2d 773, 779 (2d Cir.1986). And the change in the law pursuant to which dangerousness to the community, by itself, could support pretrial detention was controversial. Salerno was a 6-3 decision and was the subject of much judicial and academic debate when it came out. Compare Salerno, 481 U.S. at 741-55, 107 S.Ct. at 2098-2106 (majority opinion) with id. at 755-67, 107 S.Ct. at 2105-12 (Marshall, J., dissenting) and id. at 767-69, 107 S.Ct. at 2112-13 (Stevens, J., dissenting). As the historical and doctrinal underpinnings of pretrial detention based on potential dangerousness are less longstanding and more politically controversial than the foundations of pretrial detention based on risk of flight or a threat to the trial process, it might be argued that the constitutional limits on detention based on potential danger to society should be stricter than the limits on detention imposed in order to prevent flight or protect the trial process. Substantive due process analysis, after all, is supposed to turn on norms that are "so rooted in the traditions and conscience of our people as to be ranked as fundamental.” See Rochin, 342 U.S. at 169, 72 S.Ct. at 208 (quoting Snyder, 291 U.S. at 105, 54 S.Ct. at 332).

We also note, with greater analytical ambition and less analytical confidence, that there have been developments in how the high court distinguishes between "regulatory measures” and "punishment” since Salerno was decided that might make the Court more demanding of the government when it tries to support lengthy pretrial detentions on the ground of dangerousness. While these developments have been made under different constitutional norms than substantive due process, and while the reach of their influence may well be circumscribed, they at least suggest the possibility that the high court might be open to using somewhat more demanding standards when reviewing long pre-trial detentions to determine whether they have crossed the line that separates "regulation” from "punishment.” Compare Salerno, 481 U.S. at 747, 107 S.Ct. at 2101-02 (citing Schall v. Martin, 467 U.S. 253, 269, 104 S.Ct. 2403, 2412, 81 L.Ed.2d 207 (1984); Kennedy v. Mendoza-Martinez, 372 U.S. 144, 168-69, 83 S.Ct. 554, 567-58, 9 L.Ed.2d 644 (1963)); Allen v. Illinois, 478 U.S. 364, 106 S.Ct. 2988, 92 L.Ed.2d 296 (1986) and United States v. Ward, 448 U.S. 242, 248-49, 100 S.Ct. 2636, 2641-42, 65 L.Ed.2d 742 (1980) with Department of Revenue of Montana v. Kurth Ranch, — U.S. —, 114 S.Ct. 1937, 128 L.Ed.2d 767 (1994); Austin v. United States, 509 U.S. 602, 609-623, 113 S.Ct. 2801, 2805-12, 125 L.Ed.2d 488 (1993) and United States v. Halper, 490 U.S. 435, 448, 109 S.Ct. 1892, 1901-02, 104 L.Ed.2d 487 (1989). In Austin, the Court stated that the analysis used in cases such as Mendoza-Martinez, 372 U.S. at 168-69, 83 S.Ct. at 567-68, and Ward, 448 U.S. at 248-49, 100 S.Ct. at 2641-42, applies when the question is "whether a nominally civil penalty should be reclassified as criminal and the safeguards that attend a criminal prosecution should be required,” but not in cases where the "question [is] whether punishment has been imposed" (for certain Eighth Amendment purposes). See 509 U.S. at 610 n. 6, 113 S.Ct. at 2806 n. 6. Unlike the approach in Ward, 448 U.S. at 248-49, 100 S.Ct. at 2641-42, where articulated Congressional intent seems to play a dominant role, under Halper, 490 U.S. at 448, 109 S.Ct. at 1901-02, and Austin, 509 U.S. at 609-611, 113 S.Ct. at 2805-06, the Court is more concerned with the purposes actually served by legislation.

As will be obvious in the text, doctrinal possibilities of the kinds just mentioned play no role in my development of the recommendations in this Report.

24 . The prosecution could attempt to speed the resolution of the appeal of the dismissal of the charges against Ailemen by filing a motion to sever the part of the appeal which concerns Ailemen from the part which concerns his two co-defendants.

25 . These predicted time frames are probably unrealistically short. Hopefully it goes without saying that work on Mr. Ailemen’s case is not my only judicial assignment — and that it is unreasonable to assume that, over a six or eight month period, I will devote every one of my working moments to this one case.

Moreover, the predicted time frames assume that the translation work will not slow down the process of conducting the suppression hearings. Recall that conducting a suppression hearing for a specific wiretap authorization will require translation of all the wiretap evidence gathered prior to that authorization. If the translations necessary for a suppression hearing are not completed before that hearing, the hearing will have to be postponed. Given the difficulties encountered with translation so far, it is not safe to assume that all the translations will be completed in a timely manner. Note also that if the Court of Appeals acts quickly, this will increase the probability that the translation process will delay the suppression hearings, because there will be less time to complete translation before the hearings begin.

26 . Two cases suggest aiding the due process inquiry by comparing the length of expected detention to the prison sentence a defendant would serve if convicted. See Shareef, 907 F.Supp. at 1484 ("it is also appropriate to consider the potential terms of imprisonment to which the defendants may be sentenced if ultimately found guilty of the charges as compared to the prospective length of pretrial detention in determining whether the due process rights of a person may be violated”); Lofranco, 620 F.Supp. at 1325 ("holding a defendant without bail for longer than he would serve if tried and convicted must also violate due process”).

It appears that this consideration can lead to release of the defendant where the potential sentence the defendant faces is very short, but that it cannot justify continued detention once detention has become very long. For example, in Shareef, the court ordered release of a defendant who had already been detained nine months and who would have faced a prison sentence of eighteen to twenty-four months if convicted. See 907 F.Supp. at 1484-85.

However, in cases where the expected length of detention exceeded two years, courts have found that the length of detention weighed in favor of release even though the defendants faced very serious charges and were subject to long prison sentences. See Millan, 4 F.3d at 1043; Ojeda Rios, 846 F.2d at 169; Melendez-Carrion II, 820 F.2d at 60; Gonzales Claudio, 806 F.2d at 341; Gatto, 750 F.Supp. at 675. Thus, my conclusion that the expected length of detention weighs strongly in favor of Ailemen’s release is not altered by the long prison sentence that Ailemen will receive if the dismissed charges against him are reinstated and he is convicted.

27 . Chief Judge Feinberg’s dissenting position that the Bail Reform Act is not facially unconstitutional anticipated the Supreme Court's decision in Salerno, 481 U.S. 739, 107 S.Ct. 2095.

28 . The district court did not authorize the required expenditures until December 27, 1994, seven months after the application was filed. See supra Preliminary Findings of Fact ¶¶ 6-8. Defense counsel also contends that during this seven month hiatus the team that had been assembled to do the translating dispersed — and could not be fully re-assembled when the authorization finally was forthcoming, thus causing additional time to be lost.

29 . While this court-caused delay in translation of the tapes is best treated as a "more neutral reason” than intentional or even negligent conduct, nonetheless "ultimate responsibility for [it] must rest with the government.” Cf. Barker, 407 U.S. at 531, 92 S.Ct. at 2192.

30 . Ailemen suggests that the government's decision to seek a stay of the case on the money-laundering count pending the double jeopardy appeal should be charged against the government. See 12/19/95 Supplemental Memorandum in Support of Motion-for Pre-Trial Release at 16. However, Ailemen did not oppose the govemment's motion for a stay and later filed a cross-appeal arguing for dismissal of the money laundering count on double jeopardy grounds. See supra Preliminary Findings of Fact ¶ 14. For this reason, the government's decision to move for a stay does not weigh in favor of Ailemen's release.

31 . Without the delays occasioned by the slow delivery to the defense of discoverable information and the judicial system’s processing of the request for funds to translate the tapes, most of the case-development could have been concluded before the double jeopardy appeals intervened. And a relatively prompt response by the Court of Appeals, even after Cretacci, 62 F.3d 307, to the government’s challenge of the dismissals of the counts against Pius Ailemen, would have enabled the case against him to go to trial by January or February of 1996. But because of the unnecessary delays, this case now cannot be trial ready until at least the end of this year, and probably until sometime in the first half of 1997.

32 . Under the conditions suggested for his release, Ailemen could pose a serious threat to the safety of the community only if he were to escape from the half-way house and resume his drug trafficking. But if he chose to flee, there is a substantial likelihood that he would not remain in this country, where he would know that he would be the subject of a relentless manhunt. And while he might try from abroad to orchestrate drug transactions that could affect the United States, a threat of this kind is too remote to weigh heavily in our analytical scales.

In any case, since Ailemen could only pose a serious danger to society if he were to escape from the half-way house, the magnitude of the threat that he might pose to society certainly cannot justify continued detention to any greater extent than can the risk of flight.

33 . Ailemen argues that it is improper to base his detention on charges which have been dismissed on double jeopardy grounds. See 9/21/95 Supplemental Memorandum of Points and Authorities in Support of Motion for Pre-Trial Release at 2 & n. 1; 7/24/95 Memorandum of Points and Authorities in Support of Motion for Pre-Trial Release at 6 & n. 4. Ailemen’s argument is aided by Gatto, 750 F.Supp. 664. In Gatto, the court found that the government had a strong case against the defendants on charges of running an illegal gambling business, but a weak case on much more serious charges of violent crimes. See id. at 674. The court stated that the illegal gambling charges alone could not justify detention, and ordered release of the defendants. Id.

However, in this case, the dismissed charges against Ailemen will probably be reinstated by the Ninth Circuit, and the evidence supporting those charges is fairly strong. Moreover, the Bail Reform Act permits a defendant to be detained pending an appeal by the government of the dismissal of an indictment. See 18 U.S.C.A. §§ 3142, 3143(c), 3731 (West Supp.1995); see also Shareef, 907 F.Supp. at 1483. No challenge to the constitutionality of that provision has been made, and it is unlikely, after Salerno, 481 U.S. 739, 107 S.Ct. 2095, that such a challenge would be successful. If the high court does not consider pretrial detention under the Act facially unconstitutional, it does not seem likely that it would hold that detention pending appeal is unconstitutional per se.

Thus, at least in the circumstances presented in this case, where it is probable that the Court of Appeals will reinstate the dismissed charges, and where that probability might well intensify the pressure on the defendant to flee, there does not seem to be any statutory or constitutional bar to considering the dismissed charges in ruling on this motion.

34 . While he was acquitted of all the drug charges, he was convicted of the substantially less serious passport fraud count. He served his relatively modest time on that conviction, apparently without being much the worse for the wear.

35 . The adverse consequences would be much more immediate if the sureties had property to post. Since none of the sureties proffered to date have property to post, they would feel the consequences of a violation by Ailemen over a longer period of time — e.g., in the form of wage garnishments and losses of tax refunds, etc. But the fact that the consequences would be experienced over a longer period of time, in installments, does not necessarily make them less real, especially to people whose only assets are their capacities to earn an income.

36 . These conditions are listed in section V, "Recommendations.” They do not include a requirement that property be posted on Ailemen’s behalf. Because no such property has been available, imposing that condition would be tantamount to ordering his continued detention.

37 . Of course, the standard release conditions also will apply (e.g., that he obey the law, appear as required in court and to serve any sentence imposed, and that he not harass, intimidate, or tamper with any witness, victim, informant, juror or officer of the court).