MOHAWK INDUSTRIES, INC. v. CARPENTER

MOHAWK INDUSTRIES, INC. v. CARPENTER No. 08-678. Argued October 5, 2009 — Decided December 8, 2009 Sotomayor, J., delivered the opinion of the Court, in which Roberts, C. J., and Stevens, Scalia, Kennedy, Ginsburg, Breyer, and Alito, JJ., joined, and in which Thomas, J., joined, as to Part II-C. Thomas, J., filed an opinion concurring in part and concurring in the judgment, post, p. 114. Randall L. Allen argued the cause for petitioner. With him on the briefs was Daniel F. Diffley. Judith Resnik argued the cause for respondent. With her on the brief were J. Craig Smith, Dennis E. Curtis, Thomas J. Munger, Alan B. Morrison, Deepak Gupta, Brian Wolf-man, and Sean K. McElligott. Deputy Solicitor General Kneedler argued the cause for the United States as amicus curiae in support of respondent. With him on the brief were Solicitor General Kagan, Assistant Attorney General West, Pratik A. Shah, and Michael S. Raab. Briefs of amici curiae urging reversal were filed for the American Bar Association by H. Thomas Wells, Jr., and Paul Mogin; for the Chamber of Commerce of the United States of America by Paul D. Clement, Jeffrey S. Bucholtz, Robin S. Conrad, and Amar D. Sarwal; and for DRI-The Voice of the Defense Bar by Constantine L. Trela, Jr., and Quin M. Sorenson. Stephen I. Vladeck, Charles S. Sims, Mark D. Harris, and Anna G. Kaminska filed a brief for Former Article III Judges et al. as amici curiae urging affirmance. [MAJORITY — Justice Sotomayor] Justice Sotomayor delivered the opinion of the Court. Section 1291 of the Judicial Code confers on federal courts of appeals jurisdiction to review “final decisions of the district courts.” 28 U. S. C. § 1291. Although “final decisions” typically are ones that trigger the entry of judgment, they also include a small set of prejudgment orders that are “collateral to” the merits of an action and “too important” to be denied immediate review. Cohen v. Beneficial Industrial Loan Corp., 337 U. S. 541, 546 (1949). In this case, petitioner Mohawk Industries, Inc., attempted to bring a collateral order appeal after the District Court ordered it to disclose certain confidential materials on the ground that Mohawk had waived the attorney-client privilege. The Court of Appeals dismissed the appeal for want of jurisdiction. The question before us is whether disclosure orders adverse to the attorney-client privilege qualify for immediate appeal under the collateral order doctrine. Agreeing with the Court of Appeals, we hold that they do not. Postjudgment appeals, together with other review mechanisms, suffice to protect the rights of litigants and preserve the vitality of the attorney-client privilege. I In 2007, respondent Norman Carpenter, a former shift supervisor at a Mohawk manufacturing facility, filed suit in the United States District Court for the Northern District of Georgia, alleging that Mohawk had terminated him in violation of 42 U. S. C. § 1985(2) and various Georgia laws. According to Carpenter’s complaint, his termination came after he informed a member of Mohawk’s human resources department in an e-mail that the company was employing undocumented immigrants. At the time, unbeknownst to Carpenter, Mohawk stood accused in a pending class-action lawsuit of conspiring to drive down the wages of its legal employees by knowingly hiring undocumented workers in violation of federal and state racketeering laws. See Wil liams v. Mohawk Indus., Inc., No. 4:04-cv-00003-HLM (ND Ga., Jan. 6, 2004). Company officials directed Carpenter to meet with the company’s retained counsel in the Williams case, and counsel allegedly pressured Carpenter to recant his statements. When he refused, Carpenter alleges, Mohawk fired him under false pretenses. App. 57a-64a. After learning of Carpenter’s complaint, the plaintiffs in the Williams case sought an evidentiary hearing to explore Carpenter’s allegations. In its response to their motion, Mohawk described Carpenter’s accusations as “pure fantasy” and recounted the “true facts” of Carpenter’s dismissal. App. 208a. According to Mohawk, Carpenter himself had “engaged in blatant and illegal misconduct” by attempting to have Mohawk hire an undocumented worker. Id., at 209a. The company “commenced an immediate investigation,” during which retained counsel interviewed Carpenter. Id., at 210a. Because Carpenter’s “efforts to cause Mohawk to circumvent federal immigration law” “blatantly violated Mohawk policy,” the company terminated him. Ibid. As these events were unfolding in the Williams case, discovery was underway in Carpenter’s case. Carpenter filed a motion to compel Mohawk to produce information concerning his meeting with retained counsel and the company’s termination decision. Mohawk maintained that the requested information was protected by the attorney-client privilege. The District Court agreed that the privilege applied to the requested information, but it granted Carpenter’s motion to compel disclosure after concluding that Mohawk had implicitly waived the privilege through its representations in the Williams case. See App. to Pet. for Cert. 51a. The court declined to certify its order for interlocutory appeal under 28 U. S. C. § 1292(b). But, recognizing “the seriousness of its [waiver] finding,” it stayed its ruling to allow Mohawk to explore other potential “avenues to appeal . . . , such as a petition for mandamus or appealing this Order under the collateral order doctrine.” App. to Pet. for Cert. 52a. Mohawk filed a notice of appeal and a petition for a writ of mandamus to the Eleventh Circuit. The Court of Appeals dismissed the appeal for lack of jurisdiction under 28 U. S. C. § 1291, holding that the District Court’s ruling did not qualify as an immediately appealable collateral order within the meaning of Cohen, 337 U. S. 541. “Under Cohen,” the Court of Appeals explained, “an order is appealable if it (1) conclusively determines the disputed question; (2) resolves an important issue completely separate from the merits of the action; and (3) is effectively unreviewable on appeal from a final judgment.” 541 F. 3d 1048, 1052 (2008) (per curiam). According to the court, the District Court’s waiver ruling satisfied the first two of these requirements but not the third, because “a discovery order that implicates the attorney-client privilege” can be adequately reviewed “on appeal from a final judgment.” Ibid. The Court of Appeals also rejected Mohawk’s mandamus petition, finding no “clear usurpation of power or abuse of discretion” by the District Court. Id., at 1055. We granted certiorari, 555 U. S. 1152 (2009), to resolve a conflict among the Circuits concerning the availability of collateral appeals in the attorney-client privilege context. II A By statute, courts of appeals “have jurisdiction of appeals from all final decisions of the district courts of the United States, . . . except where a direct review may be had in the Supreme Court.” 28 U. S. C. § 1291. A “final decisio[n]” is typically one “by which a district court disassociates itself from a case.” Swint v. Chambers County Comm’n, 514 U. S. 35, 42 (1995). This Court, however, “has long given” § 1291 a “practical rather than a technical construction.” Cohen, 337 U. S., at 546. As we held in Cohen, the statute encompasses not only judgments that “terminate an action,” but also a “small class” of collateral rulings that, although they do not end the litigation, are appropriately deemed “final.” Id., at 545-546. “That small category includes only decisions that are conclusive, that resolve important questions separate from the merits, and that are effectively unreviewable on appeal from the final judgment in the underlying action.” Swint, 514 U. S., at 42. In applying Cohen’s collateral order doctrine, we have stressed that it must “never be allowed to swallow the general rule that a party is entitled to a single appeal, to be deferred until final judgment has been entered.” Digital Equipment Corp. v. Desktop Direct, Inc., 511 U. S. 863, 868 (1994) (citation omitted); see also Will v. Hallock, 546 U. S. 345, 350 (2006) (“emphasizing [the doctrine’s] modest scope”). Our admonition reflects a healthy respect for the virtues of the final-judgment rule. Permitting piecemeal, prejudgment appeals, we have recognized, undermines “efficient judicial administration” and encroaches upon the prerogatives of district court judges, who play a “special role” in managing ongoing litigation. Firestone Tire & Rubber Co. v. Risjord, 449 U. S. 368, 374 (1981); see also Richardson-Merrell Inc. v. Koller, 472 U. S. 424, 436 (1985) (“[T]he district judge can better exercise [his or her] responsibility [to police the prejudgment tactics of litigants] if the appellate courts do not repeatedly intervene to second-guess prejudgment rulings”). The justification for immediate appeal must therefore be sufficiently strong to overcome the usual benefits of deferring appeal until litigation concludes. This requirement finds expression in two of the three traditional Cohen conditions. The second condition insists upon “important questions separate from the merits.” Swint, 514 U. S., at 42 (emphasis added). More significantly, “the third Cohen question, whether a right is ‘adequately vindicable’ or ‘effectively reviewable,’ simply cannot be answered without a judgment about the value of the interests that would be lost through rigorous application of a final judgment requirement.” Digital Equipment, 511 U. S., at 878-879. That a ruling “may burden litigants in ways that are only imperfectly reparable by appellate reversal of a final district court judgment... has never sufficed.” Id., at 872. Instead, the decisive consideration is whether delaying review until the entry of final judgment “would imperil a substantial public interest” or “some particular value of a high order.” Will, 546 U. S., at 352-353. In making this determination, we do not engage in an “individualized jurisdictional inquiry.” Coopers & Lybrand v. Livesay, 437 U. S. 463, 473 (1978). Rather, our focus is on “the entire category to which a claim belongs.” Digital Equipment, 511 U. S., at 868. As long as the class of claims, taken as a whole, can be adequately vindicated by other means, “the chance that the litigation at hand might be speeded, or a ‘particular injustic[e]’ averted,” does not provide a basis for jurisdiction under § 1291. Ibid, (quoting Van Cauwenberghe v. Biard, 486 U. S. 517, 529 (1988); alteration in original). B In the present case, the Court of Appeals concluded that the District Court’s privilege-waiver order satisfied the first two conditions of the collateral order doctrine — conclusiveness and separateness — but not the third — effective unreviewability. Because we agree with the Court of Appeals that collateral order appeals are not necessary to ensure effective review of orders adverse to the attorney-client privilege, we do not decide whether the other Cohen requirements are met. Mohawk does not dispute that “we have generally denied review of pretrial discovery orders.” Firestone, 449 U. S., at 377; see also 15B C. Wright, A. Miller, & E. Cooper, Federal Practice and Procedure §3914.23, p. 123 (2d ed. 1992) (hereinafter Wright & Miller) (“[T]he rule remains settled that most discovery rulings are not final”). Mohawk contends, however, that rulings implicating the attorney-client privilege differ in kind from run-of-the-mill discovery orders because of the important institutional interests at stake. According to Mohawk, the right to maintain attorney-client confidences — the sine qua non of a meaningful attorney-client relationship — is “irreparably destroyed absent immediate appeal” of adverse privilege rulings. Brief for Petitioner 23. We readily acknowledge the importance of the attorney-client privilege, which “is one of the oldest recognized privileges for confidential communications.” Swidler & Berlin v. United States, 524 U. S. 399, 403 (1998). By assuring confidentiality, the privilege encourages clients to make “full and frank” disclosures to their attorneys, who are then better able to provide candid advice and effective representation. Upjohn Co. v. United States, 449 U. S. 383, 389 (1981). This, in turn, serves “broader public interests in the observance of law and administration of justice.” Ibid. The crucial question, however, is not whether an interest is important in the abstract; it is whether deferring review until final judgment so imperils the interest as to justify the cost of allowing immediate appeal of the entire class of relevant orders. We routinely require litigants to wait until after final judgment to vindicate valuable rights, including rights central to our adversarial system. See, e.g., Richardson-Merrell, 472 U. S., at 426 (holding an order disqualifying counsel in a civil case did not qualify for immediate appeal under the collateral order doctrine); Flanagan v. United States, 465 U. S. 259, 260 (1984) (reaching the same result in a criminal case, notwithstanding the Sixth Amendment rights at stake). In Digital Equipment, we rejected an assertion that collateral order review was necessary to promote “the public policy favoring voluntary resolution of disputes.” 511 U. S., at 881. “It defies common sense,” we explained, “to maintain that parties’ readiness to settle will be significantly dampened (or the corresponding public interest impaired) by a rule that a district court’s decision to let allegedly barred litigation go forward may be challenged as a matter of right only on appeal from a judgment for the plaintiff’s favor.” Ibid. We reach a similar conclusion here. In our estimation, postjudgment appeals generally suffice to protect the rights of litigants and ensure the vitality of the attorney-client privilege. Appellate courts can remedy the improper disclosure of privileged material in the same way they remedy a host of other erroneous evidentiary rulings: by vacating an adverse judgment and remanding for a new trial in which the protected material and its fruits are excluded from evidence. Dismissing such relief as inadequate, Mohawk emphasizes that the attorney-client privilege does not merely “prohibi[t] use of protected information at trial”; it provides a “right not to disclose the privileged information in the first place.” Brief for Petitioner 25. Mohawk is undoubtedly correct that an order to disclose privileged information intrudes on the confidentiality of attorney-client communications. But deferring review until final judgment does not meaningfully reduce the ex ante incentives for full and frank consultations between clients and counsel. One reason for the lack of a discernible chill is that, in deciding how freely to speak, clients and counsel are unlikely to focus on the remote prospect of an erroneous disclosure order, let alone on the timing of a possible appeal. Whether or not immediate collateral order appeals are available, clients and counsel must account for the possibility that they will later be required by law to disclose their communications for a variety of reasons — for example, because they misjudged the scope of the privilege, because they waived the privilege, or because their communications fell within the privilege’s crime-fraud exception. Most district court rulings on these matters involve the routine application of settled legal principles. They are unlikely to be reversed on appeal, particularly when they rest on factual determinations for which appellate deference is the norm. See, e. g., Richardson-Merrell, 472 U. S., at 434 (“Most pretrial orders of district judges are ultimately affirmed by appellate courts”); Reise v. Board of Regents, 957 F. 2d 293, 295 (CA7 1992) (noting that “almost all interlocutory appeals from discovery orders would end in affirmance” because “the district court possesses discretion, and review is deferential”). The breadth of the privilege and the narrowness of its exceptions will thus tend to exert a much greater influence on the conduct of clients and counsel than the small risk that the law will be misapplied. Moreover, were attorneys and clients to reflect upon their appellate options, they would find that litigants confronted with a particularly injurious or novel privilege ruling have several potential avenues of review apart from collateral order appeal. First, a party may ask the district court to certify, and the court of appeals to accept, an interlocutory appeal pursuant to 28 U. S. C. § 1292(b). The preconditions for § 1292(b) review — “a controlling question of law,” the prompt resolution of which “may materially advance the ultimate termination of the litigation” — are most likely to be satisfied when a privilege ruling involves a new legal question or is of special consequence, and district courts should not hesitate to certify an interlocutory appeal in such cases. Second, in extraordinary circumstances — i. e., when a disclosure order “amount[s] to a judicial usurpation of power or a clear abuse of discretion,” or otherwise works a manifest injustice — a party may petition the court of appeals for a writ of mandamus. Cheney v. United States Dist. Court for D. C., 542 U. S. 367, 390 (2004) (citation and internal quotation marks omitted); see also Firestone, 449 U. S., at 378-379, n. 13. While these discretionary review mechanisms do not provide relief in every case, they serve as useful “safety valve[s]” for promptly correcting serious errors. Digital Equipment, 511 U. S., at 883. Another long-recognized option is for a party to defy a disclosure order and incur court-imposed sanctions. District courts have a range of sanctions from which to choose, including “directing that the matters embraced in the order or other designated facts be taken as established for purposes of the action,” “prohibiting the disobedient party from supporting or opposing designated claims or defenses,” or “striking pleadings in whole or in part.” Fed. Rules Civ. Proc. 37(b)(2)(A)(i)-(iii). Such sanctions allow a party to obtain post judgment review without having to reveal its privileged information. Alternatively, when the circumstances warrant it, a district court may hold a noncomplying party in contempt. The party can then appeal directly from that ruling, at least when the contempt citation can be characterized as a criminal punishment. See, e. g., Church of Scientology of Cal. v. United States, 506 U. S. 9, 18, n. 11 (1992); Firestone, 449 U. S., at 377; Cobbledick v. United States, 309 U. S. 323, 328 (1940); see also Wright & Miller §3914.23, at 140-155. These established mechanisms for appellate review not only provide assurances to clients and counsel about the security of their confidential communications; they also go a long way toward addressing Mohawk’s concern that, absent collateral order appeals of adverse attorney-client privilege rulings, some litigants may experience severe hardship. Mohawk is no doubt right that an order to disclose privileged material may, in some situations, have implications beyond the case at hand. But the same can be said about many categories of pretrial discovery orders for which collateral order appeals are unavailable. As with these other orders, rulings adverse to the privilege vary in their significance; some may be momentous, but others are more mundane. Section 1292(b) appeals, mandamus, and appeals from contempt citations facilitate immediate review of some of the more consequential attorney-client privilege rulings. Moreover, protective orders are available to limit the spillover effects of disclosing sensitive information. That a fraction of orders adverse to the attorney-client privilege may nevertheless harm individual litigants in ways that are “only imperfectly reparable” does not justify making all such orders immediately appealable as of right under §1291. Digital Equipment, 511 U. S., at 872. In short, the limited benefits of applying “the blunt, categorical instrument of §1291 collateral order appeal” to privilege-related disclosure orders simply cannot justify the likely institutional costs. Id., at 883. Permitting parties to undertake successive, piecemeal appeals of all adverse attorney-client rulings would unduly delay the resolution of district court litigation and needlessly burden the courts of appeals. See Wright & Miller §3914.23, at 123 (“Routine appeal from disputed discovery orders would disrupt the orderly progress of the litigation, swamp the courts of appeals, and substantially reduce the district court’s ability to control the discovery process”); cf. Cunningham v. Hamilton County, 527 U. S. 198, 209 (1999) (expressing concern that allowing immediate appeal as of right from orders fining attorneys for discovery violations would result in “the very sorts of piecemeal appeals and concomitant delays that the final judgment rule was designed to prevent”). Attempting to downplay such concerns, Mohawk asserts that the three Circuits in which the collateral order doctrine currently applies to adverse privilege rulings have seen only a trickle of appeals. But this may be due to the fact that the practice in all three Circuits is relatively new and not yet widely known. Were this Court to approve collateral order appeals in the attorney-client privilege context, many more litigants would likely choose that route. They would also likely seek to extend such a ruling to disclosure orders implicating many other categories of sensitive information, raising an array of line-drawing difficulties. C In concluding that sufficiently effective review of adverse attorney-client privilege rulings can be had without resort to the Cohen doctrine, we reiterate that the class of collaterally appealable orders must remain “narrow and selective in its membership.” Will, 546 U. S., at 350. This admonition has acquired special force in recent years with the enactment of legislation designating rulemaking, “not expansion by court decision,” as the preferred means for determining whether and when prejudgment orders should be immediately appealable. Swint, 514 U. S., at 48. Specifically, Congress in 1990 amended the Rules Enabling Act, 28 U. S. C. § 2071 et seq., to authorize this Court to adopt rules “defin[ing] when a ruling of a district court is final for the purposes of appeal under section 1291.” § 2072(c). Shortly thereafter, and along similar lines, Congress empowered this Court to “prescribe rules, in accordance with [§ 2072], to provide for an appeal of an interlocutory decision to the courts of appeals that is not otherwise provided for under [§ 1292].” § 1292(e). These provisions, we have recognized, “warran[t] the Judiciary’s full respect.” Swint, 514 U. S., at 48; see also Cunningham, 527 U. S., at 210. Indeed, the rulemaking process has important virtues. It draws on the collective experience of bench and bar, see 28 U. S. C. § 2073, and it facilitates the adoption of measured, practical solutions. We expect that the combination of standard postjudgment appeals, § 1292(b) appeals, mandamus, and contempt appeals will continue to provide adequate protection to litigants ordered to disclose materials purportedly subject to the attorney-client privilege. Any further avenue for immediate appeal of such rulings should be furnished, if at all, through rulemaking, with the opportunity for full airing it provides. * * * In sum, we conclude that the collateral order doctrine does not extend to disclosure orders adverse to the attorney-client privilege. Effective appellate review can be had by other means. Accordingly, we affirm the judgment of the Court of Appeals for the Eleventh Circuit. It is so ordered. Three Circuits have permitted collateral order appeals of attorney-client privilege rulings. See In re Napster, Inc. Copyright Litigation, 479 F. 3d 1078, 1087-1088 (CA9 2007); United States v. Philip Morris Inc., 314 F. 3d 612, 617-621 (CADC 2003); In re Ford Motor Co., 110 F. 3d 954, 957-964 (CA3 1997). The remaining Circuits to consider the question have found such orders nonappealable. See, e.g., Boughton v. Cotter Corp., 10 F. 3d 746, 749-750 (CA10 1993); Texaco Inc. v. Louisiana Land & Exploration Co., 995 F 2d 43, 44 (CA5 1993); Reise v. Board of Regents of Univ. of Wisconsin System, 957 F. 2d 293, 295 (CA7 1992); Chase Manhattan Bank, N. A. v. Turner & Newall, PLC, 964 F. 2d 159, 162-163 (CA2 1992); Quantum Corp. v. Tandon Corp., 940 F. 2d 642, 643-644 (CA Fed. 1991). Perhaps the situation would be different if district courts were systematically underenforeing the privilege, but we have no indication that this is the case. Mohawk itself petitioned the Eleventh Circuit for a writ of mandamus. See supra, at 105. It has not asked us to review the Court of Appeals’ denial of that relief. Participating as amicus curiae in support of respondent Carpenter, the United States contends that collateral order appeals should be available for rulings involving certain governmental privileges “in light of their structural constitutional grounding under the separation of powers, relatively rare invocation, and unique importance to governmental functions.” Brief for United States 28. We express no view on that issue. [CONCURRENCE — Justice Thomas,] Justice Thomas, concurring in part and concurring in the judgment. I concur in the judgment and in Part II-C of the Court’s opinion because I wholeheartedly agree that “Congress’s designation of the rulemaking process as the way to define or refine when a district court ruling is ‘final’ and when an interlocutory order is appealable warrants the Judiciary’s full respect.” Swint v. Chambers County Comm’n, 514 U. S. 35, 48 (1995); ante, at 114 (quoting Swint, supra; citing Cunningham v. Hamilton County, 527 U. S. 198, 210 (1999)). It is for that reason that I do not join the remainder of the Court’s analysis. The scope of federal appellate jurisdiction is a matter the Constitution expressly commits to Congress, see Art. I, § 8, cl. 9, and that Congress has addressed not only in 28 U. S. C. §§ 1291 and 1292, but also in the Rules Enabling Act amendments to which the Court refers. See ante, at 113-114 (citing §§2072-2073). The Court recognizes that these amendments “designat[e] rulemaking, ‘not expansion by court decision,’ as the preferred means for determining whether and when prejudgment orders should be immediately appealable.” Ante, at 113 (quoting Swint, supra, at 48). Because that designation is entitled to our full respect, and because the privilege order here is not on all fours with orders we previously have held to be appealable under the collateral order doctrine, see Cohen v. Beneficial Industrial Loan Corp., 337 U. S. 541 (1949), I would affirm the Eleventh Circuit’s judgment on the ground that any “avenue for immediate appeal” beyond the three avenues addressed in the Court’s opinion must be left to the “rulemaking process.” Ante, at 114; see ante, at 110-113 (discussing certification under 28 U. S. C. § 1292(b), petitions for mandamus, and appeals from contempt orders). We need not, and in my view should not, further justify our holding by applying the Cohen doctrine, which prompted the rulemaking amendments in the first place. In taking this path, the Court needlessly perpetuates a judicial policy that we for many years have criticized and struggled to limit. See, e. g., Ashcroft v. Iqbal, 556 U. S. 662, 671-675 (2009); Will v. Hallock, 546 U. S. 345, 349 (2006); Sell v. United States, 539 U. S. 166, 177 (2003); Cunningham, supra, at 210; Digital Equipment Corp. v. Desktop Direct, Inc., 511 U. S. 863, 884 (1994); Swint, supra, at 48; Lauro Lines s.r.l. v. Chasser, 490 U. S. 495, 498-501 (1989); Van Cauwenberghe v. Biard, 486 U. S. 517, 527 (1988). The Court’s choice of analysis is the more ironic because applying Cohen to the facts of this case requires the Court to reach conclusions on, and thus potentially prejudice, the very matters it says would benefit from “the collective experience of bench and bar” and the “opportunity for full airing” that rulemaking provides. Ante, at 114. “Finality as a condition of review is an historic characteristic of federal appellate procedure” that was incorporated in the first Judiciary Act and that Congress itself has “departed from only when observance of it would practically defeat the right to any review at all.” Cobbledick v. United States, 309 U. S. 323, 324-325 (1940). Until 1949, this Court’s view of the appellate jurisdiction statute reflected this principle and the statute’s text. See, e.g., Catlin v. United States, 324 U. S. 229, 233 (1945) (holding that § 128 of the Judicial Code (now 28 U. S. C. § 1291) limits review to decisions that “en[d] the litigation on the merits and leav[e] nothing for the court to do but execute the judgment”). Cohen changed all that when it announced that a “small class” of collateral orders that do not meet the statutory definition of finality nonetheless may be immediately appealable if they satisfy certain criteria that show they are “too important to be denied review.” 337 U. S., at 546. Cohen and the early decisions applying it allowed § 1291 appeals of interlocutory orders concerning the posting of a bond, see id., at 545-547, the attachment of a vessel in admiralty, see Swift & Co. Packers v. Compania Colombiana Del Caribe, S. A., 339 U. S. 684, 688-689 (1950), and the imposition of notice costs in a class action, see Eisen v. Carlisle & Jacquelin, 417 U. S. 156, 170-172 (1974). As the Court’s opinion notes, later decisions sought to narrow Cohen lest its exception to § 1291 “ ‘swallow’ ” the final judgment rule. Ante, at 106 (quoting Digital Equipment, supra, at 868); see generally Coopers & Lybrand v. Livesay, 437 U. S. 463, 467-468 (1978). The Court has adhered to that narrowing approach, principally by raising the bar on what types of interests are “important enough” to justify collateral order appeals. See, e. g., Will, supra, at 352-353 (explaining that an interlocutory order typically will be “important” enough to justify Cohen review only where “some particular value of a high order,” such as “honoring the separation of powers, preserving the efficiency of government . . . , [or] respecting a State’s dignitary interests,” is “marshaled in support of the interest in avoiding trial” and the Court determines that denying review would “imperil” that interest); Digital Equipment, supra, at 878-879 (noting that appealability under Cohen turns on a “judgment about the value of the interests that would be lost through rigorous application of a final judgment requirement,” and that an interest “qualifies as ‘important’ in Cohen’s sense” if it is “weightier than the societal interests advanced by the ordinary operation of final judgment principles”). As we recognized last Term, however, our attempts to contain the Cohen doetrine have not all been successful or persuasive. See Ashcroft, supra, at 672 (“As a general matter, the collateral-order doctrine may have expanded beyond the limits dictated by its internal logic and the strict application of the criteria set out in Cohen”). In my view, this case presents an opportunity to improve our approach. The privilege interest at issue here is undoubtedly important, both in its own right and when compared to some of the interests (e. g., in bond and notice-cost rulings) we have held to be appealable under Cohen. Accordingly, the Court’s Cohen analysis does not rest on the privilege order’s relative unimportance, but instead on its effective reviewability after final judgment. Ante, at 108-113. Although I agree with the Court’s ultimate conclusion, I see two difficulties with this approach. First, the Court emphasizes that the alternative avenues of review it discusses (which did not prove adequate in this case) would be adequate where the privilege ruling at issue is “particularly injurious or novel.” Ante, at 110. If that is right, and it seems to me that it is, then the opinion raises the question why such avenues were not also adequate to address the orders whose unusual importance or particularly injurious nature we have held justified immediate appeal under Cohen. See, e. g., Sell, 539 U. S., at 177. Second, the facts of this particular case seem in several respects to undercut the Court’s conclusion that the benefits of collateral order review “cannot justify the likely institutional costs.” Ante, at 112. The Court responds that these case-specific arguments miss the point because the focus of the Cohen analysis is whether the “entire category” or “class of claims” at issue merits appellate review under the collateral order doctrine. Ante, at 107 (internal quotation marks omitted). That is exactly right, and illustrates what increasingly has bothered me about making this kind of appealability determination via case-by-case adjudication. The exercise forces the reviewing court to subordinate the realities of each case before it to generalized conclusions about the “likely” costs and benefits of allowing an exception to the final judgment rule in an entire “class of cases.” The Court concedes that Congress, which holds the constitutional reins in this area, has determined that such value judgments are better left to the “collective experience of bench and bar” and the “opportunity for full airing” that rulemaking provides. Ante, at 114. This determination is entitled to our full respect, in deed as well as in word. Accordingly, I would leave the value judgments the Court makes in its opinion to the rulemaking process, and in so doing take this opportunity to limit — effectively, predictably, and in a way we should have done long ago — the doctrine that, with a sweep of the Court’s pen, subordinated what the appellate jurisdiction statute says to what the Court thinks is a good idea. The Court concludes, for example, that in most cases final judgment review of an erroneous privilege ruling will suffice to vindicate the injured party’s rights because the appellate court can vacate the adverse judgment and remand for a new trial in which the protected material is excluded. Ante, at 109. But this ease appears to involve one of the (perhaps rare) situations in which final judgment review might not be sufficient because it is a case in which the challenged order already has had “implications beyond the case at hand,” namely, in the separate class action in Williams v. Mohawk Indus., Inc., No. 4:04-CV-00003-HLM (ND Ga.). Ante, at 112. The Court also concludes that the “likely institutional costs” of allowing collateral order review would outweigh its benefits because, inter alia, such review would "needlessly burden the courts of appeals.” Ibid. But as the Court concedes, it must speculate on this point because the three Circuits that allow Cohen appeals of privilege rulings have not been overwhelmed. See ante, at 113.