As the American economy recovers from the financial crisis, courts and agencies continue to debate how best to sanction the conduct that sparked the collapse. In its postcrisis response, the Securities and Exchange Commission (SEC) has used consent decrees as a tool of choice. Designed to promote prompt resolution of disputes and efficient use of judicial resources, consent decrees are court-approved settlements that combine judicial enforcement power with agency settlement discretion.^1×1. See Anthony DiSarro, Six Decrees of Separation: Settlement Agreements and Consent Orders in Federal Civil Litigation, 60 Am. U. L. Rev. 275, 277 (2010) (explaining that consent decrees differ from ordinary settlements insofar as they contain injunctions backed by the court’s contempt power). The SEC may also resolve disputes through an out-of-court administrative process. See Russell G. Ryan, Why the SEC Needs ‘No-Admit’ Settlements, Wall St. J., May 21, 2013, http://online.wsj.com/news/articles/SB10001424127887324767004578491410503567462. Although the use of consent decrees is an established practice, in most contexts courts have struggled to define the appropriate level of deference due when they review decrees.^2×2. See Thomas M. Mengler, Consent Decree Paradigms: Models Without Meaning, 29 B.C. L. Rev. 291, 292–93 (1988). Recently, in SEC v. Citigroup Global Markets, Inc.,^3×3. 752 F.3d 285 (2d Cir. 2014). the Second Circuit clarified its standard of review for consent decrees, emphasizing that district courts should focus on ensuring that the decree is “procedurally proper.”^4×4. Id. at 295. The court rightly vacated the lower court’s rejection of the settlement at issue. By reformulating the standard of review as a highly deferential procedural test, however, the Second Circuit overcorrected the district court’s exacting substantive review. An ideal standard would split the difference between the two approaches and review the substantive merits of the decree with a deferential posture.

In October 2011, after a four-year investigation, the SEC filed a complaint against Citigroup Global Markets, Inc. (Citigroup), alleging that the firm negligently misrepresented its role and economic interest in creating a billion-dollar fund.^5×5. SEC v. Citigroup Global Mkts. Inc., 827 F. Supp. 2d 328, 329–30, 333 (S.D.N.Y. 2011). The SEC also filed a parallel complaint against Citigroup employee Brian Stoker for his role in structuring and marketing the fund. Id. at 329–30. Citigroup marketed the fund’s assets as sound investments selected by an independent adviser, but, according to the complaint, the firm packed the portfolio with dubious assets and then shorted the securities it had helped select.^6×6. Id. at 329. On the same day that it filed the complaint, the SEC filed a proposed consent decree to settle its claims against Citigroup.^7×7. Id. at 330. The decree permanently restrained Citigroup from future violations of the Securities Act of 1933^8×8. 15 U.S.C. §§ 77a–77aa (2012). and required the firm to pay $285 million in disgorged profits and civil penalties.^9×9. Citigroup, 827 F. Supp. 2d at 330. By betting against the fund’s performance, Citigroup realized net profits of around $160 million, while investors lost more than $700 million. Id. Citigroup agreed to turn over the profits plus $30 million in interest and $95 million as a civil penalty. Id. Absent from the settlement was any admission of culpability.^10×10. Id.

The district court refused to approve the consent decree.^11×11. Id. at 335. Although Judge Rakoff recognized the “substantial deference” owed to an administrative body,^12×12. Id. at 331. he concluded that the court must still satisfy itself that the settlement is “fair, reasonable, adequate, and in the public interest.”^13×13. Id. at 332 (quoting Memorandum by Plaintiff Securities & Exchange Commission in Support of Proposed Settlement at 5, Citigroup, 827 F. Supp. 2d 328 (No. 11-CIV-7388)). Acknowledging that “all these requirements inform each other,” id., Judge Rakoff analyzed the four factors together, id. at 332–35. Turning to these factors, he noted the asymmetry between the SEC’s suggestions of fraud^14×14. Judge Rakoff highlighted language in the parallel complaint against Brian Stoker — also charged with negligence — which suggested that Citigroup was guilty of fraud. Id. at 330. and its decision to charge Citigroup with only negligence.^15×15. Id. Further, Judge Rakoff criticized the settlement’s $95 million civil penalty as “pocket change to any entity as large as Citigroup”^16×16. Id. at 334. and wondered what the SEC was getting from the deal, “other than a quick headline.”^17×17. Id. at 333.

Judge Rakoff also found particularly problematic the SEC’s longstanding policy of settling claims with defendants who neither admit nor deny wrongdoing. The practice both deprived the court of the means to assess the factual basis for the requested relief^18×18. Id. at 332. and ignored the “overriding public interest in knowing the truth.”^19×19. Id. at 335. Therefore, without “cold, hard, solid facts, established either by admissions or by trials,” the district court concluded that the consent decree was not fair, reasonable, adequate, or in the public interest.^20×20. Id. The district court refused to become a “mere handmaiden to a settlement privately negotiated on the basis of unknown facts.” Id. at 332.

The Second Circuit vacated and remanded.^21×21. Citigroup, 752 F.3d at 298. Writing for the panel, Judge Pooler^22×22. Judge Pooler was joined by Judge Lohier, who also wrote a concurring opinion, and by Judge Carney. clarified that the proper standard for reviewing a proposed consent decree requires the district court to determine whether the settlement is “fair and reasonable,”^23×23. Citigroup, 752 F.3d at 294. and, if the decree includes injunctive relief, that the “public interest would not be disserved” by the agreement.^24×24. Id. (quoting eBay Inc. v. MercExchange, L.L.C., 547 U.S. 388, 391 (2006)) (internal quotation marks omitted). The Second Circuit explicitly excluded “adequacy” from the test.^25×25. Id. The court reasoned that the adequacy requirement was an interloper from the class action context: because a consent decree, unlike a class action judgment, does not bind future claimants, it “does not pose the same concerns regarding adequacy.”^26×26. Id. Pursuant to Federal Rule of Civil Procedure 23(e)(2), a court must consider the adequacy of any class action settlement that would bind class members. Citing institutional competencies,^27×27. Id. at 296 (explaining that the “responsibilities for . . . resolving the struggle between competing views of the public interest are not judicial ones” (quoting Chevron U.S.A. Inc. v. Natural Res. Def. Council, Inc., 467 U.S. 837, 866 (1984))). the court also concluded that the “job of determining” the public interest “rests squarely with the S.E.C., and its decision merits significant deference.”^28×28. Id. The court concluded that “discretionary matters of policy,” such as the SEC’s decision to settle without requiring Citigroup to admit liability, were outside the scope of the district court’s public interest inquiry. Id. at 297. Judge Pooler acknowledged that “[c]onsent decrees vary,” and in certain instances the district court “may need to make additional inquiry to ensure that the consent decree is fair and reasonable.”^29×29. Id. at 295. Nevertheless, the Second Circuit emphasized that the “primary focus” of the district court’s review should be “ensuring the consent decree is procedurally proper, using objective measures”^30×30. Id. (emphasis added). such as whether the decree (1) is lawful,^31×31. Id. at 294. (2) is clear, (3) resolves the claims, and (4) is not “tainted by improper collusion.”^32×32. Id. at 295.

After establishing the proper test for review, Judge Pooler determined that the district court abused its discretion. Recognizing that consent decrees “are primarily about pragmatism,” she concluded, first, that the district court had no right to demand that the SEC establish the truth of the allegations in the consent decree.^33×33. Id. Second, she found that the district court incorrectly defined the public interest as “an overriding interest in knowing the truth”^34×34. Id. at 297 (quoting SEC v. Citigroup Global Mkts. Inc., 827 F. Supp. 2d 328, 335 (S.D.N.Y. 2011)) (internal quotation marks omitted). — a more appropriate inquiry would be, for example, whether the consent decree would bar private litigants from pursuing independent claims.^35×35. Id. Third, Judge Pooler stressed that second-guessing the charging and settlement decisions of an agency is “not the job of the courts.”^36×36. Id.

Judge Lohier concurred, expressing his view that the four factors identified by the majority are the only factors a district court may consider in its “fair and reasonable” analysis.^37×37. Id. at 298 (Lohier, J., concurring) (internal quotation marks omitted). Judge Lohier also stated that he was inclined to reverse based on the factual record.^38×38. Id. Nevertheless, he saw no harm in remanding “to permit the very able and distinguished District Judge” to determine whether the decree met the court’s standard.^39×39. Id. On remand, Judge Rakoff voiced his misgivings that agency settlements will now be subject “to no meaningful oversight whatsoever.” SEC v. Citigroup Global Mkts. Inc., No. 11-cv-7387, 2014 WL 3827497, at *1 (S.D.N.Y. Aug. 5, 2014). He upheld the decree, however, admitting that the Second Circuit had “fixed the menu, leaving this Court with nothing but sour grapes.” Id.

The court was right to vacate Judge Rakoff’s nondeferential decision. In an effort to curb the excesses below, however, the Second Circuit overcorrected by adopting a purely procedural test that will, practically speaking, result in the rubber-stamping of consent decrees. To restore a meaningful check on an agency subject to regulatory capture, the court should adopt a balanced standard of review, reincorporating a deferential adequacy requirement while rejecting general presumptions against no-admit/no-deny settlements.

The district court’s rejection of the SEC settlement exceeded the bounds of precedent. First, Judge Rakoff overreached in his demand that the SEC establish the “truth” of the allegations against Citigroup.^40×40. SEC v. Citigroup Global Mkts. Inc., 827 F. Supp. 2d 328, 332, 333 (S.D.N.Y. 2011). The Supreme Court has declared that a substantive inquiry has its limits: lower courts should not attempt to resolve the factual disputes of cases in their review of consent decrees.^41×41. See United States v. Armour & Co., 402 U.S. 673, 682 (1971) (recognizing that a consent decree “must be construed as . . . written, and not as it might have been written had the plaintiff established his factual claims and legal theories in litigation”); see also United States v. Oregon, 913 F.2d 576, 582 (9th Cir. 1990) (“The reviewing court should not determine contested issues of fact that underlie the dispute.”). Moreover, the district court improperly considered the charges levied against the defendant.^42×42. See Citigroup, 827 F. Supp. 2d at 330 (suggesting that the defendant should have been charged with an “allegation of knowing and fraudulent intent” rather than “negligence”). Again, courts have repeatedly held that the review of a consent decree is not the forum for a district judge to “reach beyond the complaint to evaluate claims that the government did not make.”^43×43. United States v. Microsoft Corp., 56 F.3d 1448, 1459 (D.C. Cir. 1995) (emphasis omitted); see also Heckler v. Chaney, 470 U.S. 821, 831 (1985) (noting that enforcement decisions are “generally committed to an agency’s absolute discretion”).

Nonetheless, although the Second Circuit properly vacated Judge Rakoff’s decision, it did so by effectively imposing a procedural standard of review for consent decrees. Judge Pooler established a default procedural norm, directing the district court to focus on “objective measures” of fairness and reasonableness.^44×44. Citigroup, 752 F.3d at 295. To be sure, the opinion suggested the possibility of substantive review in its recognition that “[c]onsent decrees vary,” and that some may require “additional inquiry.”^45×45. Id. The panel majority, unlike Judge Lohier, did not hold that the district court could consider only the four “objective measures.” Compare id. at 294 (holding that a court should, “at a minimum,” consider the four objective factors), with id. at 298 (Lohier, J., concurring) (declaring that the fair and reasonable standard “involves a straightforward analysis of only the four factors”). However, between the court’s explicit exclusion of the adequacy factor and its emphasis on procedural propriety, the court likely foreclosed all meaningful substantive review.^46×46. If a court cannot consider the size of the settlement, and charging decisions are committed to an agency’s discretion, then little seems left for the court to review. Even though Judge Pooler retained the public interest inquiry, she offered only examples of a procedural review,^47×47. Citigroup, 752 F.3d at 297 (suggesting that a district court could consider whether a consent decree would have a res judicata effect on private litigants pursuing their own separate claims). and it is not clear how a district court could consider substantive factors — such as deterrence — if adequacy is beyond its purview.

The Second Circuit’s procedural standard of review starkly diverges from established practice. Indeed, courts have considered the adequacy of consent decrees for more than three decades.^48×48. See, e.g., EEOC v. Product Fabricators, Inc., 666 F.3d 1170, 1172–73 (8th Cir. 2012); SEC v. Randolph, 736 F.2d 525, 529 (9th Cir. 1984). Formerly, district courts “enter[ed] a consent decree . . . only after considering [its] substantive validity.”^49×49. Adams v. Bell, 711 F.2d 161, 170 n.40 (D.C. Cir. 1983). Nor was the public interest inquiry always confined to questions of res judicata: courts had earlier determined whether the “proposed decree ha[d] an adequate deterrent effect for it to be in the public interest.”^50×50. Randolph, 736 F.2d at 530 (emphasis added). Moreover, the context from which the court borrowed its definition of the public interest inquiry — permanent injunctions — supports an open-ended examination of policy concerns. See, e.g., Metro-Goldwyn-Mayer Studios, Inc. v. Grokster, Ltd., 518 F. Supp. 2d 1197, 1222 (C.D. Cal. 2007); Commonwealth Scientific & Indus. Research Org. v. Buffalo Tech. Inc., 492 F. Supp. 2d 600, 607 (E.D. Tex. 2007).

An ideal standard of review would achieve a middle ground between the circuit and district court decisions. To guard against collusive settlements and protect the public’s interest in deterrence, the Second Circuit should reincorporate a deferential adequacy requirement, which in turn should permit a more robust public interest inquiry, into its standard of review. The court should also eschew Judge Rakoff’s general presumption against no-admit/no-deny settlements. Rather, district courts should deferentially review the adequacy of each individual settlement by comparing the penalty — and any admission of liability — against both the gains realized by the defendant and the losses suffered by investors.^51×51. Of course, the district court should also discount the alleged gains and losses according to the SEC’s probability of success at trial. Although no magic formula exists, if the comparison drastically differs from the court’s expectation, then in rare cases it may be appropriate for the court to hold further hearings or reject the decree. By avoiding the extremes of either court’s approach, the proposed standard would be superior on grounds of both administrability and policy.

First, the Second Circuit’s tidy procedural test will likely be difficult to administer. Procedure is not so easily divorced from substance,^52×52. Indeed, courts have long struggled to separate procedure from substance in administrative law. See, e.g., Ethyl Corp. v. EPA, 541 F.2d 1, 68–69 (D.C. Cir. 1976) (en banc) (Leventhal, J., concurring). Scholars and students of civil procedure have been similarly baffled by Hanna v. Plumer, 380 U.S. 460 (1965), and its “arguably procedural” test. John Hart Ely, The Irrepressible Myth of Erie, 87 Harv. L. Rev. 693, 697 (1974); see also id. 722–25. and thus certain elements of the court’s procedural standard are not amenable to procedural definition. For instance, it is not clear how adequacy is severable from the court’s definition of “fair and reasonable.” Given the limited record before the district court, a lenient settlement may be the only evidence of improper collusion. A similar dissonance plagues the court’s definition of the public interest inquiry. Because modest penalties may disserve the public interest by failing to deter misconduct, adequacy is the focal point of the public’s interest in a consent decree. Thus, although the precise limits of the Second Circuit’s standard will likely remain unclear, if judges err on the cautious side and consider only “procedural” factors, they will likely not test for collusion or examine the public interest concerns.

By contrast, a deferential adequacy requirement would be well within judicial competency. After all, courts painstakingly review consent decrees in other contexts. Under the Tunney Act,^53×53. Antitrust Procedures and Penalties Act, 15 U.S.C. § 16(b)–(h) (2012). for instance, judges may consider a variety of substantive factors — including the “competitive impact of such judgment” and the “adequacy” of the remedy — when reviewing proposed antitrust settlements.^54×54. Id. § 16(e)(1)(A). Courts embrace a similar substantive standard when reviewing environmental settlements: among the considerations are “corrective justice and accountability,” United States v. Cannons Eng’g Corp., 899 F.2d 79, 87 (1st Cir. 1990), and adequate compensation to the public for the costs of remedial measures, see id. at 90. Moreover, judges are well equipped to review even complex settlement arrangements. Indeed, when judges evaluate an injunction, they weigh it in light of the whole package of relief: some combination of retrospective damages, prospective damages, and equitable relief.^55×55. See, e.g., eBay Inc. v. MercExchange, L.L.C., 547 U.S. 388, 391 (2006). If judges can determine the adequacy of a bundle of remedies, they should also be competent to assess whether a single civil penalty is too light.

Second, as a matter of policy, the proposed standard of review would strike the appropriate balance between judicial deference and adequate oversight. On the one hand, the proposed standard would allow greater room for agency decisionmaking than did the district court’s exacting standard. Although Judge Rakoff did not explicitly proscribe no-admit/no-deny settlements as a matter of law,^56×56. Judge Rakoff was equally skeptical of the settlement’s “very modest penalty.” SEC v. Citigroup Global Mkts. Inc., 827 F. Supp. 2d 328, 333 (S.D.N.Y. 2011). Presumably, if the settlement price were high enough to offset the lack of admissions in the consent decree, Judge Rakoff might have approved the settlement. See SEC v. Bank of Am. Corp., Nos. 09 Civ. 6829, 10 Civ. 0215, 2010 WL 624581, at *5–6 (S.D.N.Y. Feb. 22, 2010) (Rakoff, J.) (approving a no-admit/no-deny consent decree after the SEC increased the settlement’s civil penalty). the import of his “truth” rhetoric set a prohibitive bar for this category of consent decrees.^57×57. See Citigroup, 827 F. Supp. 2d at 335. In so doing, the district court intruded on a policymaking function that inheres in the executive branch.^58×58. See Maryland v. United States, 460 U.S. 1001, 1004 (1983) (Rehnquist, J., dissenting) (explaining that the decision to settle is an executive function); In re Cuyahoga Equip. Corp., 980 F.2d 110, 118 (2d Cir. 1992) (“Appellate courts ordinarily defer to the agency’s expertise and the voluntary agreement of the parties in proposing the settlement.”). After all, the SEC’s policy of pursuing no-admit/no-deny settlements is informed by budgetary constraints^59×59. See, e.g., John C. Coffee, Jr., SEC Enforcement: What Has Gone Wrong?, CLS Blue Sky Blog (Jan. 2, 2013), http://clsbluesky.law.columbia.edu/2013/01/02/sec-enforcement-what-has-gone-wrong [http://perma.cc/W4NQ-DYDU] (“[T]he SEC is an overworked, underfunded agency that is subject to severe resource constraints.”). and the strategic balancing of the costs of negotiating failure against any deterrence benefits that might flow from an admission of liability.^60×60. See Robert Khuzami, Dir., Div. of Enforcement, SEC, Remarks Before the Consumer Federation of America’s Financial Services Conference (Dec. 1, 2011), http://www.sec.gov/news/speech/2011/spch120111rk.htm [http://perma.cc/Y3RZ-MNSF]. If consent decrees were required to include admissions of culpability, scholars predict that private parties would be less likely to settle due to the threat of private litigation premised on such stipulations.^61×61. Professor Joseph Grundfest predicts that “no rational [corporate] defendant” would admit wrongdoing. Jean Eaglesham & Chad Bray, Citi Ruling Could Chill SEC, Street Legal Pacts, Wall St. J., Nov. 29, 2011, http://online.wsj.com/news/articles/SB10001424052970203935604577066242448635560. Or the SEC and settling parties might prefer to settle through the SEC administrative process, avoiding federal courts altogether.^62×62. See Ryan, supra note 1. Given his interest in promoting truth and transparency in financial markets, it is unlikely that Judge Rakoff sought to push the SEC toward venues that are even less observable and accountable. The proposed deferential adequacy standard encourages transparent, efficient resolution of enforcement actions by respecting the SEC’s expertise in crafting consent decrees. Unlike the district court’s test, which all but fashioned a categorical rule that would restrict the SEC’s general settlement strategy, the proposed standard would confine its review to the adequacy of the individual settlement package.

On the other hand, the Second Circuit’s procedural test grants too much deference. A modest substantive review may be necessary to protect the public interest. In the antitrust field, scholars recognize that the Tunney Act “remains a significant deterrent” to sweetheart deals.^63×63. Lloyd C. Anderson, United States v. Microsoft, Antitrust Consent Decrees, and the Need for a Proper Scope of Judicial Review, 65 Antitrust L.J. 1, 37 (1996). In fact, “since the Act became law there appear to have been almost no controversies like the cries of foul play surrounding” the approval of earlier consent decrees.^64×64. Id. at 38. The SEC could similarly benefit from substantive judicial review. Scholars^65×65. See, e.g., Lawrence G. Baxter, Essay, “Capture” in Financial Regulation: Can We Channel It Toward the Common Good?, 21 Cornell J.L. & Pub. Pol’y 175, 182 (2011). and government watchdogs^66×66. See, e.g., Michael Smallberg, Project on Gov’t Oversight, Dangerous Liaisons: Revolving Door at SEC Creates Risk of Regulatory Capture 6 (2013), http://pogoarchives.org/ebooks/20130211-dangerous-liaisons-sec-revolving-door.pdf [http://perma.cc/CY46-7PJ8]. have both observed that the SEC is vulnerable to regulatory capture. A “revolving door” between the SEC and Wall Street creates significant conflicts of interest that can undermine the SEC’s protection of investors.^67×67. Id. at 2. Lucrative job prospects in the private sector may encourage SEC staff to curry favor with potential employers by “soft-pedaling” cases.^68×68. Id. at 27–28. Thus, as the interests of SEC investigators might not always align with the public interest, courts necessarily must provide an additional check on agency settlements.^69×69. See Sanford I. Weisburst, Judicial Review of Settlements and Consent Decrees: An Economic Analysis, 28 J. Legal Stud. 55, 67 (1999) (“The greater the divergence in interest, the greater the need for judicial review.”).

Ultimately, the procedural review that the Second Circuit articulated both deviates from precedent and fails to recognize judicial competency to deferentially review the adequacy of civil penalties and guard against regulatory capture. Then again, the Citigroup decision is probably not a watershed in the court’s consent-decree jurisprudence. Given the flurry of publicity generated by Judge Rakoff’s opinion, and the rising tide of district courts emulating his expansive standard of review,^70×70. See, e.g., Ashby Jones, In Rejecting SEC Settlement, Has Pauley Pulled a Rakoff?, Wall St. J.L. Blog (Mar. 18, 2010, 12:58 PM), http://blogs.wsj.com/law/2010/03/18/in-rejecting-sec-settlement-has-pauley-pulled-a-rakoff [http://perma.cc/5VMQ-3V98]. the Second Circuit may have wanted to send a particularly clear statement regarding the proper test for consent decrees. Intent on differentiating itself from the excesses of the lower court, the Second Circuit then “overshot” and set up an unduly permissive test.