In Re Atlantic Pipe Corp.

SELYA, Circuit Judge. This mandamus proceeding 1 requires us to resolve an issue of importance to judges and practitioners alike: Does a district court possess the authority to compel an unwilling party to participate in, and share the costs of, non-binding mediation conducted by a private mediator? We hold that a court may order mandatory mediation pursuant to an explicit statutory provision or local rule. We further hold that where, as here, no such authorizing medium exists, a court nonetheless may order mandatory mediation through the use of its inherent powers as long as the case is an appropriate one and the order contains adequate safeguards. Because the mediation order here at issue lacks such safeguards (although it does not fall far short), we vacate it and remand the matter for further proceedings. I. BACKGROUND In January 1996, Thames-Dick Supera-queduct Partners (Thames-Dick) entered into a master agreement with the Puerto Rico Aqueduct and Sewer Authority (PRA-SA) to construct, operate, and maintain the North Coast Superaqueduct Project (the Project). Thames-Dick granted subcontracts for various portions of the work, including a subcontract for construction management to Dick Corp. of Puerto Rico (Dick-PR), a subcontract for the operation and maintenance of the Project to Thames Water International, Ltd. (Thames Water), and a subcontract for the fabrication of pipe to Atlantic Pipe Corp. (APC). After the Project had been built, a segment of the pipeline burst. Thames-Dick incurred significant costs in repairing the damage. Not surprisingly, it sought to recover those costs from other parties. In response, one of PRASA’s insurers filed a declaratory judgment action in a local court to determine whether Thames- *139 Dick’s claims were covered under its policy. The litigation ballooned, soon involving a number of parties and a myriad of issues above and beyond insurance coverage. On April 25, 2001, the hostilities spilled over into federal court. Two entities beneficially interested in the master agreement — CPA Group International and Chiang, Patel & Yerby, Inc. (collectively CPA) — sued Thames-Dick, Dick-PR, Thames Water, and various insurers in the United States District Court for the District of Puerto Rico, seeking remuneration for consulting services rendered in connection with repairs to the Project. A googol of claims, counterclaims, cross-claims, and third-party complaints followed. Some of these were brought against APC (the petitioner here). To complicate matters, one of the defendants moved to dismiss on grounds that, inter alia, (1) CPA had failed to join an indispensable party whose presence would destroy diversity jurisdiction, and (2) the existence of the parallel proceeding in the local court counseled in favor of abstention. While this motion was pending before the district court, Thames-Dick asked that the case be referred to mediation and suggested Professor Eric Green as a suitable mediator. The district court granted the motion over APC’s objection and ordered non-binding mediation to proceed before Professor Green. The court pronounced mediation likely to conserve judicial resources; directed all parties to undertake mediation in good faith; stayed discovery pending completion of the mediation; and declared that participation in the mediation would not prejudice the parties’ positions vis-á-vis the pending motion or the litigation as a whole. The court also stated that if mediation failed to produce a global settlement, the case would proceed to trial. After moving unsuccessfully for reconsideration of the mediation order, APC sought relief by way of mandamus. Its petition alleged that the district court did not have the authority to require mediation (especially in light of unresolved questions as to the court’s subject-matter jurisdiction) and, in all events, could not force APC to pay a share of the expenses of the mediation. We invited the other parties and the district judge to respond. See Fed. RApp. P. 21(b)(4)-(5). Several entities (including Thames-Dick, Dick-P.R., and Thames Water) opposed the petition. Two others (third-party defendants United States Fidelity & Guaranty Company and United Surety and Indemnity Company) filed a brief in support of APC. We assigned the case to the oral argument calendar and stayed the contemplated mediation pending our review. Prior to argument in this court, two notable developments occurred. First, the district court considered and rejected the challenges to its exercise of jurisdiction. Second, APC rejected an offer by Thames-Dick to pay its share of the mediator’s fees. II. JURISDICTION In an effort to shut off further debate, the respondents asseverate that mandamus is improper because APC will not suffer irreparable harm in the absence of such relief. They rest this asseveration on the notion that “[mjandamus is ordinarily appropriate [only] in those rare cases in which the issuance (or nonissuance) of an order presents a question anent the limits of judicial power, poses some special risk of irreparable harm to the appellant, and is palpably erroneous.” United States v. Horn, 29 F.3d 754, 769 (1st Cir.1994). The problem, however, is that these limitations typically apply only to supervisory mandamus. Id. at 769 & n. *140 19. In the tiny class of cases in which advisory mandamus is appropriate, irreparable harm need not be shown. Id. at 769-70. We believe that this case is fit for advisory mandamus because the extent of a trial court’s power to order mandatory mediation presents a systemieally important issue as to which this court has not yet spoken. See In re Prov. Journal Co., 293 F.3d 1, 9 (1st Cir.2002) (discussing criteria for advisory mandamus). Moreover, that issue is capable of significant repetition prior to effective review. See Jennifer O’Hearne, Comment, Compelled Participation in Innovative Pretrial Proceedings, 84 Nw. U.L.Rev. 290, 317 (1989) (noting that, as a practical matter, lawyers often are unable to challenge pretrial innovations even when they may be invalid). That fact militates in favor of advisory mandamus. See Horn, 29 F.3d at 770. We conclude, therefore, that invoking advisory mandamus is prudent under the circumstances. Consequently, the existence vel non of irreparable harm is a non-issue. We turn, then, to the merits. III. THE MERITS There are four potential sources of judicial authority for ordering mandatory non-binding mediation of pending cases, namely, (a) the court’s local rules, (b) an applicable statute, (c) the Federal Rules of Civil Procedure, and (d) the court’s inherent powers. Because the district court did not identify the basis of its assumed authority, we consider each of these sources. A. The Local Rules. A district court’s local rules may provide an appropriate source of authority for ordering parties to participate in mediation. See Rhea v. Massey-Ferguson, Inc., 767 F.2d 266, 268-69 (6th Cir.1985) (per curiam). In Puerto Rico, however, the local rules contain only a single reference to any form of alternative dispute resolution (ADR). That reference is embodied in the district court’s Amended Civil Justice Expense and Delay Reduction Plan (CJR Plan). See D.P.R. R. app. III. The district court adopted the CJR Plan on June 14, 1993, in response to the directive contained in the Civil Justice Reform Act of 1990 (CJRA), 28 U.S.C. §§ 471-482. Rule V of the CJR Plan states: Pursuant to 28 U.S.C. § 473(b)(4), this Court shall adopt a method of Alternative Dispute Resolution (“ADR”) through mediation by a judicial officer. Such a program would allow litigants to obtain from an impartial third party — the judicial officer as mediator — a flexible non-binding, dispute resolution process to facilitate negotiations among the parties to help them reach settlement. D.P.R. R. app. III (R. V.). In addition to specifying who may act as a mediator, Rule V also limns the proper procedure for mediation sessions and assures confidentiality. See id. The respondents concede that the mediation order in this case falls outside the boundaries of the mediation program envisioned by Rule V. It does so most noticeably because it involves mediation before a private mediator, not a judicial officer. Seizing upon this discrepancy, APC argues that the local rules limit the district court in this respect, and that the court exceeded its authority thereunder by issuing a non-conforming mediation order (i.e., one that contemplates the intervention of a private mediator). The respondents counter by arguing that the rule does not bind the district court because, notwithstanding the unambiguous promise of the CJR Plan (which declares that the district court *141 “shall adopt a method of Alternative Dispute Resolution”), no such program has been adopted to date. This is a powerful argument. APC does not contradict the respondents’ assurance that the relevant portion of the CJR Plan has remained unimplemented, and we take judicial notice that there is no formal, ongoing ADR program in the Puerto Rico federal district court. Because that is so, we conclude that the District of Puerto Rico has no local rule in force that dictates the permissible characteristics of mediation orders. Consequently, APC’s argument founders. 2 B. The ADR Act. There is only one potential source of statutory authority for ordering mandatory non-binding mediation here: the Alternative Dispute Resolution Act of 1998 (ADR Act), 28 U.S.C. §§ 651-658. Congress passed the ADR Act to promote the utilization of alternative dispute resolution methods in the federal courts and to set appropriate guidelines for their use. The Act lists mediation as an appropriate ADR process. Id. § 651(a). Moreover, it sanctions the participation of “professional neutrals from the private sector” as mediators. Id. § 653(b). Finally, the Act requires district courts to obtain litigants’ consent only when they order arbitration, id. § 652(a), not when they order the use of other ADR mechanisms (such as nonbinding mediation). Despite the broad sweep of these provisions, the Act is quite clear that some form of the ADR procedures it endorses must be adopted in each judicial district by local rule. See id. § 651(b) (directing each district court to “devise and implement its own alternative dispute resolution program, by local rule adopted under [28 U.S.C.] section 2071(a), to encourage and promote the use of alternative dispute resolution in its district”). In the absence of such local rules, the ADR Act itself does not authorize any specific court to use a particular ADR mechanism. Because the District of Puerto Rico has not yet complied with the Act’s mandate, the mediation order here at issue cannot be justified under the ADR Act. The respondents essay an end run around this lacuna: they contend (borrowing a phrase from the court below) that the “spirit” of the ADR Act authorizes the mediation order because the Act was intended to promote experimentation with ADR techniques. We reject this attempt to press the ADR Act into service by indirection. Although the ADR Act was designed to promote the use of ADR techniques, Congress chose a very well-defined path: it granted each judicial district, rather than each individual judge, the authority to craft an appropriate ADR program. In other words, Congress permitted experimentation, but only within the disciplining format of district-wide local rules adopted with notice and a full opportunity for pub- *142 lie comment. See 28 U.S.C. § 2071(b). To say that the Act authorized each district judge to disregard a district-wide ADR plan (or the absence of one) and fashion innovative procedures for use in specific cases is simply too much of a stretch. We add, however, that although the respondents cannot use the ADR Act as a justification, neither can APC use it as a nullification. Noting that the Act requires the adoption of local rules establishing a formal ADR program, APC equates the absence of such rules with the absence of power to employ an ADR procedure (say, mediation) in a specific case. But that is wishful thinking: if one assumes that district judges possessed the power to require mediation prior to the passage of the ADR Act, there is nothing in the Act that strips them of that power. After all, even the adoption of a federal procedural rule does not implicitly abrogate a district court’s inherent power to act merely because the rule touches upon the same subject matter. See Chambers v. NASCO, Inc., 501 U.S. 32, 42-43, 111 S.Ct. 2123, 115 L.Ed.2d 27 (1991) (rejecting the argument that the adoption of various provisions of the Civil Rules eliminated the district court’s inherent power to impose other sanctions); Link v. Wabash R.R., 370 U.S. 626, 630, 82 S.Ct. 1386, 8 L.Ed.2d 734 (1963) (explaining that neither the permissive language of Fed.R.Civ.P. 41(b) nor the policy behind it justified a conclusion that it was meant to limit the district courts’ inherent power to dismiss a case for want of prosecution). The case before us is analogous to Chambers and Link. Even though Congress may cabin the district courts’ inherent powers, its intention to do so must be clear and unmistakable. See Weinberger v. Romero-Barcelo, 456 U.S. 305, 313, 102 S.Ct. 1798, 72 L.Ed.2d 91 (1982). Not so here: we know of nothing in either the ADR Act or the policies that undergird it that can be said to restrict the district courts’ authority to engage in the case-by-case deployment of ADR procedures. Hence, we conclude that where, as here, there are no implementing local rules, the ADR Act neither authorizes nor prohibits the entry of a mandatory mediation order. C. The Civil Rules. The respondents next argue that the district court possessed the authority to require mediation by virtue of the Federal Rules of Civil Procedure. They concentrate their attention on Fed.R.Civ.P. 16, which states in pertinent part that “the court may take appropriate actionf] with respect to ... (9) settlement and the use of special procedures to assist in resolving the dispute when authorized by statute or local rule....” Fed.R.Civ.P. 16(c)(9). But the words “when authorized by statute or local rule” are a frank limitation on the district courts’ authority to order mediation thereunder, 3 and we must adhere to that circumscription. See Schlagenhauf v. Holder, 379 U.S. 104, 121, 85 S.Ct. 234, 13 L.Ed.2d 152 (1964) (explaining that the Civil Rules “should not be expanded by disregarding plainly expressed limitations”). Because there is no statute or local rule authorizing mandatory private mediation in the District of Puerto Rico, *143 see supra Parts III(A)-(B), Rule 16(c)(9) does not assist the respondents’ cause. 4 D. Inherent Powers. Even apart from positive law, district courts have substantial inherent power to manage and control their calendars. See Link, 370 U.S. at 630-31, 82 S.Ct. 1386; see generally Brockton Sav. Bank v. Peat, Marwick, Mitchell & Co., 771 F.2d 5, 11 (1st Cir.1985) (explaining that “the rules of civil procedure do not completely describe and limit the power of district courts”). This inherent power takes many forms. See Fed.R.Civ.P. 83(b) (providing that judges may regulate practice in any manner consistent with federal law and applicable rules). By way of illustration, a district court may use its inherent power to compel represented clients to attend pretrial settlement conferences, even though such a practice is not specifically authorized in the Civil Rules. See Heileman Brewing Co. v. Joseph Oat Corp., 871 F.2d 648, 650 (7th Cir.1989) (en banc). Of course, a district court’s inherent powers are not infinite. There are at least four limiting principles. First, inherent powers must be used in a way reasonably suited to the enhancement of the court’s processes, including the orderly and expeditious disposition of pending cases. Coyante v. P.R. Ports Auth., 105 F.3d 17, 23 (1st Cir.1997). Second, inherent powers cannot be exercised in a manner that contradicts an applicable statute or rule. Chambers, 501 U.S. at 47, 111 S.Ct. 2123. Third, the use of inherent powers must comport with procedural fairness. Id. at 50, 111 S.Ct. 2123. And, finally, inherent powers “must be exercised with restraint and discretion.” Id. at 44, 111 S.Ct. 2123. At one time, the inherent power of judges to compel unwilling parties to participate in ADR procedures was a hot-button issue for legal scholars. Compare, e.g., O’Hearne, supra at 320 (arguing that inherent power should not be used to compel participation in pretrial settlement proceedings), with Lucille M. Ponte, Putting Mandatory Summary Jury Trial Back on the Docket: Recommendations on the Exercise of Judicial Authority, 63 Fordham L.Rev. 1069, 1094 (1995) (urging the opposite conclusion). Although many federal district courts have forestalled further debate by adopting local rules that authorize specific ADR procedures and outlaw others, e.g., D.N.H. R. 53.1 (permitting mandatory mediation); D. Me. R. 83.11 (permitting only voluntary mediation); D. Mass. R. 16.4 (permitting mandatory summary jury trials but only voluntary mediation), the District of Puerto Rico is not among them. Thus, we have no choice but to address the question head-on. We begin our inquiry by examining the case law. In Strandell v. Jackson County, 838 F.2d 884 (7th Cir.1987), the Seventh Circuit held that a district court does not possess inherent power to compel participation in a summary jury trial. 5 In the court’s view, Fed.R.Civ.P. 16 occupied the *144 field and prevented a district court from forcing “an unwilling litigant [to] be sidetracked from the normal course of litigation.” Id. at 887. But the group that spearheaded the subsequent revision of Rule 16 explicitly rejected that interpretation. See Fed.R.Civ.P. 16, advisory committee’s note (1993 Amendment) (“The [amended] rule does not attempt to resolve questions as to the extent a court would be authorized to require [ADR] proceedings as an exercise of its inherent powers.”). Thus, we do not find Strandell persuasive on this point. The Strandell court also expressed concern that summary jury trials would undermine traditional discovery and privilege rules by requiring certain disclosures prior to an actual trial. 838 F.2d at 888. We find this concern unwarranted. Because a summary jury trial (like a non-binding mediation) does not require any disclosures beyond what would be required in the ordinary course of discovery, its principal disadvantage to the litigants is that it may prevent them from saving surprises for the time of trial. Since trial by ambush is no longer in vogue, that interest does not deserve protection. See Fed. Reserve Bank v. Carey-Canada, Inc., 123 F.R.D. 603, 606 (D.Minn.1988). Relying on policy arguments, the Sixth Circuit also has found that district courts do not possess inherent power to compel participation in summary jury trials. See In re NLO, Inc., 5 F.3d 154, 157-58 (6th Cir.1993). The court thought the value of a summary jury trial questionable when parties do not engage in the process voluntarily, and it worried that “too broad an interpretation of the federal courts’ inherent power to regulate their procedure ... encourages judicial high-handedness.... ” Id. at 158 (citation and internal quotation marks omitted). The concerns articulated by these two respected courts plainly apply to mandatory mediation orders. When mediation is forced upon unwilling litigants, it stands to reason that the likelihood of settlement is diminished. Requiring parties to invest substantial amounts of time and money in mediation under such circumstances may well be inefficient. Cf. Richard A. Posner, The Summary Jury Trial and Other Methods of Alternative Dispute Resolution: Some Cautionary Observations, 53 U. Chi. L.Rev. 366, 369-72 (1986) (offering a model to evaluate ADR techniques in terms of their capacity to encourage settlements). The fact remains, however, that none of these considerations establishes that mandatory mediation is always inappropriate. There may well be specific cases in which such a protocol is likely to conserve judicial resources without significantly burdening the objectors’ rights to a full, fair, and speedy trial. Much depends on the idio-syncraeies of the particular case and the details of the mediation order. In some cases, a court may be warranted in believing that compulsory mediation could yield significant benefits even if one or more parties object. After all, a party may resist mediation simply out of unfamiliarity with the process or out of fear that a willingness to submit would be perceived as a lack of confidence in her legal position. See Campbell C. Hutchinson, The Case for Mandatory Mediation, 42 Loy. L.Rev. 85, 89-90 (1996). In such an instance, the party’s initial reservations are likely to evaporate as the mediation progresses, and negotiations could well produce a beneficial outcome, at reduced cost and greater speed, than would a trial. While the possibility that parties will fail to reach agreement remains ever present, the boon of settlement can be worth the risk. *145 This is particularly true in complex cases involving multiple claims and parties. The fair and expeditious resolution of such cases often is helped along by creative solutions — solutions that simply are not available in the binary framework of traditional adversarial litigation. Mediation with the assistance of a skilled facilitator gives parties an opportunity to explore a much wider range of options, including those that go beyond conventional zero-sum resolutions. Mindful of these potential advantages, we hold that it is within a district court’s inherent power to order non-consensual mediation in those cases in which that step seems reasonably likely to serve the interests of justice. Cf. Reilly v. United States, 863 F.2d 149, 156-57 (1st Cir.1988) (finding that district courts have inherent power to appoint technical advisors in especially complex cases). E. The Mediation Order. Our determination that the district courts have inherent power to refer cases to non-binding mediation is made with a recognition that any such order must be crafted in a manner that preserves procedural fairness and shields objecting parties from undue burdens. We thus turn to the specifics of the mediation order entered in this case. As with any exercise of a district court’s inherent powers, we review the entry of that order for abuse of discretion. See Chambers, 501 U.S. at 50, 111 S.Ct. 2123; Reilly, 863 F.2d at 156. As an initial matter, we agree with the lower court that the complexity of this case militates in favor of ordering mediation. At last count, the suit involves twelve parties, asserting a welter of claims, counterclaims, cross-claims, and third-party claims predicated on a wide variety of theories. The pendency of nearly parallel litigation in the Puerto Rican courts, which features a slightly different cast of characters and claims that are related to but not completely congruent with those asserted here, further complicates the matter. Untangling the intricate web of relationships among the parties, along with the difficult and fact-intensive arguments made by each, will be time-consuming and will impose significant costs on the parties and the court. Against this backdrop, mediation holds out the dual prospect of advantaging the litigants and conserving scarce judicial resources. In an effort to parry this thrust, APC raises a series of objections. Its threshold claim is that the district court erred in ordering mediation before resolving a pending motion to dismiss for lack of subject-matter jurisdiction (or, alternatively, to abstain). See, e.g., Bouchard Transp. Co. v. Fla. Dep’t of Envtl. Prot., 91 F.3d 1445, 1448-19 (11th Cir.1996) (vacating a mediation order and directing the lower court first to consider the objecting party’s assertion of Eleventh Amendment immunity). Given what has transpired, this argument is fruitless. While this proceeding was pending, the district court denied the motion in question and confirmed the existence of its subject-matter jurisdiction. See CPA Group Int’l, Inc. v. Am. Int’l Ins. Co., No. 01-1483, slip op. at 16-25 (D.P.R. May 23, 2002). Thus, even if it were error to enter the mediation order before passing upon the motion to dismiss, 6 the error *146 was harmless: it would be an empty exercise to vacate the mediation order on this ground when the lower court has already rejected the challenges to its exercise of jurisdiction. See, e.g., Gibbs v. Buck, 307 U.S. 66, 78, 59 S.Ct. 725, 83 L.Ed. 1111 (1939); Aoude v. Mobil Oil Corp., 862 F.2d 890, 895 (1st Cir.1988). Next, APC posits that the appointment of a private mediator proposed by one of the parties is per se improper (and, thus, invalidates the order). We do not agree. The district court has inherent power to “appoint persons unconnected with the court to aid judges in the performance of specific judicial duties.” Ex parte Peterson, 253 U.S. 300, 312, 40 S.Ct. 543, 64 L.Ed. 919 (1920). In the context of non-binding mediation, the mediator does not decide the merits of the case and has no authority to coerce settlement. Thus, in the absence of a contrary statute or rule, it is perfectly acceptable for the district court to appoint a qualified and neutral private party as a mediator. The mere fact that the mediator was proposed by one of the parties is insufficient to establish bias in favor of that party. Cf. TechSearch, L.L.C. v. Intel Corp., 286 F.3d 1360, 1379 n. 3 (Fed.Cir.2002) (noting that technical advisors typically would be selected from a list of candidates submitted by the parties). We hasten to add that the litigants are free to challenge the qualifications or neutrality of any suggested mediator (whether or not nominated by a party to the case). APC, for example, had a full opportunity to present its views about the suggested mediator both in its opposition to the motion for mediation and in its motion for reconsideration of the mediation order. Despite these opportunities, APC offered no convincing reason to spark a belief that Professor Green, a nationally recognized mediator with significant experience in sprawling cases, is an unacceptable choice. When a court enters a mediation order, it necessarily makes an independent determination that the mediator it appoints is both qualified and neutral. Because the court made that implicit determination here in a manner that was procedurally fair (if not ideal), we find no abuse of discretion in its selection of Professor Green. 7 APC also grouses that it should not be forced to share the costs of an unwanted mediation. We have held, however, that courts have the power under Fed.R.Civ.P. 26(f) to issue pretrial cost-sharing orders in complex litigation. See In re San Juan Dupont Plaza Hotel Fire Litig., 994 F.2d 956, 965 (1st Cir.1993). Given the difficulties facing trial courts in cases involving multiple parties and multiple claims, we are hesitant to limit that power to the traditional discovery context. See id. This is especially true in complicated cases, where the potential value of mediation lies not only in promoting settlement but also in clarifying the issues remaining for trial. The short of the matter is that, without default cost-sharing rules, the use of valuable ADR techniques (like mediation) becomes hostage to the parties’ ability to agree on the concomitant financial *147 arrangements. This means that the district court’s inherent power to order private mediation in appropriate cases would be rendered nugatory absent the corollary power to order the sharing of reasonable mediation costs. To avoid this pitfall, we hold that the district court, in an appropriate case, is empowered to order the sharing of reasonable costs and expenses associated with mandatory non-binding mediation. The remainder of APC’s arguments are not so easily dispatched. Even when generically appropriate, a mediation order must contain procedural and substantive safeguards to ensure fairness to all parties involved. The mediation order in this case does not quite meet that test. In particular, the order does not set limits on the duration of the mediation or the expense associated therewith. 8 We need not wax longiloquent. As entered, the order simply requires the parties to mediate; it does not set forth either a timetable for the mediation or a cap on the fees that the mediator may charge. The figures that have been bandied about in the briefs — $900 per hour or $9,000 per mediation day — are quite large and should not be left to the mediator’s whim. Relat-edly, because the mediator is to be paid an hourly rate, the court should have set an outside limit on the number of hours to be devoted to mediation. Equally as important, it is trite but often true that justice delayed is justice denied. An unsuccessful mediation will postpone the ultimate resolution of the case — indeed, the district court has stayed all discovery pending the completion of the mediation — and, thus, prolong the litigation. For these reasons, the district court should have set a definite time frame for the mediation. The respondents suggest that the district court did not need to articulate any limitations in its mediation order because the mediation process will remain under the district court’s ultimate supervision; the court retains the ability to curtail any excessive expenditures of time or money; and a dissatisfied party can easily return to the court at any time. While this might be enough of a safeguard in many instances, the instant litigation is sufficiently complicated and the mediation efforts are likely to be sufficiently expensive that, here, reasonable time limits and fee constraints, set in advance, are appropriate. 9 A court intent on ordering non-consensual mediation should take other precautions as well. For example, the court should make it clear (as did the able district court in this case) that participation in mediation will not be taken as a waiver of any litigation position. The important point is that the protections we have mentioned are not intended to comprise an exhaustive list, but, rather, to illustrate that when a district court orders a party to participate in mediation, it should take care to assuage legitimate concerns about the possible negative consequences of such an order. To recapitulate, we rule that a mandatory mediation order issued under the district court’s inherent power is valid in an appropriate case. We also rule that this is an appropriate case. We hold, however, that the district court’s failure to set rea *148 sonable limits on the duration of the mediation and on the mediator’s fees dooms the decree. IV. CONCLUSION We admire the district court’s pragmatic and innovative approach to this massive litigation. Our core holding — that ordering mandatory mediation is a proper exercise of a district court’s inherent power, subject, however, to a variety of terms and conditions — validates that approach. We are mindful that this holding is in tension with the opinions of the Sixth and Seventh Circuits in NLO and Strandell, respectively, but we believe it is justified by the important goal of promoting flexibility and creative problem-solving in the handling of complex litigation. That said, the need of the district judge in this case to construct his own mediation regime ad hoc underscores the greater need of the district court as an institution to adopt an ADR program and memorialize it in its local rules. In the ADR Act, Congress directed that “[ejach United States district court shall authorize, by local rule under section 2071(a), the use of alternative dispute resolution processes in all civil actions....” 28 U.S.C. § 651(b). While Congress did not set a firm deadline for compliance with this directive, the statute was enacted four years ago. This omission having been noted, we are confident that the district court will move expediently to bring the District of Puerto Rico into compliance. We need go no further. For the reasons set forth above, we vacate the district court’s mediation order and remand for further proceedings consistent with this opinion. The district court is free to order mediation if it continues to believe that such a course is advisable or, in the alternative, to proceed with discovery and trial. Vacated and remanded. Costs shall be taxed in favor of the petitioner. Footnotes [1] Although the petition seeks the issuance of a writ of mandamus, the relief sought is more in the nature of a writ of prohibition. Because the two writs have much in common— one is merely the obverse of the other — we follow past practice and make no distinction between them. See, e.g., United States v. Horn, 29 F.3d 754, 769 n. 18 (1st Cir.1994). [2] This holding renders it unnecessary for us to discuss the respondents' alternate contention that the CJR Plan is a dead letter because the legislation that prompted its enactment— the CJRA — expired in 1997. See, e.g., Carl Tobias, Did the Civil Justice Reform Act of 1990 Actually Expire?, 31 U. Mich. J.L. Reform 887, 892 (1998) (exploring the uncertainty regarding whether the CJRA expired and whether local plans adopted pursuant to it are still effective). By like token, this holding renders moot the respondents' claim that the district court could disregard Appendix III even if an ADR program were in force. See D.P.R. R. 105 ("When a judge of this court issues any order in a specific case which is not consistent with these rules, such order shall constitute a suspension of these rules for such case and only to the extent that it is inconsistent therewith.''). Accordingly, we take no view of the meaning or validity of Local Rule 105. [3] We think it is pertinent here to quote the advisory committee’s note: The rule acknowledges the presence of statutes and local rules or plans that may authorize use of some [ADR] procedures even when not agreed to by the parties. The rule does not attempt to resolve questions as to the extent a court would be authorized to require such proceedings as an exercise of its inherent powers. Fed.R.Civ.P. 16, advisory committee’s note (1993 Amendment) (citations omitted). [4] The cases that the respondents cite for the proposition that Rule 16 sanctions mandatory ADR procedures even in the absence of an enabling statute or local rule are inapposite. All of them predate the 1993 amendments, which added the pertinent language to Rule 16. See, e.g., Fed. Reserve Bank v. Carey-Canada, Inc., 123 F.R.D. 603, 606 (D.Minn.1988); Arabian Am. Oil Co. v. Scarfone, 119 F.R.D. 448, 448-49 (M.D.Fla.1988). [5] A summary jury trial is an ADR technique in which the opposing attorneys present their case, in abbreviated form, to a mock jury, which proceeds to render a non-binding verdict. See In re NLO, Inc., 5 F.3d 154, 156 (6th Cir.1993); see generally Thomas D. Lam-bros, The Summary Jury Trial Report to the Judicial Conference of the United States, 103 F.R.D. 461 (1984). [6] Although we take no view of this aspect of APC’s argument, we note that when a jurisdictional question involves disputed facts, the district court sometimes "may defer resolution of the jurisdictional issue until the time of trial.” Valentin v. Hosp. Bella Vista, 254 F.3d 358, 363 n. 3 (1st Cir.2001). How this proposition affects a trial court's ability to issue *146 interim orders designed to move litigation forward while it ponders a challenge to its jurisdiction is a matter that we leave for another day. [7] We say "not ideal” because, in an ideal world, it would be preferable for the district court, before naming a mediator, to solicit the names of potential nominees from all parties and to provide an opportunity for the parties to comment upon each others' proposed nominees. [8] We do not assign significant weight to Thames-Dick’s belated offer to pay APC’s share of the mediator’s fee. There are other expenses involved, and there is too much of a risk that “free rider” status will itself breed problems. [9] We do not mean that a mediation order in such a case must be etched in stone. The mediator and the parties remain free, for good cause shown, to ask the district court to extend or modify the original order.