UNITED STATES OF AMERICA PLAINTIFF v. LOUISVILLE JEFFERSON COUNTY METRO GOVERNMENT DEFENDANT (2025)

ORDER GRANTING THE UNITED STATES' MOTION TO DISMISS

A year ago, in the waning days of a presidential administration, the Department of Justice sued the Louisville Metro Government, alleging that the Louisville Metro Police Department had violated residents' constitutional and statutory rights in a systematic and ongoing manner. That same day, the Justice Department and the City proposed to resolve the lawsuit through a 248-page consent decree that would have put a judge (with assistance from a so-called “independent monitor”) in charge of supervising LMPD reforms until at least 2030.

Why such a drastic remedy? The parties insisted that reallocating core local authority from elected representatives to an appointed judge was essential: judicial insulation from the political process was the best (and likely only) way to conform Louisville's policing to the demands of federal law. True, the City denied that it was engaging in a “pattern or practice” of violations that the Attorney General could sue to eliminate. 34 U.S.C. § 12601. Yet both sides insisted that immediate judicial oversight of the police department, according to a negotiated plan of reforms, was superior to adversarial litigation in open court. Joint Motion for Entry of Consent Decree (DN 4-2) at 5.

Perhaps settlement was the right decision at the time. Establishing facts in open court about alleged LMPD policies—toward protestors, assault complainants, warrant subjects, black people, arrestees, and the disabled, see Complaint (DN 1) ¶ 18—would've been difficult, to say the least. But this truth-or-reconciliation choice wasn't inevitable. Purchasing reform through a court monitorship wouldn't have been costless (democratically, financially, or otherwise), either. And progress toward lawful policing—no matter who's in charge—requires measurement against some historical baseline, which the City and the Justice Department never agreed on. One year later, hindsight teaches that a court order wasn't the only path to reform after all. The Justice Department has changed direction, the City has adopted its own set of reforms, and the courts remain open to individuals aggrieved by lawless policing. The success of these efforts and the safety of Louisville's streets will, as ever, depend not on the decisions of a judge, but on the diligence of its officers, elected officials, and voters.

A. Judicial Review of Proposed Consent Decree. The parties' original plan to “settle first, litigate never” was an abnormal, but not unheard-of, approach to federal law enforcement. Typically, in response to alleged lawbreaking, a prosecutor would indict officials criminally or pursue civil damages in court. Vindicating the law would require proving those charges in open court before a neutral judge, who would then decide on an appropriate remedy. But a consent-decree case often switches the roles and sequence: it may begin, rather than end, with a set of remedial reforms agreed to by the parties and presented to the judge for entry as a court order. See Simon Brewer, The Attorney General's Settlement Authority and the Separation of Powers, 130 YALE L.J. 174, 180 (2020) (describing this “undertheorized form of unorthodox policymaking” that allows the executive branch to “bin[d] itself and others (including future administrations) through litigation”). The judge, moreover, remains directly involved in ensuring ongoing compliance with that decree, which now stands as a proxy for compliance with the law.

Many district judges, attorneys general, and municipalities have accepted such “judicial administration”—however oxymoronic that might sound to devotees of the separation of powers—in lieu of the jury box or the ballot box. Indeed, governing caselaw permits settlements that rely on judicial supervision so long as they “spring from and serve to resolve a dispute within the court's subject-matter jurisdiction,” “come within the general scope of the case made by the pleadings,” and “further the objective of the law upon which the complaint was based.” Lexington Insurance Co. v. Ambassador Group, 581 F. Supp. 3d 863, 868 (W.D. Ky. 2021) (citations omitted). Some decisions even point to “a presumption in favor of voluntary settlement,” United States v. Lexington-Fayette Urb. County Gov't, 591 F.3d 484, 490 (6th Cir. 2010) (quotation marks omitted), as one reason to approve consent decrees, see United States v. Akzo Coatings of America, Inc., 949 F.2d 1409, 1436 (6th Cir. 1991).

But approval of such settlements is not automatic, and a judge is not a “rubber stamp” available to settling government officials. Id. at 1435 (quoting City of Detroit v. Grinnell Corp., 495 F.2d 448, 463 (2d Cir. 1974)). Even in decisions citing a presumption favoring negotiated agreements, courts have carefully scrutinized the nature of the remedial proposals in light of the statutory violations they aim to redress. See id. at 1435–54; Lexington-Fayette, 591 F.3d at 489–91. Indeed, judges must conduct an “independent evaluation” of proposed decrees to determine whether they are “fair, adequate, and reasonable, as well as consistent with the public interest.” Akzo, 949 F.2d at 1435; United States v. County of Muskegon, 298 F.3d 569, 581 (6th Cir. 2002).¹ This evaluation, to be sure, “must not be turned into a trial or a rehearsal of the trial,” Grinnell, 495 F.2d at 462, as counsel for the Justice Department emphasized, see Hearing Tr. (DN 52) at 118:2–3. Yet only by understanding “sufficient facts” about an agreement may a judge “intelligently” decide whether to approve it. Grinnell, 495 F.2d at 462–63.

Whether a proposed consent decree is “fair, adequate, and reasonable, as well as consistent with the public interest” draws on open-ended terms that may not, at first blush, signal much constraint or even caution to a presiding judge seeking to facilitate reform and agreement. Indeed, the standard seems to invite a judge to consider his or her own notions of fairness, justice, or the common good.² If that centripetal standard were all the law offered, it might pull extreme power into the hands of a single judge. And in some instances judges approving consent decrees do seem to draw more on notions of fairness or even conscience, divorced from normal legal text and precedent.³ But consent decrees are “strange hybrid[s].” Brown v. Neeb, 644 F.2d 551, 557 (6th Cir. 1981). They are not just contracts but also injunctions, United States v. Swift & Co., 286 U.S. 106, 114 (1932), “subject to the rules generally applicable to other judgments and decrees,” Frew ex rel. Frew v. Hawkins, 540 U.S. 431, 437 (2004) (quotation marks omitted). And the traits produced by this doctrinal cross-breeding are in obvious tension: deference due the parties' agreement, rooted in consent and contract law, alongside the scrutiny due remedies involving the court's supervision, constrained by the law of equity.

Those roots in equity provide more guidance, and more constraint, than a superficial review of contemporary decisions might suggest. A consent decree is a particular, and particularly aggressive, form of equitable remedy. See Bergmann v. Michigan State Transp. Comm'n, 665 F.3d 681, 683 (6th Cir. 2011) (citing Cook v. City of Chicago, 192 F.3d 693, 695 (7th Cir. 1999)). As with other equitable remedies, judges may enter forward-looking consent decrees to redress actual injuries—almost always ones that remain ongoing. See Swift & Co. v. United States, 276 U.S. 311, 326 (1928) (“[A] suit for an injunction deals primarily, not with past violations, but with threatened future ones.”). Lawful decrees extinguish unlawful conduct, while “decrees exceed appropriate limits if they are aimed at eliminating a condition that does not violate [federal law] or ․ flow from such a violation.” Milliken v. Bradley (“Milliken II”), 433 U.S. 267, 282 (1977) (emphasis added). In other words, proper decrees follow the “settled rule that in federal equity cases the nature of the violation determines the scope of the remedy.” Rizzo v. Goode, 423 U.S. 362, 378 (1976) (quotation marks omitted).

True, parties don't have to litigate all the way through trial to earn a consent decree; earlier compromise is often the whole point of a carefully negotiated settlement agreement. But any judge-ordered remedies that parties include in that agreement nevertheless must hew closely to the facts of the controversy at hand and the laws as enacted by Congress—including the law of equitable remedies as it existed when Congress afforded federal judges these remedial powers. See generally Grupo Mexicano de Desarrollo S.A. v. Alliance Bond Fund, 527 U.S. 308, 332 (1999) (federal “equitable powers” do not include “the power to create remedies previously unknown to equity jurisprudence.”). After all, consent decrees (particularly the institutional-reform variety at issue here) don't just confine the current authority of government agencies and others made party to the judgment; they often affect the rights of future officeholders and third parties extending into the future. And they do so with the court's imprimatur and on the threat of contempt sanctions for noncompliance. Spallone v. United States, 493 U.S. 265, 276 (1990). This is sterner stuff than basic contractual remedies-at-law.

Like all equitable remedies, consent decrees require judges to exercise discretion about the decree's relationship to the underlying legal controversy. And debates about the nature and scope of this discretion reach back to the Constitution's framing—and even earlier.⁴ But courts and commentators have long agreed that Congress' grant of equitable discretion to federal trial courts did not codify judges' personal views of fairness. “[T]he discretion” granted “is not the mere personal discretion of the ․ judge, but is a judicial discretion, which means that the judge consults precedents to find the principles ․ applicable to a given situation, and then determines, from all the facts in the case, what relief will best give effect to the various principles involved.” HENRY L. MCCLINTOCK, HANDBOOK OF THE PRINCIPLES OF EQUITY 51–52 (1948) (emphasis added). The rule requiring judges to ask whether a decree is “fair, adequate, and reasonable,” Akzo, 949 F.2d at 1436, therefore, is not merely legal realism masquerading as positive law. Rather, it's tied to Congress' specific decision to afford judges a specific, if malleable, form of remedial authority. See Judiciary Act of 1789, 1 Stat. 73, §§ 11, 16; Trump v. CASA, 606 U.S. 831, 841– 42 (2025). And the limits of that equitable discretion are not traced by a single judge's personal sense of fairness, but instead by “doctrines” and “habits of decisionmaking” honed over centuries of incremental development. Samuel L. Bray, The System of Equitable Remedies, 63 UCLA L. REV. 530, 545 (2016); see also Grupo Mexicano, 527 U.S. at 332; Guaranty Trust Co. of N.Y. v. York, 326 U.S. 99, 105–06 (1945).

The law of equity therefore provides careful judges with at least two important limiting principles pertinent here.

First, the nature and scope of the alleged lawbreaking matter because, among other reasons, “a court of equity” can do no more than afford “complete relief for the parties.” CASA, 606 U.S. at 847. And “[t]he parties cannot,” merely “by giving each other consideration, purchase from a court of equity a continuing injunction” that stretches beyond what the underlying law entails. System Federation No. 91, Railway Employees' Dep't, AFL-CIO v. Wright, 364 U.S. 642, 651 (1961). Instead, “the District Court's authority to adopt a consent decree comes only from the statute which the decree is intended to enforce.” Id. So although consent decrees are not strictly limited to the relief that might have been available had the parties litigated their case to judgment, the relief must address conduct within the “general scope of the case made by the pleadings.” See Local No. 93, Intern. Ass'n of Firefighters v. City of Cleveland, 478 U.S. 501, 525 (1986) (citation omitted). Only by understanding the “nature and scope” of the underlying violation, Milliken II, 433 U.S. at 280 (quoting Swann v. Charlotte-Mecklenburg Board of Education, 402 U.S. 1, 16 (1971)), may “a federal court ․ tailor ‘the scope of the remedy' to fit ‘the nature and extent of the constitutional violation,’ ” United States v. Louisville Jefferson County Metro Gov't (“LMPD II”), No. 3:24-cv-722, 2025 WL 238010 at *5 (W.D. Ky. Jan. 18, 2025) (quoting Hills v. Gautreaux, 425 U.S. 284, 293–94 (1976)).

Second, and relatedly, federal equitable relief should follow a principle of remedial parsimony—stretching no further than necessary to enforce federal law. In that way, consent decrees must be “limited to reasonable and necessary implementations of federal law,” Horne v. Flores, 557 U.S. 443, 450 (2009), and—like other equitable remedies—be “no more burdensome ․ than necessary,” Madsen v. Women's Health Center, Inc., 512 U.S. 753, 765 (1994) (quoting Califano v. Yamasaki, 442 U.S. 682, 702 (1979)). So in prior instances when the Supreme Court has blessed consent decrees, the ongoing legal injury has been both clear and redressable by a tailored remedy.⁵ “Once a right and a violation have been shown,” of course, “the scope of a district court's equitable powers to remedy past wrongs is broad, for breadth and flexibility are inherent in equitable remedies.” Swann, 402 U.S. at 15. But that flexibility serves a purpose: “restor[ing] the plaintiff to his rightful position,” which is itself a function of the substantive entitlements established by law. Bray, Equitable Remedies, at 568. So despite—or perhaps because of—that equitable discretion, establishing the connection between the alleged wrongs and proposed remedies is crucial. LMPD II, at *5.

B. This Litigation. The Consent Decree offered here implicated all the hazards and opportunities of institutional-reform adjudication: federal judicial encroachment on policymaking by local elected officials, insulation of policy choices from the political process beyond the terms of elected representatives, questions of cost and institutional competence, and more. Id. at *1; see also Michael W. McConnell, Why Hold Elections? Using Consent Decrees to Insulate Policies from Political Change, 1987 U CHI. L. FORUM 297, 314). Given those risks, the Court responded to the proposed Consent Decree by asking the parties three main questions. First, was this proposed order lawful? Second, was it necessary? Third, who would decide when LMPD's compliance with federal law was sufficient—and based on what metric? See LMPD II at *1. Only by understanding “sufficient facts” about an agreement, after all, could a judge “intelligently” decide whether to approve it. Grinnell, 495 F.2d at 462–63.

During a public hearing, the parties' lawyers struggled to answer these questions. So, too, even more basic factual questions regarding the underlying alleged legal violations by LMPD officers. During the prior decade, how many times had LMPD officers shot civilians without justification? How often have officers unlawfully tased, punched, or kicked suspects? How did the Justice Department determine when and how LMPD policing disproportionately violates the rights of black people? See DN 52. The Justice Department refused to answer, while the City admitted to “some” erroneous shootings, but disputed that a pattern or custom of such violations existed. The record was even more spare with respect to alleged ADA and sex-discrimination violations: the former weren't tied to any apparent theory of legal liability, while the latter weren't supported by any factual allegations whatsoever. See LMPD I (DN 18) at 3. So the Court directed the parties to say more about this set of seven claims, see Complaint ¶ 18, and the important questions of law and fact that undergirded them, LMPD II at *6–7.

Beyond the obvious public interest in transparency and accountability regarding such serious allegations, the equitable basis of the proposed decree rendered these answers crucial from legal perspective. If a judge is to rely on constitutional violations to justify a takeover of police decisonmaking and—one hopes—to eventually measure progress sufficient to hand power back to the City's elected representatives, then a baseline of noncompliance is necessary to determine future reform.

The parties responded, but their briefing raised more questions than it answered. The City said it would take these reform steps regardless of whether the Court entered the proposed order. So why was judicial supervision necessary? Was there even a live dispute for a court to resolve? The Justice Department, meanwhile, refused to answer questions about the number and nature of the constitutional violations it had alleged.

The ensuing months brought three further developments.

First, a new Attorney General took office, assigned new lawyers to the case, and repeatedly asked the Court for time to review the Justice Department's case. DNs 80, 87 & 90. Like their predecessors, these lawyers declined to supply the factual and legal support the Court previously requested.

Second, the City (as it had promised it would) rolled out its own plan for police reform, which it pledged to undertake no matter what happened in this litigation. That voluntary plan's curious title? “Community Commitment – Louisville's Consent Decree.” But a new breed of consent, it seems—one drawing on the consent of the people's representatives, not the court. (Indeed, in addition to the Mayor's support, the Metro Council narrowly passed a resolution in June 2025 endorsing the reform plan.) The Justice Department's new leadership hasn't taken a position on the lawfulness or advisability of the “Community Consent Decree”— at least not in a manner shared with the Court.

Third, the Justice Department moved to dismiss this case. Motion to Dismiss (DN 93). Its motion said that the Justice Department “no longer ha[d] confidence in the strength of its case,” at least in part because “it has not yet presented sufficient evidence to support its allegations.” Id. at 2. At that point, the City could have ended this case itself—without this Court's stamp of approval—by signing a stipulation of dismissal. See FED. R. CIV. P. 41(a)(1)(A)(ii). Yet, once again, the parties' choices entangled the judiciary further: the City stated only that it “does not oppose” the Justice Department's motion to dismiss. Motion to Dismiss at 1.

C. Motion to Dismiss. Deciding whether to dismiss a case—even one the plaintiff doesn't want to keep litigating—implicates another standard that on its face offers judges little guidance. After a defendant has answered (as here) or moved for summary judgment, dismissal requires a court order, “on terms that the court considers proper.” United States ex rel. Polansky v. Exec. Health Resources, Inc., 599 U.S. 419, 435 (2023) (quoting FED. R. CIV. P. 41(a)(2)). Ordinarily, dismissal is appropriate at the plaintiff's request “in all but the most exceptional cases.” Id. at 437. Id. And although the Supreme Court “has never set out a grand theory of what ․ Rule [41(a)(2)] requires, ․ [t]he inquiry is necessarily ‘contextual,’ ” with “the Government's views ․ entitled to substantial deference.” Id. (quoting 9 WRIGHT & MILLER, FEDERAL PRACTICE & PROCEDURE § 2364 (4th ed. 2022)).

In part, that's because plaintiffs “suffer the consequences” when they decide not “to offer evidence in support of [their] claim[s]”—namely, dismissal with prejudice. Smoot v. Fox, 340 F.2d 301, 303 (6th Cir. 1964). The “primary purpose” of Rule 41(a)(2) in requiring court approval, then, “is to protect the nonmovant from unfair treatment.” Grover ex rel. Grover v. Eli Lilly & Co., 33 F.3d 716, 718 (6th Cir. 1994). Unfair treatment might come when a dismissal is without prejudice—meaning that the plaintiff is free to sue the defendant on the same claims in the future. See Cone v. West Virginia Pulp & Paper Co., 330 U.S. 212, 217 (1947). But here the Justice Department proposes to dismiss its case with prejudice, which all but eliminates concerns about unfairness to the City.

Still, a case like this implicates interests beyond the parties' own institutional concerns. As the Sixth Circuit has recognized, “there may be rare cases in which dismissal with prejudice adversely affects the interests of defendants or third parties in a way that causes them plain legal prejudice.” Pedreira v. Sunrise Children's Services, 79 F.4th 741, 751 (6th Cir. 2023). After all, “[m]odern litigation involves ever-more complex configurations of parties, and courts may face circumstances that require a fuller inquiry of a voluntary motion to dismiss with prejudice” than the ordinary bilateral case. Id. at 750. In consent-decree cases like this one—in which the Court had already held hearings, received extensive briefing, heard from amici, and imposed scheduling orders before the Justice Department walked away from the agreement—that inquiry accounts for third parties' interests as well.

Here, determining whether dismissal would be “proper” requires considering the same questions that would have governed—and quite likely prevented—issuance of the Consent Decree in the first place. 1. Was the proposed decree lawful? As an initial matter, consent decrees fit curiously into the framework of “Cases and Controversies” subject to federal adjudication under Article III. Cases that begin with an agreed order involve parties who have agreed about what the law demands of them going forward.⁶ This means the parties in a real sense have ceased to be adverse—if they ever were to begin with. See Michael T. Morley, Consent of the Governed or Consent of the Government? The Problems with Consent Decrees in Government-Defendant Cases, 16 U. PA. J. CONST. L. 638, 668 (2014); Lexington Insurance Co., 581 F. Supp .3d at 867–68 (questioning necessity of court-ordered injunction after and atop agreed private resolution). Yet similar orders enjoy at least some historical pedigree. See James E. Pfander & Daniel D. Birk, Article III Judicial Power, the Adverse-Party Requirement, and Non-Contentious Jurisdiction, 124 YALE L.J. 1346, 1410–13, 1441–43 (2015) (describing history of “non-contentious jurisdiction” and comparing it with modern consent-decree practice).⁷

Even apart from these concerns about Article III justiciability, separate questions persisted as a matter of equity. Were all the Consent Decree's remedial provisions traceable to underlying alleged violations? The reforms offered here— more than 240 pages covering complex subjects as diverse as protest security and 911 response policy—reached deep into the innerworkings of municipal policing. And “in equity,” as noted above, “the broader and deeper the remedy the plaintiff wants, the stronger the plaintiff 's story needs to be.” CASA, 606 U.S. at 854 (quoting Samuel L. Bray & Paul B. Miller, Getting into Equity, 97 NOTRE DAME L. REV. 1763, 1797 (2022)).

On at least two major points, however, the Justice Department's theories offered far too little to justify a court's intervention.

First, the decree seeks to overhaul the LMPD's protocols governing the investigation of sexual assaults (both those alleged to have been committed by civilians and those alleged against LMPD officers). See Proposed Order ¶¶ 351–96. Yet the Complaint didn't even allege unlawful sex discrimination caused by current policies, as counsel conceded. See Hearing Tr. at 129. To the contrary, the Justice Department's 2023 investigation found no “reasonable cause ․ to believe that LMPD's practices result in gender bias in violation of federal law.” 2023 Findings at 69. Yet the Consent Decree prescribed wholesale change in the Department's procedures for investigating sexual assaults. The Court identified these omissions, but received no response from the parties regarding any justification for an extensive equitable remedy that redressed no identifiable factual allegations or plausible legal liability. See LMPD II at *4 (“Where in the Complaint or 2023 Findings does the United States allege the LMPD's sexual-assault protocols violate federal law, thus justifying the robust reform process for those protocols set forth in the decree?”). To order such a remedy—unconnected to any identifiable alleged violation of federal law—surely lies beyond the broadest notions of the Court's equitable authority, see Milliken II, 433 U.S. at 282, and perhaps even beyond some conceptions of its subject-matter jurisdiction, cf. Bell v. Hood, 327 U.S. 678, 682–83 (1946).

A second and related problem afflicts the Justice Department's claim that aspects of the City's “emergency response program” discriminate against persons with “behavioral health disabilities” in violation of the Americans with Disabilities Act. Complaint (DN 1) ¶ 101. Beyond that bare allegation, the Complaint never offered a theory of liability or any more detailed allegations of how the current 911 program discriminates against this population.⁸ To remedy this unspecified potential violation, the Consent Decree sought to overhaul the City's existing “deflection” program, which aims to dispatch unarmed mental-health professionals rather than armed officers under certain circumstances. Proposed Order ¶¶ 296–330. From the start, though, the Court was skeptical that the ADA prescribed a rule specifying, for example, how 911 dispatchers may confront calls that possibly involve mental illness. And the parties' proposed reforms don't obviously flow from the statute of existing doctrine, either. So the Court asked the parties to identify “caselaw from the Sixth Circuit or elsewhere supporting the legal or remedial theories advanced by the Justice Department (for example, whether any Court has imposed ADA liability on a police department for lack of a behavioral-health response unit the consent decree contemplates).” LMPD I at 7. At neither the hearing nor in any written filing did the Justice Department provide any response at all. So here again, the Court could not identify any connection between legal violation and remedy against which to measure the extensive and intrusive proposed reforms.

At least with respect to these claims, therefore, entering judgment against the City and ordering compliance with reform provisions set forth in the Consent Decree would have stretched the Court's considerable equitable discretion past its breaking point. Because without an identifiable legal violation to resolve, no equitable remedy may follow. See, e.g., Horne, 557 U.S. at 450.

2. Was the proposed order necessary? The remainder of the Justice Department's claims—based on the First, Fourth, and Fourteenth Amendments—are at least discernible, if still contested. Nevertheless, these aspects of the proposed decree likewise fail to withstand the scrutiny that the law requires.

As noted above, the extraordinary judicial intrusion that comes from a consent decree should remain a last resort—when other options to enforce the law will not work. See Milliken II, 433 U.S. at 280–81; Meredith v. City of Winter Haven, 320 U.S. 228, 235–36 (1943). And developments during the pendency of this litigation made clear that judicial receivership of the LMPD is no longer necessary to pursue reform— even assuming it once was. Perhaps the threat of litigation was necessary, as the Justice Department and many advocates contended, to drag the City to the bargaining table. But once there, the City agreed it would—and in most instances already had—proceed with voluntary reform efforts regardless of a court order. See Mayor Greenberg Announces Historic Consent Decree (Dec. 12, 2024), [https://perma.cc/A8UZ-GJA8] (quoting Police Chief: LMPD is “already well on [its] way to implementing many of the requirements in this consent decree”). This progress continued after the change in presidential administrations, with the City soliciting and hiring an independent monitor—albeit one who would serve under the supervision of the Mayor, not the Court—and implementing some 260 new policies to “enhance accountability, transparency, and effectiveness in policing.” See Mayor Greenberg Announces Independent Monitor (Sept. 26, 2025) [https://perma.cc/F7VU-X74K]; Community Commitment: Louisville's Consent Decree, at 3 [https://perma.cc/J4MU-2XTQ]; Community Commitment Press Release (May 21, 2025) [https://perma.cc/8KE6-CYMW]. And the City reiterated its intent to adhere to the “Community Consent Decree” during the hearing on the federal motion to dismiss. See May 23, 2025, Hearing Tr. 14.

That the City promised to voluntarily undertake these many reforms (and indeed is already undertaking them) casts doubt on the notion that the proposed order ever required a court's involvement. To be sure, political officials might have preferred a judicial imprimatur, whether to “insulate the policies embodied in the decree from future political change,” see McConnell, Why Hold Elections at 300, or to evade political accountability by thrusting responsibility for less-popular or less-efficacious reforms onto a court's shoulders, see Horne, 557 U.S. at 471.⁹ But if the City was able and ultimately willing to take on reform itself—whether to alter 911 protocols, or limit use-of force policies, or institute new forms of oversight—then why should that same reform occur under the banner of a court order?

These developments within the municipal government make clear that open-ended judicial supervision is no longer essential—if it ever was—and is therefore no longer warranted. Our constitutional system tolerates judicial intrusion on the democratic process only begrudgingly and suspiciously: the “least dangerous branch” is well equipped to interpret and vindicate the specific legal rights of specific parties, but is not designed to oversee general police department reforms for years on end. Absent identifiable rights violations, our system of separated powers leaves those decisions to the lawful discretion of local, state, and sometimes even federal officials— empowered and constrained as they are by voters at the ballot box—rather than judges behind courthouse doors.

3. Who would decide—and on what basis—whether and when the LMPD had sufficiently complied with federal law? Given these elected leaders' authority to make such voluntary choices, our system likewise commits to those leaders the authority to assess those choices' success—and to adjust their plans accordingly. The proposed Consent Decree, meanwhile, would have instead interposed a judge to supervise an independent monitor who would serve as an “agent of the Court” and wield substantial authority to manage the day-to-day operations of the LMPD. See Proposed Order (DN 4-1) at 200. The Court also would have had to decide when enough reform was enough—that is, when the City had attained “substantial compliance” with the decree's objectives. Id. at 227–28. Though the Consent Decree spends pages and pages describing its bespoke process for tiered exit, 223–28, that level of detail is almost entirely procedural, providing little guidance to the Court—or clarity to the parties or public—about how to know when its benchmarks have been achieved. See also id. at 223–26 (envisioning a separate process for “self-monitoring” in some circumstances).

Prudent judges, like all effective decisionmakers, should by habit “begin with the end in mind.” as Stephen Covey taught. Mindful that exiting this Consent Decree could prove far trickier than signing it, the Court expressed doubt at the outset that an articulable and measurable baseline could guide its later termination efforts. See LMPD II, at *6. This was no small matter, moreover, given the difficulties and delays that have afflicted other lengthy consent-decree litigation. See generally LMPD I & II (describing difficulties terminating consent decrees in New Orleans, Cleveland, and Tennessee). The Justice Department's repeated refusal to provide even basic factual data about historical LMPD practices and errors exacerbated these concerns.¹⁰ Absent that information, it's hardly clear how this Court could have “limited” its supervision of the LMPD “to reasonable and necessary implementations of federal law.” Horne, 557 U.S. at 450 (quoting Frew, 540 U.S. at 441). Apparently the Court, perhaps with the aid of a monitor, would have been left to define “substantial compliance” for itself and assess progress against its own definition. This inquisitorial approach is not one that American courts have considered conducive to the sort of reasoned decisionmaking that adjudication requires. See, e.g., Allentown Mack Sales & Service v. NLRB, 522 U.S. 359, 374–75 (1998); Rizzo, 423 U.S. at 380.

D. Conclusion. These reasons—independent, yet intertwined—should leave little doubt that wholesale dismissal, as the Justice Department now requests, is “proper” under Rule 41(a)(2). Dismissal with prejudice, as the Justice Department further requests, will bar the federal government from alleging future constitutional violations arising from the same facts as its complaint set out here. But it will not shield individual LMPD actors from individual claims that they violated the constitutional rights of other citizens. Indeed, as counsel acknowledged at the May 2025 hearing, this dismissal applies only to claims against the City—“not against other third-party individuals who might have claims brought against them” or who presently have claims pending. See May 23, 2025, Hearing Tr. 12. So all that ends here is the federal government's litigation against the City—and a presumably well-meaning but ultimately misdirected effort toward judicial supervision of the police department.

These parties can hardly be blamed for running to federal court in hopes of untangling a dangerously knotty problem of public policy. The United States in 2025, after all, is a litigious place where many look first to litigation rather than legislation to address societal ills. But our Constitution assigns the democratic process and republican institutions the principal roles in institutional reform. Judges, meanwhile, play a more limited part, one Chief Justice Marshall summarized more than two centuries ago: “The province of the court is, solely, to decide on the rights of individuals, not to enquire how the executive, or executive officers, perform duties in which they have a discretion.” Marbury v. Madison, 5 U.S. (1 Cranch) 137, 170 (1803). Even when litigants believe courts are the “only institution in the United States capable of solving [a] problem,” the perceived need for judicial intervention is not a sufficient justification for judicial power. Gill v. Whitford, 585 U.S. 48, 64–65 (2018) (citing Marbury).

Clearly, and commendably, members of the Louisville community hold strong and diverse views on the necessity and requisite degree of police reform. Throughout this litigation, stakeholders spanning the ideological spectrum have raised legitimate, credible, and sincere concerns about the proposed Consent Decree. See, e.g., Amicus Briefs by Judge David L. Bazelon Center for Mental Health Law (DN 29); ACLU of Kentucky (DN 32); Heritage Foundation (DN 33). But absent an identifiable rule of law, set down by those with authority to legislate, an individual judge would always struggle to find a valid principle—beyond his or her own limited judgment—to resolve the inevitable tensions between competing policy positions. Given the parties' failure to connect the Consent Decree to specific factual allegations about historical and ongoing violations of specific legal rules, the Court lacked a basis to order such sweeping reforms.

But that doesn't mean that no violations existed. And, equally important, it doesn't prevent the parties from undertaking the hard work of reform themselves— work they have already begun, outside the judge's gaze. During the process, should the United States or another aggrieved party believe Louisville has violated the law, it may sue for an appropriate remedy that would vindicate its rights. Otherwise, the responsibility to lead the Louisville Metro Police Department in compliance with federal law must remain with the City's elected representatives and the people they serve.

The Court therefore grants the United States' motion to dismiss this case with prejudice (DN 93).

FOOTNOTES

1.   The consent-decree caselaw repeatedly refers to the public interest, implying a singular policy ideal for an undifferentiated community. See, e.g., Williams v. Vukovich, 720 F.2d 909, 923 (6th Cir. 1983). That approach sits in considerable tension with the teachings of public-choice theory, which accounts for the self-interest of differentiated constituencies within the body politic. See, e.g., James Buchanan, Public Choice: Politics Without Romance, POLICY, Spring 2003, at 13. The notion of a unitary public interest is also hard to square with the common-sense recognition that any contested policy issue (and law-enforcement policy is among the most contested these days) yields multiple views from multiple good-faith actors regarding how to pursue the common good. See, e.g., § D below. A judge, of course, can and should strive to stand outside the political arena. But doing so effectively requires some external principle or yardstick aside from his or her own view of good policy—a “neutral principle” that offers an “intellectually coherent statement of the reason for a result which in like cases will produce a like result.” ALEXANDER M. BICKEL, THE LEAST DANGEROUS BRANCH 59 (2d ed. 1986). Which is why an injunction meant to conform policing to constitutional standards should focus on officers' actual conduct and policies, measured against the governing legal requirements. Without preexisting law to apply, what distinguishes judicial decisionmaking, which should be objective, from democratic policymaking, which is necessarily pluralistic? Cf. William Baude & Stephen E. Sachs, The Law of Interpretation, 130 HARV. L. REV. 1079, 1145 (2017) (rejecting adjudication based principally on “the normative desiderata of substantive desirability, consistency with rule of law principles, and promotion of political democracy”) (quoting Richard Fallon, The Meaning of Legal “Meaning” and Its Implications for Theories of Legal Interpretation, 82 U. CHI. L. REV. 1235, 1305 (2015)).

2.   Such an approach, the Court previously noted, finds relatively “little in the way of specific guidance for district courts deciding whether to enter a proposed consent decree,” partly due to a “dearth of caselaw ․ presumably correlated with the lack of an incentive for a consenting party to take an appeal.” United States v. Louisville Jefferson Metro County Gov't (“LMPD II”), No. 3:24-cv-722, 2025 WL 238010 at *3 n.3 (W.D. Ky. Jan. 18, 2025) (quoting Lexington Insurance Co. v. Ambassador Group LLC, 581 F. Supp. 3d 863, 868 n.2 (W.D. Ky. 2021)). Given this lacuna in the appellate reports, the decision whether to fill it with a district judge's own assessment of fairness, reasonableness, and the public interest— on the one hand—or with careful reasoning by analogy from related equitable remedies—on the other—raises a consequential and undertheorized question for those confronting the modern institutional-reform consent decree.

3.   The law makes clear, however, that judicial review of proposed decrees does not invite judges to “substitute [their] own judgment for that of the parties to the decree.” Akzo, 949 F.2d at 1435 (citing United States v. Jones & Laughlin Steel Corp., 804 F.2d 348 (6th Cir. 1986) (judge may approve or reject—but not alter—consent decree)).

4.   The Ratification debates, for instance, addressed concerns that the proposed federal equity power would give judges too much power. See, e.g., Missouri v. Jenkins, 515 U.S. 70, 128–29 (1995) (Thomas, J., concurring) (collecting citations). John Selden, after all, had famously remarked more than a century earlier that a decision based on equity might well be determined by something as arbitrary as the length of “the chancellor's foot.” JOHN SELDEN, TABLE TALK 43 (Frederick Pollock, ed., 1927). Alexander Hamilton, however, argued that by the time of the Constitution's framing those equity powers were “governed” by “principles ․ now reduced to a regular system.” THE FEDERALIST NO. 83; see also 3 BLACKSTONE, COMMENTARIES 440–41 (1768) (describing then-modern equity as a system of more-or-less-settled principles: a “regular science”); 1 Joseph Story, Commentaries on Equity Jurisprudence § 20 (“Courts of Equity” decide cases according to “principles as fixed and certain, as the principles on which the Courts of Common Law proceed.”) (quotation marks omitted).

5.   See, e.g., Wright, 364 U.S. at 643–44 (specific labor practices by union and railroad discriminated against non-union rail workers in violation of Railway Labor Act); Local No. 93, Intern. Ass'n of Firefighters v. City of Cleveland, 478 U.S. 501, 504–06 (1986) (specific employment practices discriminated against non-white firefighters in violation of Title VII); Rufo v. Inmates of Suffolk County Jail, 502 U.S. 367, 372–75 (1992) (specified prison conditions violated Eighth Amendment rights of inmates); cf. Horne, 557 U.S. at 459–60 (changed factual circumstances that abate unlawful conditions may require relief from existing consent decree).

6.   This critique wouldn't apply, of course, to a consent decree negotiated during litigation but before final judgment, in which the hallmarks of adversity are self-evident. Cf. Kokkonen v. Guardian Life Ins., 511 U.S. 375, 381–82 (1994) (permitting courts to retain jurisdiction to enforce settlement agreements).

7.   As discussed below and before, the lack of any adversarial presentation during this litigation raises similar concerns that sound in prudential as well as constitutional registers. See, e.g., Paul J. Mishkin, Federal Courts As State Reformers, 35 WASH. & LEE L. REV. 949, 959 n.42 (1978) (noting, among other concerns, the “special need for effective adversary process” in “institutional reform cases where the nature of the issues strains the capacities of the judicial process in the first place”); Henry P. Monaghan, Constitutional Adjudication: The Who and When, 82 YALE L.J. 1363, 1373–74 (1973) (“Concededly, the lack of an adversary presentation increases the risk that the Court's constitutional pronouncements will not be sufficiently considered.”).

8.   Circuits have split on the question whether the ADA applies to arrest procedures. See City & County of San Francisco v. Sheehan, 575 U.S. 600, 609–10 (2015) (dismissing certiorari petition as improvidently granted with respect to this question). The Sixth Circuit has generally avoided the question. See, e.g., Roell v. Hamilton County, 870 F.3d 471, 489 (6th Cir. 2017) (rejecting proposed ADA accommodations—including de-escalation techniques—in light of arrest-specific safety concerns without deciding whether Title II applied).

9.   Similarly, the City's decision to label its voluntary plan a “consent decree”—despite the lack of any judicial involvement whatsoever—likewise has the potential to obfuscate responsibility for the success or failure of the reform proposals.

10.   The Justice Department was not only unwilling to share relevant data and findings with the Court; it also refused to share that data with its putative partners at the City. This made little sense, given the LMPD's expressed willingness to commence the reforms that both sides deemed warranted. The only justification the Justice Department offered was unwillingness to sacrifice “litigation leverage,” a curious reason for a public agency to advance when the aim is to avoid litigation by inducing compliance with the law. Now, given this case's dismissal, any possible justification to withhold that data based on the possibility of future litigation has surely evaporated. As, one hopes, has the Justice Department's continued recalcitrance, given the LMPD's agreement that such data would “be useful to the city” as it pursues reform. See Hearing Tr. 57:22–58:1.

Benjamin Beaton, District Judge United States District Court