United States v. HSBC Bank USA, N.A., No. 16-308 (2d Cir. 2017)

Annotate this Case
Justia Opinion Summary

HSBC and the government appealed the district court's grant of a motion by a member of the public to unseal the Monitor's Report in a case involving a deferred prosecution agreement (DPA) with HSBC. The Second Circuit held that the Monitor's Report is not a judicial document because it is not relevant to the performance of the judicial function. By sua sponte invoking its supervisory power at the outset of this case to oversee the government's entry into and implementation of the DPA, the court explained that the district court impermissibly encroached on the Executive's constitutional mandate to "take Care that the Laws be faithfully executed." Furthermore, even assuming arguendo that a district court could revoke a speedy trial waiver were it to later come to question the bona fides of a DPA, the presumption of regularity precludes a district court from engaging in the sort of proactive and preemptive monitoring of the prosecution undertaken here.

Download PDF
16 308(L) United States v. HSBC Bank USA, N.A. UNITED STATES COURT OF APPEALS FOR THE SECOND CIRCUIT _________________________ August Term, 2016 (Argued: March 1, 2017 Decided: July 12, 2017) Docket Nos. 16 308(L), 16 353, 16 1068, 16 1094 _________________________ UNITED STATES OF AMERICA, Plaintiff Appellant, v. HSBC BANK USA, N.A., and HSBC HOLDINGS PLC, Defendants Appellants, HUBERT DEAN MOORE, JR., Before: Appellee. _________________________ KATZMANN, Chief Judge, POOLER, and LYNCH, Circuit Judges. _________________________ In December 2012, the government entered into a five year deferred prosecution agreement (the “DPA”) with HSBC Holdings plc and HSBC Bank, USA, N.A. (collectively, “HSBC”), deferring prosecution of charges under the Bank Secrecy Act, the International Emergency Economic Powers Act, and the Trading with the Enemy Act. The DPA provided for the appointment of an independent monitor charged with preparing periodic reports on HSBC’s ongoing compliance with anti money laundering laws and with the DPA itself. When the parties moved to exclude time under the Speedy Trial Act, the district court concluded that it had supervisory authority to approve or reject the DPA and conditioned its approval of the DPA on its own continued monitoring of the DPA’s implementation. In the exercise of that asserted supervisory authority, the district court later ordered the government to file an annual report prepared by the monitor (the “Monitor’s Report”) regarding HSBC’s compliance efforts. The government did so. In November 2015, the district court received a letter from a member of the public, which it construed as a motion to unseal the Monitor’s Report. The district court granted the motion, subject to redactions, finding that the Monitor’s Report was a judicial document to which the public enjoyed a qualified First Amendment right of access. HSBC and the government appealed the district court’s unsealing and redaction orders, which the district court stayed pending appeal. We hold that the Monitor’s Report is not a judicial document because it is not now relevant to the performance of the judicial function. First, the district court misapprehended its supervisory authority. By sua sponte invoking its supervisory power at the outset of this case to oversee the government’s entry into and implementation of the DPA, the district court impermissibly encroached on the Executive’s constitutional mandate to “take Care that the Laws be faithfully executed.” U.S. Const. art. II, § 3. In the absence of a showing of impropriety, a district court has no authority to supervise the implementation of a DPA. To hold otherwise would be to turn the presumption of regularity on its head. Second, the Monitor’s Report is not relevant to the district court’s role under the Speedy Trial Act. See 18 U.S.C. § 3161(h)(2). Section 3161(h)(2) excludes from the speedy trial clock “[a]ny period of delay during which prosecution is deferred by the attorney for the Government pursuant to written agreement with the defendant, with the approval of the court, for the purpose of allowing the defendant to demonstrate his good conduct.” 18 U.S.C. § 3161(h)(2). We hold that 2 the statute’s requirement of court approval does not imbue the judiciary with ongoing authority to oversee a DPA, but rather authorizes courts to determine that a DPA is genuine and not merely a means of circumventing the speedy trial clock. Third, though the Monitor’s Report might one day be relevant to the judicial function, that is not sufficient to render the Monitor’s Report a “judicial document” today. Accordingly, we reverse. JUDGE POOLER concurs in a separate opinion. _________________________ JENNY C. ELLICKSON (Kenneth A. Blanco and Sung Hee Suh, Deputy Assistant Attorneys General; M. Kendall Day, Chief, Asset Forfeiture and Money Laundering Section; Laura Billings, Trial Attorney, Asset Forfeiture and Money Laundering Section, on the brief ), U.S. Department of Justice, Criminal Division, Appellate Section, for Leslie R. Caldwell, Assistant Attorney General, Washington, DC; Julia Nestor and Alexander A. Solomon (on the brief ), Assistant United States Attorneys for the Eastern District of New York, for Robert L. Capers, United States Attorney for the Eastern District of New York. PAUL D. CLEMENT (Viet D. Dinh, Jeffrey M. Harris, Megan M. Wold, Christopher G. Michel, on the brief ), Kirkland & Ellis LLP, Washington, DC; Samuel W. Seymour and Alexander J. Willscher (on the brief ), Sullivan & Cromwell LLP, New York, NY; Jeffrey B. Wall (on the brief ), Sullivan & Cromwell LLP, Washington, DC, for Defendants Appellants. DAVID A. SCHULZ (Thomas B. Sullivan, Maxwell S. Mishkin, on the brief ), Levine Sullivan Koch & Schulz, LLP, New York, NY, for Appellee. Kevin W. Goering, Norwick, Schad & Goering, New York, NY, for Amicus Curiae Brandon L. Garrett, in support of Appellee. Dennis M. Kelleher, Better Markets, Inc., Washington, DC, for Amicus Curiae Better Markets, Inc., in support of Appellee. 3 Bruce D. Brown (Gregg P. Leslie, D. Victoria Baranetsky, on the brief ), The Reporters Committee for Freedom of the Press, Washington, DC, for Amici Curiae The Reporters Committee for Freedom of the Press, American Society of News Editors, Association of Alternative Newsmedia, The Center for Investigative Reporting, CNBC, Daily News, L.P., Dow Jones & Company, Inc., The E.W. Scripps Company, First Look Media Works, Inc., Forbes Media LLC, The Foundation for National Progress, Gannett Co., Inc., Hearst Corporation, International Documentary Assn., Investigative Reporting Workshop at American University, Digital First Media, MPA – The Association of Magazine Media, National Newspaper Association, National Press Photographers Association, The New York Times Company, The Newspaper Guild Communications Workers of America, Online News Association, Reporters Without Borders, Seattle Times Company, Tully Center for Free Speech, and National Public Radio, Inc., in support of Appellee. _________________________ KATZMANN, Chief Judge: We are called upon in this case to address the role of a district court in monitoring the implementation of a deferred prosecution agreement. In December 2012, plaintiff appellant the United States entered into a five year deferred prosecution agreement (the “DPA”) with defendants appellants HSBC Holdings plc and HSBC Bank, USA, N.A. (collectively, “HSBC”), deferring prosecution of charges under the Bank Secrecy Act, the International Emergency Economic Powers Act, and the Trading with the Enemy Act. The still pending agreement provides that if HSBC complies with its extensive obligations under 4 the DPA, the government will seek the dismissal of those charges at the conclusion of the DPA’s term. If, on the other hand, HSBC breaches the DPA, the government may seek to convict HSBC on the deferred charges. To inform that determination, the DPA provides for the appointment of an independent monitor charged with preparing periodic reports on HSBC’s ongoing compliance with anti money laundering laws and with the DPA itself. When the government and HSBC jointly moved for a speedy trial waiver, the district court (Gleeson, J.) invoked its supervisory power both to review and “approve” the DPA on its merits and to condition its approval on the court’s monitoring of the DPA’s implementation. In the exercise of that asserted authority, the district court subsequently ordered the government to file a confidential report prepared by the independent monitor regarding HSBC’s compliance with the DPA (the “Monitor’s Report”). In November 2015, appellee Hubert Dean Moore, Jr., a member of the public, moved to unseal the Monitor’s Report. The district court granted the motion, subject to redactions, finding that the Monitor’s Report was a “judicial document” to which the public enjoyed a qualified First Amendment right of access. The government and HSBC appeal the district court’s unsealing and redaction orders, arguing that the district court 5 ran afoul of separation of powers principles in involving itself in the implementation of the DPA. We agree. By sua sponte invoking its supervisory power at the outset of this case to oversee the government’s entry into and implementation of the DPA, the district court impermissibly encroached on the Executive’s constitutional mandate to “take Care that the Laws be faithfully executed.” U.S. Const. art. II, § 3. In the absence of evidence to the contrary, the Department of Justice is entitled to a presumption of regularity—that is, a presumption that it is lawfully discharging its duties. Though that presumption can of course be rebutted in such a way that warrants judicial intervention, it cannot be preemptively discarded based on the mere theoretical possibility of misconduct. Absent unusual circumstances not present here, a district court’s role vis à vis a DPA is limited to arraigning the defendant, granting a speedy trial waiver if the DPA does not represent an improper attempt to circumvent the speedy trial clock, and adjudicating motions or disputes as they arise. Because the Monitor’s Report is not now relevant to the performance of the judicial function, it is not a “judicial document” and the district court erred in ordering it unsealed. Accordingly, we reverse. 6 BACKGROUND A. The DPA HSBC Holdings plc (“HSBC Holdings”), incorporated and headquartered in England, is the ultimate parent company of one of the largest banking and financial services groups in the world. HSBC Bank USA, N.A., headquartered in the United States, is a federally chartered banking institution and an indirect subsidiary of HSBC Holdings. In December 2012, following an investigation that lasted more than four years, the United States entered into a five year deferred prosecution agreement with HSBC. See Joint App. 30–104. Under a typical DPA with a corporate defendant, the defendant admits to a statement of facts, submits to the filing of criminal charges against it on the basis of those facts, and agrees to a forfeiture or fine and to institute remedial measures. In exchange, the government agrees to defer prosecution and to ultimately seek dismissal of all charges if the defendant complies with the DPA. If the government determines that the defendant has breached the DPA, however, the government may rip up the agreement and pursue the prosecution. 7 The government’s DPA with HSBC followed this framework. As contemplated by the DPA, the government filed a four count criminal information (the “Information”) charging HSBC Bank, USA, N.A. with willfully violating the Bank Secrecy Act by failing to develop, implement, and maintain an effective anti money laundering program (Count 1) and by failing to conduct due diligence on correspondent bank accounts held on behalf of foreign persons (Count 2). As a result of these failures, some $881 million in drug trafficking proceeds were laundered through HSBC Bank USA, N.A. The government also charged HSBC Holdings with violating U.S. sanctions laws by willfully facilitating financial transactions in the United States for various sanctioned entities, in violation of the International Emergency Economic Powers Act (Count 3), and with willfully facilitating financial transactions for sanctioned entities in Cuba, in violation of the Trading with the Enemy Act (Count 4). In addition, HSBC admitted to a 30 page Statement of Facts and agreed to forfeit $1.256 billion to the United States. HSBC further agreed to continue to cooperate fully with the government and to adopt (or continue to adhere to) dozens of measures designed to remediate the deficiencies in its compliance program. To that end, HSBC 8 Holdings agreed to retain an independent compliance monitor (the “Monitor”) to be approved by the government. As set forth in an attachment to the DPA, the Monitor is charged with “evaluat[ing] . . . the effectiveness of the internal controls, policies and procedures of HSBC Holdings and its subsidiaries . . . as they relate to [those entities’] ongoing compliance with the Bank Secrecy Act, International Emergency Economic Powers Act, Trading With The Enemy Act and other applicable anti money laundering laws . . ., as well as [with] the enumerated remedial measures [in the DPA].” DPA, Attach. B ¶ 1. The DPA requires the Monitor to submit periodic reports to HSBC and the government detailing his findings and making recommendations designed to improve HSBC’s compliance with the DPA and with anti money laundering laws generally. By the terms of the DPA, the Monitor’s reports are intended to remain non public. Finally, the DPA provides that the government will seek to dismiss the Information with prejudice at the conclusion of the DPA’s term if the government determines that HSBC has fully complied with the DPA. If, on the other hand, the government “determines, in its sole discretion, that [HSBC] ha[s] (a) committed any crime under U.S. federal law subsequent to the signing of 9 th[e] [DPA], (b) at any time provided in connection with th[e] [DPA] deliberately false, incomplete, or misleading information, or (c) otherwise breached the [DPA],” all bets are off and the government may pursue the deferred charges. DPA ¶ 16. B. Proceedings Before the District Court For a DPA to function as intended, the parties must obtain an exemption from the Speedy Trial Act’s general requirement that a criminal trial “begin within 70 days after a defendant is charged or makes an initial appearance.” Zedner v. United States, 547 U.S. 489, 492 (2006) (citing 18 U.S.C. § 3161(c)(1)). Absent such an exemption, the logic of any DPA would unravel, as the filed charges would be subject to mandatory dismissal once the 70 day period had run. See 18 U.S.C. § 3162(a)(2) (providing that “[i]f a defendant is not brought to trial within the time limit required by section 3161(c) as extended by section 3161(h), the information or indictment shall be dismissed on motion of the defendant”). Congress anticipated this problem and excluded from the speedy trial clock “[a]ny period of delay during which prosecution is deferred by the attorney for the Government pursuant to written agreement with the defendant, 10 with the approval of the court, for the purpose of allowing the defendant to demonstrate his good conduct.” 18 U.S.C. § 3161(h)(2). Accordingly, when the government filed the Information and the DPA with the district court, the government and HSBC jointly requested that the district court “place th[e] matter into abeyance for a period of sixty months and exclude that time” under the Speedy Trial Act. Joint App. 15. At a subsequent hearing at which the defendants entered pleas of not guilty, the district court asked counsel “what [they] contemplated of the Court’s participation, if any, in the proceedings as they go forward.” Id. at 109. The government responded that it “had not asked the Court to actively take part in overseeing the deferred prosecution agreement,” but instead “simply asked the Court to accept the information for filing and [to] exclude time during the period of the deferred prosecution agreement.” Id. The district court asked the parties to put in submissions explaining why the court should accept the DPA. At the time it made this request, the district court was operating under the misconception that the DPA was “really a plea agreement in the form of an 11C1A charge bargain,” id., and that the United States Sentencing Guidelines therefore instructed it to consider whether the “charges adequately reflect[ed] the seriousness of the 11 actual offense behavior and [whether] accepting the agreement [would] undermine the statutory purposes of sentencing or the sentencing guidelines,” id. at 110. In their resulting submissions, both the government and HSBC took the position that Rule 11(c)(1)(A) of the Federal Rules of Criminal Procedure did not apply and that the district court’s authority at this juncture was limited to determining whether to exclude time under 18 U.S.C. § 3161(h)(2). 1. The July 1, 2013 Approval Order On July 1, 2013, the district court issued a Memorandum and Order granting the parties’ request to hold the case in abeyance pursuant to the Speedy Trial Act and “approv[ing] the DPA pursuant to the Court’s supervisory power.” United States v. HSBC Bank USA, N.A., No. 12 CR 763, 2013 WL 3306161, at *1 (E.D.N.Y. July 1, 2013) (“Approval Order”). After concluding that Rule 11(c)(1)(A) did not apply because the defendants had not entered pleas of guilty or nolo contendere, the district court turned to the text of the Speedy Trial Act, noting that the statute is “silent as to the standard the court should employ when evaluating whether to grant ‘approval’ to a deferred prosecution agreement under 18 U.S.C. § 3161(h)(2).” Id. at *3. Based on its review of legislative history, however, the district court surmised that “§ 3161(h)(2) appears to instruct courts 12 to consider whether a deferred prosecution agreement is truly about diversion and not simply a vehicle for fending off a looming trial date.” Id. The district court harbored no doubt that the DPA was truly about deferring prosecution. Id. Nevertheless, it concluded that it had the authority to approve or reject the DPA on the merits pursuant to its inherent supervisory power. Id. at *4. The exercise of such power was appropriate, the district court reasoned, because “it is easy to imagine circumstances in which a deferred prosecution agreement, or the implementation of such an agreement, so transgresses the bounds of lawfulness or propriety as to warrant judicial intervention to protect the integrity of the Court.” Id. at *6. Though the district court “recognize[d] that the exercise of supervisory power in this context [was] novel” and that any such power must be limited, id., it believed its supervisory power was plainly implicated by virtue of the government filing criminal charges in federal court: [F]or whatever reason or reasons, the contracting parties have chosen to implicate the Court in their resolution of this matter. There is nothing wrong with that, but a pending federal criminal case is not window dressing. Nor is the Court, to borrow a famous phrase, a potted plant. By placing a criminal matter on the docket of a federal court, the parties have subjected their DPA to the legitimate exercise of that court’s authority. 13 . . . The inherent supervisory power serves to ensure that the courts do not lend a judicial imprimatur to any aspect of a criminal proceeding that smacks of lawlessness or impropriety. . . . The parties have asked the Court to lend precisely such a judicial imprimatur to the DPA, by arranging for its implementation within the confines of a pending case. The Court will therefore exercise its supervisory authority over the DPA. Id. at *5–6 (footnote omitted). Having determined that it had supervisory power over the DPA, the district court “approve[d] the DPA . . . subject to [the court’s] continued monitoring of its execution and implementation.” Id. at *7. Pursuant to the court’s declared supervisory power “to ensure that the implementation of the DPA remains within the bounds of lawfulness and respects the integrity of this Court,” the district court directed the parties “to file quarterly reports with the Court to keep it apprised of all significant developments in the implementation of the DPA.” Id. at *11. 2. The Monitor’s Report Beginning in September 2013, in compliance with the district court’s Approval Order, the government began filing quarterly letters with the district court apprising the court of the Monitor’s progress and findings. The government’s April 2015 quarterly report advised the district court that the 14 Monitor had submitted his first annual follow up report (which, again, we refer to here as the “Monitor’s Report”) and described the Monitor’s findings in some detail. The government noted that though the Monitor believed that HSBC was acting in good faith and making meaningful progress in developing an effective compliance program, he also believed that “in certain instances . . . [HSBC’s] progress ha[d] been too slow” and that HSBC “ha[d] a substantial amount of work left to do to implement its written policies.” Joint App. 181. The government also relayed the Monitor’s finding that senior managers at one HSBC business line had failed to properly cooperate with internal compliance reviews, and the government outlined the steps HSBC was taking to address the situation. On April 28, 2015, the district court ordered the government to file the Monitor’s Report with the court. The government sought leave to file the Monitor’s Report under seal, citing confidentiality considerations and concerns that public disclosure of the Monitor’s Report would undermine the effectiveness of the Monitor and serve as a road map for criminals who wished to exploit vulnerabilities in HSBC’s compliance regime. The district court initially accepted the filing of the Monitor’s Report under seal. In November 2015, however, Hubert Dean Moore, Jr., a 15 member of the public, filed a pro se letter with the district court suggesting that the Monitor’s Report might be relevant to a complaint he had filed against HSBC with the Consumer Financial Protection Bureau. The district court entered an order construing Moore’s letter as a motion to unseal the Monitor’s Report. Both the government and HSBC filed letters opposing unsealing. 3. The January 28, 2016 Unsealing Order and the March 9, 2016 Redaction Order On January 28, 2016, the district court issued a Memorandum and Order granting Moore’s motion to unseal the Monitor’s Report in part. In this Circuit, a document filed with the court is a judicial document subject to a presumptive right of access if it is “relevant to the performance of the judicial function and useful in the judicial process.” Lugosch v. Pyramid Co. of Onondaga, 435 F.3d 110, 119 (2d Cir. 2006) (internal quotation mark omitted). The district court determined that the Monitor’s Report satisfied this test for two reasons. First, the Monitor’s Report was relevant to the exercise of its supervisory power, as the court could not “perform [the] task” of “ensur[ing] that the DPA remains within the bounds of lawfulness and respects the integrity of th[e] Court . . . without receiving at least some updates from the parties about HSBC’s compliance with the DPA.” United States v. HSBC Bank USA, N.A., No. 12 CR 763, 2016 WL 16 347670, at *3 (E.D.N.Y. Jan. 28, 2016) (“Unsealing Order”). Second, the Monitor’s Report would be “be integral to the future resolution of the case.” Id. In particular, if the government were to determine that HSBC breached the DPA and that the government would pursue the pending charges, the district court would oversee those proceedings. If, on the other hand, the government were to seek to dismiss the charges at the conclusion of the DPA’s term, such a dismissal could only be effectuated under Federal Rule of Criminal Procedure 48 with leave of court. The decision whether to grant such leave, the district court reasoned, could not “properly be made without judicial review of the [Monitor’s] Report.” Id. The district court then concluded that a qualified First Amendment right of public access attached to the Monitor’s Report and that the confidentiality concerns articulated by the government and HSBC could be addressed with targeted redactions and by maintaining five of the Monitor’s Report’s six appendices under seal. To that end, the district court invited the parties to submit proposed redactions to the Monitor’s Report, which the parties did. On March 9, 2016, the district court issued an order (the “Redaction Order”) describing the 17 redactions it had made to the Monitor’s Report and staying the unsealing of the Monitor’s Report pending appellate review. The government and HSBC timely appealed the Unsealing Order and the Redaction Order. Because the government and HSBC are aligned on this appeal, this Court granted Moore, who is now represented by pro bono counsel, leave to intervene as appellee. DISCUSSION I. Before assessing the merits, we must first determine whether we have jurisdiction over this consolidated appeal of interlocutory orders.1 Ordinarily, “interlocutory orders are not appealable as a matter of right.” United States v. Graham, 257 F.3d 143, 147 (2d Cir. 2001) (internal quotation mark omitted). An exception to this rule is the collateral order doctrine, which provides that interlocutory orders are appealable if they: (1) “conclusively determine the disputed question,” (2) “resolve an important issue completely separate from the merits of the action,” and (3) will “be effectively unreviewable on appeal from a final judgment.” Id. (internal quotation mark omitted). We find each criterion The parties agree that we have jurisdiction under the collateral order doctrine. That, of course, does not settle the issue, because “[j]urisdiction cannot be created by the consent of the parties.” New York v. Shinnecock Indian Nation, 686 F.3d 133, 138 (2d Cir. 2012). 1 18 met here. First, the challenged orders “conclusively determine the disputed question[s]” of whether the Monitor’s Report is a judicial document and the extent to which it must be disclosed. Id. Second, the challenged orders “resolve an important issue completely separate from the merits of the action,” as whether the Monitor’s Report is a judicial document that must be unsealed has nothing to do with the merits of the underlying criminal charges. Id. Third, the challenged orders will “be effectively unreviewable on appeal from a final judgment,” id. (internal quotation mark omitted), as “[o]nce the information is disclosed, the ‘cat is out of the bag’ and appellate review is futile.” Al Odah v. United States, 559 F.3d 539, 544 (D.C. Cir. 2009) (invoking collateral order doctrine to exercise jurisdiction over appeal of order compelling government to share classified information with petitioners’ counsel); see also S.E.C. v. TheStreet.Com, 273 F.3d 222, 228 (2d Cir. 2001) (finding interlocutory unsealing order “unreviewable on appeal from a final judgment” “because the alleged harm caused by disclosure . . . will be immediate and irreparable”); Graham, 257 F.3d at 147–48 (exercising jurisdiction pursuant to collateral order doctrine over appeal of district court’s interlocutory order requiring government to release tapes to media).2 2 Because we find that we have jurisdiction under the collateral order doctrine, we 19 II. The threshold merits question in this case is whether the Monitor’s Report is a judicial document, as only judicial documents are subject to a presumptive right of public access, whether on common law or First Amendment grounds.3 See Lugosch, 435 F.3d at 119–20. Though we review the “ultimate decision to seal or unseal for abuse of discretion,” we review the determination that the Monitor’s Report is a judicial document de novo. Bernstein v. Bernstein Litowitz Berger & Grossmann LLP, 814 F.3d 132, 139 (2d Cir. 2016). In this Circuit, to qualify as a “judicial document” subject to a presumptive right of public access, “the item filed must be relevant to the performance of the judicial function and useful in the judicial process.” United States v. Amodeo, 44 F.3d 141, 145 (2d Cir. 1995) (“Amodeo I”). The appellants contend that the Monitor’s Report is not a judicial document, but rather an “executive document” designed to inform the executive branch’s exercise of its prosecutorial discretion in determining whether to dismiss HSBC’s petition for a writ of mandamus as moot. HSBC asserts that because the Monitor’s Report is a confidential executive document, we should not even reach the question of whether the Monitor’s Report is a judicial document. But HSBC cites no authority for the proposition that we can simply set aside the test for whether a given document constitutes a judicial document. To do so would be to assume the answer before undertaking the inquiry. 3 20 dismiss or pursue the deferred charges. That discretion, HSBC notes, flows from the Executive’s duty under Article II of the Constitution to “take Care that the Laws be faithfully executed.” U.S. Const. art. II, § 3; see also United States v. Huerta, 878 F.2d 89, 92 (2d Cir. 1989) (“The Executive . . . has the exclusive authority to decide whether to prosecute . . . .”); United States v. Fokker Servs. B.V., 818 F.3d 733, 741 (D.C. Cir. 2016) (“Decisions to initiate charges, or to dismiss charges once brought, lie at the core of the Executive’s duty to see to the faithful execution of the laws.” (internal quotation marks and alteration omitted)). The appellants thus argue that the district court invaded the province of the Executive in preemptively invoking the court’s supervisory authority to oversee the implementation of the DPA and the Monitor’s work. Because the district court lacked the authority it claimed to have, the argument goes, the Monitor’s Report cannot be a judicial document. Moore and his amici, by contrast, assert that the Monitor’s Report is relevant to the performance of the judicial function on four separate grounds. Specifically, they assert that the Monitor’s Report is relevant to: (1) the district court’s supervisory authority to approve and supervise the implementation of the DPA; (2) the district court’s authority to approve the DPA under the Speedy 21 Trial Act; (3) the district court’s assessment of any Rule 48(a) motion for dismissal that the government might bring at the conclusion of the DPA’s term; and (4) the district court’s adjudication of the proceedings in the event that the government alleges that HSBC breached the DPA. We analyze these proffered bases for treating the Monitor’s Report as a judicial document in turn. A. The District Court’s Asserted Supervisory Power The district court ordered the filing of the Monitor’s Report for its review in the exercise of its stated supervisory power to monitor the implementation of the DPA and to condition its approval of the DPA on such monitoring. If the district court’s conception of its supervisory power in this context were correct, the Monitor’s Report would quite obviously “be relevant to the performance of the judicial function and useful in the judicial process.” Amodeo I, 44 F.3d at 145. Thus, even though the appellants did not attempt to appeal the district court’s Approval Order announcing its supervisory power over the DPA, whether the district court was correct to assert such a power is squarely at issue on this appeal. We hold that the district court erred in sua sponte invoking its supervisory power to monitor the implementation of the DPA in the absence of a showing of impropriety. 22 The supervisory power “permits federal courts to supervise ‘the administration of criminal justice’ among the parties before the bar.” United States v. Payner, 447 U.S. 727, 735 n.7 (1980) (quoting McNabb v. United States, 318 U.S. 332, 340 (1943)). “The purposes underlying use of the supervisory powers are threefold: to implement a remedy for violation of recognized rights; to preserve judicial integrity by ensuring that a conviction rests on appropriate considerations validly before the jury; and finally, as a remedy designed to deter illegal conduct.” United States v. Hasting, 461 U.S. 499, 505 (1983) (citations omitted); Mesarosh v. United States, 352 U.S. 1, 14 (1956) (“This is a federal criminal case, and this Court has supervisory jurisdiction over the proceedings of the federal courts. If it has any duty to perform in this regard, it is to see that the waters of justice are not polluted.” (footnote omitted)). Courts have also traditionally exercised their supervisory powers to establish rules of evidence and procedure to govern the administration of justice in the federal courts. See, e.g., McNabb, 318 U.S. at 340–45 (holding that confessions obtained in violation of congressional requirement to promptly present defendant to court must be excluded, even though Congress had not explicitly forbidden the admission of such confessions); Bank of Nova Scotia v. United States, 487 U.S. 250, 254 (1988) (“In 23 the exercise of its supervisory authority, a federal court may, within limits, formulate procedural rules not specifically required by the Constitution or the Congress.” (internal quotation marks omitted)). However, “[t]he supervisory power doctrine is an extraordinary one which should be ‘sparingly exercised.’” United States v. Jones, 433 F.2d 1176, 1181–82 (D.C. Cir. 1970) (quoting Lopez v. United States, 373 U.S. 427, 440 (1963)); see also Chambers v. NASCO, Inc., 501 U.S. 32, 44 (1991) (“Because of their very potency, inherent powers must be exercised with restraint and discretion.”). As the district court recognized below, “[i]n the typical supervisory power case, the defendant raises a purported impropriety in the federal criminal proceeding and seeks the court’s redress of that impropriety.” HSBC Bank USA, N.A., 2013 WL 3306161, at *6; see also United States v. Johnson, 221 F.3d 83, 96 (2d Cir. 2000) (“[G]enerally the exercise of supervisory power arises in the context of requests by defendants to vacate convictions, dismiss indictments, or invalidate sentences . . . .” (citations omitted)). Indeed, “the federal judiciary’s supervisory powers over prosecutorial activities that take place outside the courthouse is extremely limited, if it exists at all.” United States v. Lau Tung Lam, 714 F.2d 209, 210 (2d Cir. 1983). 24 The district court justified its concededly “novel” exercise of supervisory power in this context by observing that “it is easy to imagine circumstances in which a deferred prosecution agreement, or the implementation of such an agreement, so transgresses the bounds of lawfulness or propriety as to warrant judicial intervention to protect the integrity of the Court.” HSBC Bank USA, N.A., 2013 WL 3306161, at *6. We agree that it is not difficult to imagine such circumstances. But the problem with this reasoning is that it runs headlong into the presumption of regularity that federal courts are obliged to ascribe to prosecutorial conduct and decisionmaking. That presumption is rooted in the principles that undergird our constitutional structure. In particular, “because the United States Attorneys are charged with taking care that the laws are faithfully executed, there is a ‘presumption of regularity support[ing] their prosecutorial decisions and, in the absence of clear evidence to the contrary, courts presume that they have properly discharged their official duties.’” United States v. Sanchez, 517 F.3d 651, 671 (2d Cir. 2008) (alteration in original) (quoting United States v. Armstrong, 517 U.S. 456, 464 (1996)). In resting its exercise of supervisory authority on hypothesized scenarios of egregious misconduct, the district court turned this presumption on its head. See HSBC Bank USA, N.A., 2013 WL 25 3306161, at *6 (“[C]onsider a situation where the current monitor needs to be replaced. What if the replacement’s only qualification for the position is that he or she is an intimate acquaintance of the prosecutor proposing the appointment?” (citation omitted)). Rather than presume “in the absence of clear evidence to the contrary” that the prosecutors administering the DPA were “properly discharg[ing] their official duties,” the district court invoked its supervisory power—and encroached on the Executive’s prerogative—based on the mere theoretical possibility that the prosecutors might one day abdicate those duties. Sanchez, 517 F.3d at 671 (internal quotation mark omitted). To be sure, in its history, this nation has not been free of executive misconduct and abuse of power. And if misconduct in the implementation of a DPA came to a district court’s attention (for example, through a whistleblower filing a letter with the court), the district court might very well be justified in invoking its supervisory power sua sponte to monitor the implementation of the DPA or to take other appropriate action.4 The point is not that the court can or Moore suggested at oral argument that such misconduct had come to the district court’s attention via the government’s April 2015 quarterly report, the filing of which precipitated the district court’s order that the government file the Monitor’s Report with the court. But by “misconduct,” we do not mean revelations of the kind included in the April 2015 quarterly report regarding the pace of HSBC’s remediation efforts and 4 26 should disregard governmental misconduct, but rather that a federal court has no roving commission to monitor prosecutors’ out of court activities just in case prosecutors might be engaging in misconduct. See Fokker Servs. B.V., 818 F.3d at 744 (“[A]lthough charges remain pending on the court’s docket under a DPA, the court plays no role in monitoring the defendant’s compliance with the DPA’s conditions.”); In re U.S., 503 F.3d 638, 641 (7th Cir. 2007) (“[A] judicial effort to supervise [a prosecutor’s] process of reaching a decision intrudes impermissibly into the activities of the Executive Branch of government.”). In sum, because the district court has no freestanding supervisory power to monitor the implementation of a DPA, the Monitor’s Report cannot be deemed “relevant to the performance of the judicial function” on that basis. Lugosch, 435 F.3d at 119. B. The District Court’s Role under the Speedy Trial Act Amicus Curiae Professor Brandon Garrett suggests that the Monitor’s Report is relevant to the district court’s role under the Speedy Trial Act as specified by 18 U.S.C. § 3161(h)(2). Section 3161(h)(2) excludes from the speedy trial clock “[a]ny period of delay during which prosecution is deferred by the deficiencies in HSBC’s corporate culture that HSBC leadership was working to address. We mean misconduct that smacks of impropriety. 27 attorney for the Government pursuant to written agreement with the defendant, with the approval of the court, for the purpose of allowing the defendant to demonstrate his good conduct.” 18 U.S.C. § 3161(h)(2) (emphasis added). But what is it that courts are to approve and on what basis? On one hand, the text could be read to limit the court’s inquiry to the issue of whether the parties’ agreement is genuinely “for the purpose of allowing the defendant to demonstrate his good conduct”—that is, to the issue of whether the parties are acting in good faith. Id. On the other hand, at least in isolation, the text could plausibly be read to imbue courts with the authority to “approve” (or reject) a DPA as a public policy matter and to decline to grant a speedy trial waiver for virtually any reason.5 Indeed, according to Professor Garrett, “whether to approve a DPA” pursuant to § 3161(h)(2) “is necessarily combined with substantive review of [the DPA] and is joined with ongoing supervision of the case.” Amicus Br. of Prof. Brandon L. Garrett 26. In United States v. Fokker Services B.V., 818 F.3d 733 (D.C. Cir. 2016), the D.C. Circuit confronted this interpretive question when the parties in that case appealed the district court’s refusal to grant a speedy trial waiver based on its Of course, the district court granted the government and HSBC a speedy trial waiver well before the Monitor’s Report even existed, so the Monitor’s Report could not be deemed relevant to that initial determination. 5 28 view that the DPA at issue was too lenient. See id. at 737–38, 742–45. The D.C Circuit vacated the district court’s order, reasoning that Congress, in enacting § 3161(h)(2), “acted against the backdrop of long settled understandings about the independence of the Executive with regard to charging decisions” and that “[n]othing in the statute’s terms or structure suggests any intention to subvert those constitutionally rooted principles so as to enable the Judiciary to second guess the Executive’s exercise of discretion over the initiation and dismissal of criminal charges.” Id. at 738. We agree. At least in the absence of any clear indication that Congress intended courts to evaluate the substantive merits of a DPA or to supervise a DPA’s out of court implementation, the relative functions and competence of the executive and judicial branches counsel against Professor Garrett’s interpretation. Subject to constitutional constraints, “[t]he Executive . . . has the exclusive authority to decide whether to prosecute and to choose among alternative charges.” Huerta, 878 F.2d at 92 (citations omitted). There are good reasons for that delegation of authority: “Few subjects are less adapted to judicial review than the exercise by the Executive of his discretion in deciding when and whether to institute criminal proceedings, or what precise charge shall be made, 29 or whether to dismiss a proceeding once brought.” United States v. Ross, 719 F.2d 615, 620 (2d Cir. 1983); Holley v. Lavine, 553 F.2d 845, 850 (2d Cir. 1977). As the Supreme Court has explained, “[s]uch factors as the strength of the case, the prosecution’s general deterrence value, the Government’s enforcement priorities, and the case’s relationship to the Government’s overall enforcement plan are not readily susceptible to the kind of analysis the courts are competent to undertake.” Wayte v. United States, 470 U.S. 598, 607 (1985). Put simply, our role is not to act as “superprosecutors,” second guessing the legitimate exercise of core elements of prosecutorial discretion, but rather as neutral arbiters of the law. Inmates of Attica Corr. Facility v. Rockefeller, 477 F.2d 375, 380 (2d Cir. 1973) (internal quotation marks omitted). If, in the context of DPAs, Congress intended to rejigger the historical allocation of authority between the courts and the Executive, we would expect it to do so rather clearly. Congress, at least as a general matter, does not “hide elephants in mouseholes.” Whitman v. Am. Trucking Ass’ns, 531 U.S. 457, 468 (2001). But Congress did not speak clearly in § 3161(h)(2)—far from it—and we thus decline to interpret that provision’s vague “approval” requirement as imbuing courts with an ongoing oversight power over the government’s entry 30 into or implementation of a DPA. Rather, we hold that § 3161(h)(2) authorizes courts to determine that a DPA is bona fide before granting a speedy trial waiver—that is, that the DPA in question is genuinely intended to “allow[] the defendant to demonstrate his good conduct,” § 3161(h)(2), and does not constitute a disguised effort to circumvent the speedy trial clock. See Fokker Servs. B.V., 818 F.3d at 744–45 (adopting this interpretation). As the D.C. Circuit reasoned in Fokker, such an interpretation accords with the ordinary distribution of power between the judiciary and the Executive in the realm of criminal prosecution.6 See id. at 738, 741–45. To the extent that Moore and Professor Garrett contend that the district court’s mandate to assess the bona fides of a DPA survives as long as the DPA is pending—and that the Monitor’s Report was relevant to this function—we are The D.C. Circuit in Fokker, like the district court here, also pointed to legislative history in support of its interpretation of § 3161(h)(2). In particular, the court noted that the Senate Committee Report accompanying the Speedy Trial Act explained that the purpose of the “approval” requirement was to “assure[] that the court will be involved in the decision to divert and that the procedure will not be used by prosecutors and defense counsel to avoid the speedy trial time limits.” Fokker Servs. B.V., 818 F.3d at 745 (quoting S. Rep. No. 93–1021, at 37 (1974)); see also HSBC Bank USA, N.A., 2013 WL 3306161, at *3 (quoting same). While the Senate Committee Report reinforces the notion that Congress was concerned with potential abuse of DPAs as a means to circumvent the speedy trial clock, the Report’s reference to a district court’s “involv[ment] in the decision to divert” is opaque. S. Rep. No. 93–1021, at 37. By our lights, the legislative history is ambiguous and does not clearly support either the D.C. Circuit’s view or Professor Garrett’s view. 6 31 not persuaded. Even assuming arguendo that a district court could revoke a speedy trial waiver were it to later come to question the bona fides of a DPA, the presumption of regularity precludes a district court from engaging in the sort of proactive and preemptive monitoring of the prosecution undertaken here. C. The District Court’s Role when Considering a Rule 48(a) Motion or Adjudicating a Claimed Breach of the DPA Moore next argues that the Monitor’s Report is a judicial document because it will either be relevant to deciding a motion to dismiss the Information at the conclusion of the DPA’s term or, alternatively, to adjudicating any claimed breach of the DPA. These arguments fail. Even if we assume that the Monitor’s Report might be relevant to the judicial function at a later point in this case, that would not be sufficient to render the Monitor’s Report a judicial document now. Critically, as we held above, the district court erred in ordering the government to file the Monitor’s Report pursuant to its authority over the implementation of the DPA for the simple reason that the district court had no such authority. Thus, although the government complied with the district court’s order, the Monitor’s Report—for purposes of the public access doctrine—is not unlike a document exchanged by the parties in the course of litigation that has not yet been brought to the 32 attention of the court. And we have long recognized that documents “passed between the parties in discovery[] lie entirely beyond the . . . reach” of the presumption of public access. United States v. Amodeo, 71 F.3d 1044, 1050 (2d Cir. 1995) (“Amodeo II”); cf. Amodeo I, 44 F.3d at 145 (“[T]he mere filing of a paper or document with the court is insufficient to render that paper a judicial document . . . .”). If the rule were otherwise, deposition transcripts, interrogatories, and documents exchanged in discovery would become “judicial documents” to which the public could demand access before the parties had even contemplated filing such documents with the court. While “public monitoring is an essential feature of democratic control,” such an approach would constitute a radical expansion of the “public access” doctrine. Amodeo II, 71 F.3d at 1048 (“Unlimited access to every item turned up in the course of litigation would be unthinkable.”); S.E.C. v. Am. Int’l Grp., 712 F.3d 1, 4 (D.C. Cir. 2013) (“[T]hough filing a document with the court is not sufficient to render the document a judicial record, it is very much a prerequisite.”). Two of our prior cases warrant further discussion in light of their superficial similarity to this case. In Amodeo I, a “Court Officer” charged with investigating union related corruption pursuant to a consent decree filed a 33 progress report with the district court that included a sealed exhibit. 44 F.3d at 143–44. We held that the progress report and the exhibit were judicial documents. Id. at 146. We reached that conclusion in part because the court officer was permitted under the governing consent decree to apply for assistance from the court and because “the progress report certainly would be germane in assessing such an application.” Id. We also noted that the consent decree “provide[d] for any party to seek enforcement of, or relief from, any of the provisions of the [d]ecree” and that the record to be considered on such an application “surely would include the matters reported by the Court Officer.” Id. Moore thus suggests that Amodeo I stands for the proposition that if a filed document could later become relevant to the judicial function, then it is relevant to the judicial function. But, to the contrary, the court officer’s reports in Amodeo I were immediately relevant to the judicial function because they went to the district court’s authority “to make sure that the court officer [was] doing what she was appointed to do.” Id. (internal quotation mark omitted). As we explained, “[i]t certainly was helpful to the court to know that the Court Officer was fulfilling the duties assigned to her by the Consent Decree in this case.” Id. Moreover, the consent decree at issue in Amodeo I explicitly contemplated court 34 involvement in enforcing the decree, whereas the DPA entrusts the implementation of the DPA to the Justice Department. The consent decree thus “itself ma[de] the reports and exhibits filed by the Court Officer relevant to the performance of the judicial function and useful in the judicial process.” Id. The DPA does no such thing. Similarly, in United States v. Erie County, New York, 763 F.3d 235 (2d Cir. 2014), the United States and Erie County settled constitutional claims arising out of conditions at two Erie County correctional facilities. Id. at 236–37. The settlement agreement provided for the appointment of two “compliance consultants” and required the compliance consultants to file their reports with the district court. Id. at 237. The settlement agreement also empowered the district court to order “any relief permitted by law or equity” in the event that Erie County breached the settlement agreement. Id. (internal quotation marks omitted). Noting that the “the compliance reports could form the basis for the District Court reinstating the civil proceedings sua sponte if the Court believe[d] that the substance of the stipulated order of dismissal [was] not being fulfilled,” we held that the compliance reports were judicial documents subject to a right of public access. Id. at 240. Critically, the district court in Erie County was “involved 35 in effectuating [the] settlement agreement,” id. at 241, and had a “role in overseeing [the] progress” being made thereunder, id. at 242. “The fact that the District Court ha[d] not yet acted sua sponte or adjudicated an enforcement action brought by the parties d[id] not alter our analysis” because “even the District Court’s inaction [was] subject to public accountability.” Id. By contrast, because the district court here lacked authority to oversee the implementation of the DPA, the district court’s inaction during the pendency of the DPA would have been indicative of nothing other than due respect for a coordinate branch of government. We also disagree with Moore’s assertion that “the [Monitor’s] Report necessarily will be useful, one way or the other, when the DPA ends.” Appellee Br. 30. Rule 48(a) of the Federal Rules of Criminal Procedure provides that “[t]he government may, with leave of court, dismiss an indictment, information, or complaint.” Fed. R. Crim. P. 48(a) (emphasis added). In Rinaldi v. United States, 434 U.S. 22 (1977) (per curiam), the Supreme Court explained that though the “leave of court” requirement “obviously vest[s] some discretion in the court,” “[t]he principal object of the ‘leave of court’ requirement is apparently to protect a defendant against prosecutorial harassment, e.g., charging, dismissing, and 36 recharging, when the Government moves to dismiss an indictment over the defendant’s objection.” Id. at 29 n.15. The Court noted that Rule 48(a) “has also been held to permit the court to deny a Government dismissal motion to which the defendant has consented if the motion is prompted by considerations clearly contrary to the public interest.” Id. But the Court reserved decision on whether such an approach was permissible. Id. For our part, we have suggested (in dictum) that any authority a court might have to deny a Rule 48(a) motion would be limited to cases in which dismissal is “clearly contrary to manifest public interest.” United States v. Pimentel, 932 F.2d 1029, 1033 n.5 (2d Cir. 1991) (quoting United States v. Cowan, 524 F.2d 504, 513 (5th Cir. 1975)). Some courts of appeals that have elucidated this “public interest” test have stressed how “severely cabined” it is, “equat[ing] a dismissal that is clearly contrary to the public interest with one in which the prosecutor appears motivated by bribery, animus towards the victim, or a desire to attend a social event rather than trial”—in other words, bad faith. In re Richards, 213 F.3d 773, 787–88 (3d Cir. 2000); see also Rice v. Rivera, 617 F.3d 802, 811 (4th Cir. 2010) (“Put succinctly, a Rule 48 motion that is not motivated by bad faith is not clearly contrary to manifest public interest, and it must be granted.” 37 (internal quotation marks omitted)). Others have emphasized that a court deciding a Rule 48(a) motion should not “serve merely as a rubber stamp for the prosecutor’s decision.” United States v. Ammidown, 497 F.2d 615, 622 (D.C. Cir. 1973); see also Cowan, 524 F.2d at 512–13. Whatever the precise contours of a district court’s authority in resolving a Rule 48(a) motion, it is certainly not the case that the Monitor’s Report will, as Moore contends, necessarily be relevant to deciding such a motion. For example, where a Rule 48(a) motion is uncontested and the parties’ good faith is not in doubt, a district court is unlikely to have any occasion to demand a monitor’s reports (or analogous work product) for the court’s inspection before granting the Rule 48(a) motion. Nor do we accept that the Monitor’s Report will necessarily be relevant to adjudicating any claimed breach of the DPA, as the claimed breach could be based, for example, on HSBC’s obligation not to commit a crime under federal law or its obligation not to contradict the Statement of Facts in a public statement. Conversely, the Monitor’s Report may indeed be relevant in determining whether to grant an eventual Rule 48(a) motion or in adjudicating a claimed breach of the DPA. For example, there may be circumstances that suggest that the motion is being made in bad faith, thereby raising a question as 38 to whether dismissing the Information would be “clearly contrary to manifest public interest.” Pimentel, 932 F.2d at 1033 n.5. The salient point is that attempting to forecast the relevance of the Monitor’s Report at the conclusion of the DPA’s term is inherently speculative. Such prognostication cannot support treating the Monitor’s Report as a judicial document now. To be clear, we do not hold that documents can be deemed “judicial documents” only once a court has already considered them. Our case law is clear that pleadings and summary judgment papers, for example, are judicial documents upon filing—which is eminently sensible given that such documents, by definition, ask the court to grant (or reject) some relief. See Bernstein, 814 F.3d at 140; Lugosch, 435 F.3d at 120–23. But that does not mean that any document that is docketed with a court is a judicial document, regardless of the likelihood that it will ever be relevant to the judicial function. Because the Monitor’s Report is not a judicial document, the district court abused its discretion in ordering it unsealed.7 In light of our conclusion that the Monitor’s Report is not a judicial document, we do not reach whether a First Amendment or common law right of access would have ultimately required the unsealing of the document (in any form) under the circumstances of this case. In addition, that the Monitor’s Report is not a judicial document does not mean that Moore or other members of the public are necessarily without recourse. “The appropriate device” for obtaining executive records “is a 7 39 CONCLUSION Striking the appropriate balance between the prerogatives of the three branches of government is never easy. We emphasize that while the district court exceeded its authority in this case, the Take Care Clause of the Constitution is not a blank check. Where the presumption of regularity has been called into question, we do not foreclose the possibility that steps of the kind taken by the district court here could be warranted. But that is not this case. For the foregoing reasons, the orders of the district court are REVERSED. Freedom of Information Act request addressed to the relevant agency.” United States v. El Sayegh, 131 F.3d 158, 163 (D.C. Cir. 1997). We offer no view on whether any of FOIA’s exemptions would apply. 40 POOLER, Circuit Judge: I concur in the opinion and the judgment, but write separately to suggest that it is time for Congress to revisit the issue of deferred and nonprosecution agreements (collectively, “DPAs”). DPAs exist because Section 3161(h)(2) of the Speedy Trial Act excludes “[a]ny period of delay during which prosecution is deferred by the attorney for the Government pursuant to written agreement with the defendant, with the approval of the court, for the purpose of allowing the defendant to demonstrate his good conduct.” 18 U.S.C. § 3161(h)(2). Without this exception, the filing of the criminal information would trigger the running of the speedy trial clock. The Senate Judiciary Committee explained that this section was included in the Act: to encourage the current trend among United States attorneys to allow for deferral of prosecution on the condition of good behavior. A number of Federal and State courts have been experimenting with pretrial diversion or intervention programs in which prosecution of a certain category of defendants is held in abeyance on the condition that the defendant participate in a social rehabilitation program. If the defendant succeeds in the program, charges are dropped. Such diversion programs have been quite successful with first offenders in Washington, D.C. (Project Crossroads) and in New York City (Manhattan Court Employment Project). Some success has also been noted in programs where the defendant s alleged criminality is related to a specific social problem such as prostitution or heroin addiction. Of course, in the absence of a provision allowing the tolling of the speedy trial time limits, prosecutors would never agree to such diversion programs. Without such a provision the defendant could automatically obtain a dismissal of charges if prosecution were held in abeyance for a period of time in excess of the time limits set out in section 3161 (b) and (c). This section of S. 754 differs from its counterpart in S. 895. It now requires that exclusion for diversion only be allowed where deferral of prosecution is conducted “with approval of the court.” This assures that the court will be involved in the decision to divert and that the procedure will not be used by prosecutors and defense counsel to avoid the speedy trial time limits. S. Rep. No. 93–1021, at 36–37 (1974)). The two programs mentioned in the legislative history, Project Crossroads and the Manhattan Court Employment Project, were pretrial diversion programs aimed at helping individual defendants avoid the collateral consequences of a criminal conviction through programs that included education, job training, and substance abuse treatment. See, e.g., Note, Pretrial Diversion from the Criminal Process, 83 Yale L.J. 827 (1974). The programs were viewed as “a functional equivalent of a sentence to pretrial probation,” id. at 843, and were staffed with 2 paraprofessionals overseeing individuals in what was, “in effect a probationary type of supervision and control,” id. at 845. In recent years, however, DPAs increasingly are used not to divert individual defendants but rather to divert corporations from criminal charges. Unlike individuals, corporations are not diverted into probation like programs supervised by paraprofessionals. Rather, they enter into negotiated agreements with prosecutors that set forth the facts to which the corporation admits and a remedy that typically includes both a fine and an agreement for the corporation to make structural changes. The prosecution retains sole discretion to decide if the corporation adequately complied with the agreement, allowing the prosecution to act as prosecutor, jury, and judge. Prosecutors can enforce legal theories without such theories ever being tested in a court proceeding. Using DPAs in this manner is neither improper nor undesirable. An indictment alone can deal a death blow to a corporation, with severe collateral consequences for blameless employees and shareholders. As the law governing DPAs stands now, however, the prosecution exercises the core judicial functions of adjudicating guilt and imposing sentence with no meaningful oversight from the courts. 3 I respectfully suggest it is time for Congress to consider implementing legislation providing for such review. See United States v. Saena Tech Corp., 140 F. Supp. 3d 11, 30 n.9 (D.D.C. 2015) (“This ambiguity, combined with the fact that Congress s original purpose had nothing to do with the broad ranging corporate deferred prosecution agreements that have become commonplace, suggests that congressional action to clarify the standards a court should apply when confronted with a corporate deferred prosecution agreement may be appropriate.”). A bill introduced in the House in 2014 would, among other things, require the development of public, written guidelines for DPAs; require the text of DPAs to be placed on a Justice Department website; and require DPAs to be submitted to district courts for review. Accountability in Deferred Prosecution Act of 2014, H.R. 4540, 113th Cong. (2014). Courts would be charged with “approv[ing] the [DPA] if the court determines the [DPA] is consistent with the guidelines for such agreements and is in the interests of justice.” Id. at § 7(a). In addition, the parties and any monitors would be required to file quarterly reports regarding the progress toward completions of the DPA with the court. Id. at § 7(b). Finally, the court would be charged with “review[ing] the 4 implementation or termination of the [DPA], and take any appropriate action, to assure that the implementation or termination is consistent with the interests of justice.” Id. at § 7(c). Legislation along the lines of this proposed act would restore some balance in the DPA process. Accordingly, I respectfully concur. 5