ASSOCIATION FOR EDUCATION FINANCE AND POLICY INC v. MCMAHON (2025)

MEMORANDUM OPINION

Winds of political change have swept through the administrative state in recent months. In their aftermath, the Institute of Education Sciences—part of the Department of Education—has been fundamentally transformed. This upheaval is understandably jarring for those who rely on studies and data produced by the Institute. Education advocacy groups that use the Institute's data filed two suits under the Administrative Procedure Act (“APA”). They want a preliminary injunction to stop the sudden changes and restore things to how they used to be.

But the APA was never meant to be a bureaucratic windbreak insulating agencies from political gales. It cannot comprehensively undo multifaceted agency transformations wrought by political decisions. APA challenges must focus on specific agency actions and seek equally targeted remedies. But the challenges here bundle together broad lists of grievances and seek wholesale modifications to agency operations writ large. It is not this Court's place to breathe life back into wide swathes of the Institute's cancelled programs and then monitor the agency's day-to-day statutory compliance; essentially what Plaintiffs seek. Such widespread modification to agency operations must come instead from “the offices of the Department or the halls of Congress, where programmatic improvements are normally made.” Lujan v. Nat'l Wildlife Fed'n, 497 U.S. 871, 891, 110 S.Ct. 3177, 111 L.Ed.2d 695 (1990). Thus, the Court will deny both preliminary injunction motions.¹

I.

Over the last two centuries, the federal government has built up a robust framework for conducting education-related research. Assoc. Compl., ECF No. 1, ¶ 1. The earliest traces of these programs date back to the 1860s, but 2002 was a watershed year for education research. See id. That year, Congress enacted the Education Sciences Reform Act (“ESRA”), codified at 20 U.S.C. §§ 9501–84, which improved the collection, analysis, and dissemination of education data. Id. Among other things, the Act established the Institute. Acad. Compl., ECF No. 1, ¶ 1. The Institute includes four research centers: the National Center for Education Research (“Research Center”), the National Center for Education Statistics (“Statistics Center”), the National Center for Education Evaluation and Regional Assistance (“Evaluation Center”), and the National Center for Special Education Research (“Special Education Research Center”). See 20 U.S.C. §§ 9531–67. These centers have different structures, missions, and statutory mandates.

The Institute has flexibility on how to carry out its statutory missions. It can generally do so “directly or through grants, contracts, or cooperative agreements.” 20 U.S.C. § 9512. Traditionally, it has relied heavily on contracts and contractors. See, e.g., Jeffrey Mervis, Canceling Data Collections Could Imperil Efforts to Improve U.S. Education, Science (Feb. 12, 2025), Ex. 9, Acad. ECF No. 14-11, at 3 (The Statistics Center “has fewer than 100 employees, and more than 1000 contractors.”). It is also required to use contracts “to establish a networked system of 10 regional educational laboratories.” 20 U.S.C. § 9564(a). These laboratories provide “training ․ and technical assistance” to state and local education agencies and school boards. Id. § 9564(f)(1). They also “develop[e] and widely disseminat[e]” research and reports to help improve academic achievement. Id. § 9564(f)(2).

The data collected and disbursed by the Institute's subcomponents provides a high-resolution picture of the American education system. The federal government has many uses for this data, but it is not the only beneficiary of the Institute's work. The material is also an informational treasure trove for nonprofit organizations and individual researchers. This is by design. Part of the Institute's mission is to “provide parents, educators, students, researchers, policymakers, and the general public with reliable information” and data about education issues. 20 U.S.C. § 9511(b)(1). In furtherance of this mission, the Institute is statutorily required to make its research and data publicly available. 20 U.S.C. § 9574 (“Subject to [confidentiality provisions], data collected by the Institute ․ shall be made available to the public.”). The statutes for many of the individual centers echo the importance of public access, mandating varying levels of data disclosure. See, e.g., 20 U.S.C. § 9533(a)(7) (Research Center); id. § 9546(e)(2) (Statistics Center); id. §§ 9562(a)(2), (3) (Evaluation Center).

The number of studies and reports produced by the Institute is breathtaking, and many are carried out by the Statistics Center through contracts. See Assoc. Mot. Prelim. Inj., ECF No. 6-1, at 21–23. It is unnecessary to construct a comprehensive list, but describing some will set the stage for Plaintiffs’ concerns.

First, there is the National Postsecondary Student Aid Study. See Assoc. Compl. ¶ 33. The Statistics Center oversees this study, which is designed to fulfill a statutory requirement to research how students are funding their educations and how the cost of attendance and ensuing debt impacts them. See 20 U.S.C. § 1015a(k); Assoc. Mot. Prelim. Inj. at 22. The Statistics Center also oversees a National Assessment of Educational Progress, a statutorily required long-term academic achievement trend assessment. See 20 U.S.C. § 9622; Acad. Compl. ¶ 42.

Next is the Integrated Postsecondary Education Data System, a database administered by the Statistics Center. Acad. Compl. ¶ 34 & n.10. This system helps the Statistics Center “meet its mandate to report full and complete statistics on the condition of postsecondary education in the United States.” National Center for Education Statistics, NCES Handbook of Survey Methods - Integrated Postsecondary Education Data System (IPEDS) [https://perma.cc/SQ2G-GXCP]. It includes a legion of data collected from Title IV-eligible colleges and universities. See id.

There are also an assortment of Educational Longitudinal Studies that follow students over the course of their academic careers and beyond. Acad. Mot. Prelim. Inj., ECF No. 14-1, at 6. For example, the Beginning Postsecondary Students Longitudinal Study follows students through their post-high school educational journeys. See National Center for Education Statistics, Beginning Postsecondary Students (BPS) [https://perma.cc/4WM4-95ZM]. There are longitudinal studies for other age groups too, like kindergarteners. Assoc. Mot. Prelim. Inj. at 23 (discussing ECLS-K:2024).

The list goes on, but the Court will not. This overview barely scratches the surface of the agency's work, but it suffices to say that the Institute controls a complex web of studies and education-related programs spanning every corner of the American education system.

While most of its research is public, the Institute must safeguard data containing individually identifiable information. Assoc. Mot. Prelim. Inj. at 27–28. Researchers who want to see this “restricted-use” data must get a license from the Institute. The licenses can either be “physical” or “remote.” Id. at 28. Researchers with physical licenses get CDs that they can only use in “cold rooms,” meaning standalone desktop computers in secure locations disconnected from the internet. Id. Researchers with remote licenses can use a web-based portal through their own computer from anywhere. Id.

But things started to change earlier this year. On February 10, 2025, at the urging of the Department of Government Efficiency, the Institute began canceling large swathes of the contracts it uses to implement its studies and programs. Assoc. Mot. Prelim. Inj. at 21. Waves of cancellations followed in the days to come. See id. at 24–25. Many agency functions ground to a halt, including the National Postsecondary Student Aid Study, the National Assessment of Educational Progress, and the ECLS-K:2024 longitudinal study of kindergarteners, to name a few. See Assoc. Mot. Prelim. Inj. at 22–23 (identifying programs and corresponding contract award IDs that have been cancelled). But it was not just contracts getting axed. The Institute also announced in an email on February 10 that it is terminating the remote-access program for restricted-use data. See Voight Decl. Ex. 1, Assoc. ECF No. 6-10, at 11. And on March 11, the Secretary of Education announced a reduction-in-force (“RIF”) to reduce the agency's staffing by 50%, though Plaintiffs maintain the actual percentage ended up even higher. See Assoc. Mot. Prelim. Inj. at 26–27; Acad. Mot. Prelim. Inj. at 20. Finally, on March 20, the President issued an Executive Order directing the Secretary of Education to “facilitate the closure of the Department of Education” “to the maximum extent appropriate and permitted by law.” Exec. Order No. 14242, 90 Fed. Reg. 13679 (Mar. 20, 2025).

This litigation and the two preliminary injunction motions followed in late April. There are two cases, both against the Department of Education and the Secretary of Education (collectively, “the Department”). Each case has two Plaintiffs. Plaintiffs in the first case (No. 25-cv-999) are the Association for Education Finance and Policy and the Institute for Higher Education Policy. The first organization strives to “provide a forum for discussion and debate as to what drives effective education and related services” for people of all ages. Assoc. Compl. ¶ 9. The second focuses on “research, policy, and advocacy” surrounding access to postsecondary education. Id. ¶ 10. They and their members rely on Institute-produced data for these activities. See Assoc. Mot. Prelim. Inj. at 31–36.

Plaintiffs in the second case (No. 25-cv-1266) are the National Academy of Education and the National Council on Measurement in Education. The former works to “advance[ ] high-quality research to improve education policy and practice.” Acad. Compl. ¶ 12. The latter is a professional organization that promotes advancement in “assessment, evaluation, testing, and other aspects of educational measurement.” Id. ¶ 13. These Plaintiffs and their members also rely on data from the Institute in their activities. See Acad. Mot. Prelim. Inj. at 11–16, 18–19.

In both cases, Plaintiffs allege the Department is taking steps to “dismantle” the Institute. See, e.g., Assoc. Mot. Prelim. Inj. at 14 (The Department has started “to dismantle” the Institute.); Acad. Mot. Prelim. Inj. at 25 (The Department is taking “action to dismantle” the Institute.). But no Plaintiff challenges the Executive Order itself—that is, the order actually directing the drawdown of the Department. See Assoc. Reply, ECF No. 17, at 11 (“Plaintiffs do not challenge an executive order.”); Acad. Reply, ECF No. 23, at 15 (“Plaintiffs’ claims are not based on Presidential executive orders.”). Instead, Plaintiffs focus mostly on agency actions that predate the Executive Order. They challenge the decisions to cancel contracts, reduce staffing, and modify data access procedures, asserting those actions were arbitrary and capricious and violated the Institute's statutory mandates. See Assoc. Mot. Prelim. Inj. at 37; Acad. Mot. Prelim. Inj. at 20. All Plaintiffs invoke the APA, with another constitutional challenge from each. See Assoc. Mot. Prelim. Inj. at 2; Acad. Mot. Prelim. Inj. at 3.

After Plaintiffs moved for preliminary injunctions, the Court held a joint hearing. See Prelim. Inj. Hearing Tr. (“Hr'g Tr.”), Assoc. ECF No. 20 and Acad. ECF No. 25; see also Fed. R. Civ. P. 42(a) (authorizing joint hearings). All parties had an opportunity to be heard. Given the substantial overlap in factual and legal issues, the Court addresses both cases in this opinion, but it separately analyzes the distinct issues raised by the two cases when necessary. Accord Cntr. for Biological Diversity v. Trump, 453 F. Supp. 3d 11, 21 n.1 (D.D.C. 2020).

II.

A preliminary injunction is “an extraordinary remedy never awarded as of right,” but as an exercise of discretion by a court sitting in equity. Winter v. Nat. Res. Def. Council, Inc., 555 U.S. 7, 24, 129 S.Ct. 365, 172 L.Ed.2d 249 (2008). The movant faces a high bar for success, as it must establish four elements “upon a clear showing”: (1) that it is likely to succeed on the merits; (2) that it likely will suffer irreparable harm in the absence of injunctive relief; (3) that the balance of equities supports granting the relief; and (4) that the public interest favors the injunction. Id. at 20, 22, 129 S.Ct. 365. Where, as here, the Government is the party opposing injunctive relief, the latter two factors “merge.” Nken v. Holder, 556 U.S. 418, 435, 129 S.Ct. 1749, 173 L.Ed.2d 550 (2009).

While a successful plaintiff must make a clear showing across all four elements, a court need not necessarily consider each prong to deny relief. If a plaintiff fails to show that it is likely to succeed on the merits, a court may deny the preliminary injunction request without reaching the remaining factors. See Arkansas Dairy Co-op Ass'n, Inc. v. U.S. Dep't of Agr., 573 F.3d 815, 832 (D.C. Cir. 2009) (taking this approach); Guedes v. Bureau of Alcohol, Tobacco, Firearms & Explosives, 920 F.3d 1, 10 (D.C. Cir. 2019) (same); see also Winstead v. EMC Mortg. Corp., 621 F. Supp. 2d 1, 4–5 (D.D.C. 2009) (same).

III.

Neither the Association nor the Academy has shown the APA is a viable cause of action for their claims. The APA can ferry discrete challenges to particular agency actions into court. But Plaintiffs pile aboard so many broad complaints that their claims sink underneath the sheer weight of the relief they seek. The APA was not built to vindicate programmatic challenges to day-to-day agency operations, which is precisely what Plaintiffs assert.

As for the constitutional claims, they are constitutional in name only. They simply repackage the impermissible APA claims and try to sneak them in through a constitutional back door. But that door is locked by precedent. So Plaintiffs have not made a “clear showing” that they are likely to succeed on the merits of their claims. Winter, 555 U.S. at 22, 129 S.Ct. 365. The Court will thus deny both motions for preliminary injunctions.

A.

Under the APA, a person “suffering legal wrong because of agency action ․ is entitled to judicial review thereof.” 5 U.S.C. § 702. Courts are empowered to “hold unlawful and set aside” a challenged agency action if it is “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.” Id. § 706(2). To be reviewable, an agency action must be “final,” meaning it (1) “mark[s] the consummation of the agency's decisionmaking process” and (2) determines “rights or obligations” or “legal consequences will flow” from it. Bennett v. Spear, 520 U.S. 154, 177–78, 117 S.Ct. 1154, 137 L.Ed.2d 281 (1997) (cleaned up).

Courts “have long recognized that the term [agency action] is not so all-encompassing as to authorize ․ judicial review over everything done by an administrative agency.” Indep. Equip. Dealers Ass'n v. EPA, 372 F.3d 420, 427 (D.C. Cir. 2004) (cleaned up). The APA is a remedial scalpel, used to excise “circumscribed [and] discrete” improper actions. Norton v. S. Utah Wilderness All., 542 U.S. 55, 62, 124 S.Ct. 2373, 159 L.Ed.2d 137 (2004). But it is the wrong tool for plaintiffs hoping to bludgeon through broad “programmatic” attacks that “demand a general judicial review of [an agency's] day-to-day operations.” Lujan, 497 U.S. at 899, 110 S.Ct. 3177.

Some examples help shade these contours. Start with Lujan itself. There, the Bureau of Land Management (“BLM”) had a “land withdrawal review program.” 497 U.S. at 877, 110 S.Ct. 3177. The program was a collection of individual actions by which the BLM classified federally owned land into categories to designate how it could be used. See id. at 877–78, 110 S.Ct. 3177. Enter the plaintiffs. At bottom, they wanted to invalidate the BLM's classification decisions. Id. at 879, 110 S.Ct. 3177. So they mounted an APA challenge against the agency's classification program as a whole, alleging widespread deficiencies. See id. at 891, 110 S.Ct. 3177. The end goal was not merely to invalidate a few specific decisions but to vitiate all of the agency's classifications decisions. Id. at 879, 110 S.Ct. 3177.

The Supreme Court rejected the attempt. Why? Because “flaws in [an] entire ‘program’—consisting principally of the many individual actions referenced in the complaint, and presumably actions yet to be taken as well—cannot be laid before the courts for wholesale correction under the APA.”² Id. at 893, 110 S.Ct. 3177. The land classification program was not itself challengeable because it was not a “single [agency] order or regulation, or even ․ a completed universe of particular [agency] orders and regulations.” Id. at 890, 110 S.Ct. 3177. And while the plaintiffs identified some specific decisions and showed harm from those decisions, that did not entitle them to relief from “all such decisions.” Id. at 882, 110 S.Ct. 3177 (emphasis added). The power to issue that sort of holistic change to agency operations belongs not to courts but to “the offices of the Department or the halls of Congress, where programmatic improvements are normally made.” Id. at 891, 110 S.Ct. 3177.

Another group of plaintiffs launched a similar challenge in Sierra Club v. Peterson, 228 F.3d 559 (5th Cir. 2000). They objected to a Forest Service program that allowed “even-aged timber management” when the government sold lumber—a technique where all the trees in a given area are cut down at the same time. Id. at 562. The plaintiffs identified some allegedly improper sales under this program but did not limit their challenge to just those individual sales. Like the Lujan plaintiffs, they tried to use these individual sales as bellwethers and then extrapolate the remedy to modify the entire timber management process. See id. at 566 (They “challenged past, ongoing, and future timber sales.”). But the court was not having it. The Fifth Circuit rejected the plaintiffs’ attempt to “challenge an entire program by simply identifying specific allegedly improper final agency actions within that program.” Id. at 567 (emphasis added). Applying Lujan, the court concluded that large events comprised of many individual actions cannot be challenged “simply because one [action] ․ is ripe for review [and] adversely affects one of” the plaintiffs. Id. at 568 (quoting Lujan, 497 U.S. at 893, 110 S.Ct. 3177). Thus, the proper course is to challenge only those “circumscribed [and] discrete” actions within a program—not the entire program and every decision made under it. S. Utah Wilderness All., 542 U.S. at 62, 124 S.Ct. 2373.

In sum, plaintiffs cannot get “wholesale” improvement to agency operations under the APA by bundling together a collection of discrete actions, presenting them as facets of a single event, and then challenging that event instead of the individual actions. See Lujan, 497 U.S. at 893, 110 S.Ct. 3177. And plaintiffs cannot circumvent this restriction by purporting to challenge one or two discrete actions but then extrapolating the requested relief across the board so that it amounts to “wholesale” improvement. See Peterson, 228 F.3d at 567–68.

A recent D.C. Circuit decision is illustrative. See Widakuswara v. Lake, 2025 WL 1288817, at *3 (D.C. Cir. May 3, 2025) (per curiam). Much as the Lujan plaintiffs tried to unify a series of discrete actions under a land management program, the Widakuswara plaintiffs sought to cast a series of individual personnel and grant actions as the embodiments of a decision to “dismantle” an agency. See id. While they identified some specific actions, they sought to leverage these individual challenges into broad relief that would prevent “wholesale shuttering” of the agency. Id. The court recognized that many of the “individual” actions the plaintiffs challenged might allow for “viable, discrete claims.” Id. But “[t]he ‘dismantling’ that [the] plaintiffs allege[d] [was] a collection of ‘many individual actions’ that [could not] be packaged together and ‘laid before the courts for wholesale correction under the APA.’ ” Id. (quoting Lujan, 497 U.S. at 893, 110 S.Ct. 3177); accord Found. on Economic Trends v. Lyng, 943 F.2d 79, 86 (D.C. Cir. 1991).

To be sure, Widakuswara was a non-precedential emergency stay decision.³ Id. at *1. Still, it is a helpful representation of the APA's limitations in a case that bears resemblance to this dispute. And it is to that dispute that the Court now turns.

B.

1.

First on deck is the Association's case (No. 25-cv-999). The Association challenges four agency actions: (1) “the decision to terminate programs by canceling contracts en masse”; (2) “the decision to terminate the Regional Educational Laboratory program”; (3) the “reduction-in-force affecting 90% of [the Institute's] staff”; and (4) the decision to “terminate[ ] the remote access program for restricted-use data.” Assoc. Mot. Prelim. Inj. at 37. Despite purporting to target four individual actions, the Association's framing of its case gives away the game and lays bare the true scope of its challenge.

The Association repeatedly presents this case as an attempt to stop the wholesale “dismantling” of the Institute. Assoc. Mot. Prelim. Inj. at 36 (alleging “the injuries to [the Association] ․ are caused by Defendants’ dismantling of [the Institute]”); id. at 46 (“Defendants’ actions to dismantle [the Institute] are contrary to” various statutes.); id. at 49 (“Defendants’ actions effectively dismantle [the Institute].”); id. at 49 (accusing the Department of taking “actions to shut down a statutorily created agency”). As the Association succinctly put it: “While Defendants have not completed the dismantling of [the Institute], they have taken at least four discrete agency actions to carry out that goal”—the four actions the Association challenges. Id. at 37.

The Complaint bundles these four actions together as well. Count II lists them together rather than dividing them into individual challenges. See Assoc. Compl. ¶¶ 89–99. And Count III, which alleges arbitrary and capricious action, does not even list the four specific actions at all. See id. ¶¶ 100–02. Instead, it lodges a general allegation that the Department acted “arbitrarily and capriciously by attempting to render an executive agency incapable of carrying out the statutory functions that Congress has directed it to fulfill.” Id. ¶ 102.

In its quest to revivify the Institute, the Association seeks a remedy larger than the sum of the individual injuries it alleges. The overall goal is not incremental correction by fixing specific erroneous decisions, but a grand decree blocking all actions that risk “render[ing] an executive agency incapable of carrying out [its] statutory functions.” Id. To that end, the Association wants a sweeping order that “reinstate[s] any ․ research, data collection, or dissemination activity that was terminated, halted, or paused” and imposes tight restrictions on when and how the Institute can terminate any program in the future. Assoc. Rev. Prop. Order., ECF No. 12-3, ¶ 2. That relief goes far beyond asking the Court to “hold unlawful and set aside” specific actions because they were flawed. 5 U.S.C. § 706(2). Staffing levels, contract decisions, and data access procedures are merely paths by which the Association seeks to dictate large-scale alterations in day-to-day agency operations.

But the Association cannot contest the Department's process of dismantling the Institute “by simply identifying specific allegedly improper final agency actions within that program.” Peterson, 228 F.3d at 567 (emphasis added). To the extent that unilateral dismantlement (or large-scale reorganization) is underway, it is not challengeable as a final agency action any more than the land classification process in Lujan was challengeable. And it cannot be “laid before the courts for wholesale correction under the APA.” Lujan, 497 U.S. at 893, 110 S.Ct. 3177. The Department argues as much in both cases. Opp'n to Assoc. Mot., ECF No. 16, at 32–33; Opp'n to Acad. Mot., ECF No. 21, at 22–23.

The programmatic nature of the Association's challenge dooms its motion. The Court need go no further to deny the preliminary injunction as to these APA claims. But the problems with the Association's case do not end there, and explaining the other flaws helps further illuminate why this is a programmatic challenge.

2.

Examining the purported final agency actions drives the point home. Start with the contract cancellations. The Association gestures at a “decision” to “terminate programs by canceling contracts en masse.” Assoc. Mot. Prelim. Inj. at 37. Yet it is unclear from the Association's facts that these cancellations all stem from a single decision. Indeed, its timeline suggests otherwise. To the extent that there was a single decision, the Association has not identified it so the Court can hardly review it.

Remember, the contracts were not all axed in one fell swoop. Most cancellations were clustered between February 9 and February 11. See Assoc. Mot. Prelim. Inj. at 21. But others happened later. The Association specifically mentions contract cancellations on February 13, February 19, and even March 10. Id. at 24–25. And at least one cancelled contract was reinstated in late February, suggesting the agency was continuing to evaluate contracts in some capacity after February 10. Id. at 25. All of this indicates that the Association is improperly bundling together a series of actions for wholesale correction rather than challenging one “circumscribed [and] discrete” action. S. Utah Wilderness All., 542 U.S. at 62, 124 S.Ct. 2373. The Association's strategy echoes those unsuccessfully fielded by the Lujan and Peterson plaintiffs. It is trying to use a few specific contract cancellations as bellwethers to bring about large scale changes—mass program reimplementation and general statutory compliance—across the Institute.

The proposed preliminary injunction order also reflects a programmatic focus. The Association asks not just for reinstatement of contracts, but also reinstatement of “any ․ research, data collection, or dissemination activity that was terminated, halted, or paused.” Assoc. Rev. Prop. Order. ¶ 2. The supremacy of programs, not contracts, is further evident from the Association's uncertainty as to exactly how many contract terminations are included in its challenge. See Assoc. Mot. Prelim. Inj. at 21 (perhaps 90 but maybe “as many as 170 contracts”). The briefing points that direction too, with headers that focus on the big-picture “termination of [the Institute's] research and dissemination activities”—not the termination of its contracts. Assoc. Mot. Prelim. Inj. at 40; see also id. at 46. All of this indicates the true focus is wholesale correction, not remediation for a narrow band of erroneous decisions.

The Supreme Court's warning in Southern Utah rings true here: “If courts were empowered to enter general orders compelling compliance with broad statutory mandates, they would necessarily be empowered, as well, to determine whether compliance was achieved.” 542 U.S. at 66, 124 S.Ct. 2373. Such a framework would inevitably devolve into judicial supervision, and it would become “the task of the supervising court, rather than the agency, to work out compliance with the broad statutory mandate.” Id. at 66–67, 124 S.Ct. 2373. Judges were never meant to be czars overseeing the day-to-day affairs of agencies. Thus, the Association's challenge to the mass termination of programs via contract cancellations is precisely the sort of broadside attack that Lujan forbids. And the general statutory compliance it seeks as relief runs counter to Southern Utah. More, this objection to scores—hundreds?—of cancelled contracts is just one of the four targets of the Association's ire.

The Association's challenge to the RIF fares no better. It contests the “reduction-in-force generally, not ․ anything particular about how it was carried out.” Assoc. Reply at 10–11 (cleaned up). Unsurprisingly then, the primary focus is not on the RIF itself, but “the cessation of ․ activities [caused] by the reduction-in-force.” Id. at 11 (emphasis added). In other words, the Association does not assert the agency lacked the authority to conduct the RIF, that the agency violated procedural requirements in carrying out the RIF, or that it or its members have any direct stake in the underlying personnel actions.⁴

That leaves the Association to challenge the diasporic results of the RIF, rather than the decision to implement it. See Assoc. Mot. Prelim. Inj. at 42, 46 (warning the RIF will result in widespread statutory deficiencies). In essence, the Association asks the Court to decide not whether the RIF was carried out properly, but whether it should have been carried out at all. It is not the place of this Court to substitute its own judgment for that of the agency on staffing issues. Doing so would be tantamount to “general judicial review of the [Institute's] day-to-day” employment decisions. Lujan, 497 U.S. at 899, 110 S.Ct. 3177. The Association is “merely us[ing]” the RIF “as evidence to support [its] sweeping argument that the [Institute's] ‘on-the-ground’ management” of its statutorily mandated programs is deficient. Peterson, 228 F.3d at 567. Lujan closed the door on challenges like this—and Southern Utah nailed it shut.

Next, the “decision to terminate the Regional Educational Laboratory program.” Assoc. Mot. Prelim. Inj. at 37. The Association does not even mention the Laboratory terminations in its proposed injunction, much less seek any specific relief as to that decision. See Assoc. Rev. Prop. Order. The lack of a request for an independent remedy underscores that the Laboratory terminations are really one piece of a broader challenge to the “dismantling” of the Institute.

Finally, consider the termination of the remote access program. See Assoc. Mot. Prelim. Inj. at 37. On the surface, this challenge looks more like standard APA fare. It targets one clearly identifiable decision to terminate one discrete program, see id., and the proposed injunction mentions that decision. See Assoc. Rev. Prop. Order. ¶ 7. So the claim might be viable if brought as a standalone APA challenge. The problem for the Association is that it did not bring the remote access claim as a standalone challenge. See Found. on Economic Trends, 943 F.2d at 86 (“[T]he many individual actions referenced in the record ․ cannot be laid before the courts for wholesale correction even if one happens to be ripe for review and is adversely affecting one of the plaintiffs ․”) (cleaned up).

Remember, Count II (contrary to law) bundles remote access in with the other issues while challenging the Department's overall “decision to end the programs and activities of [the Institute].” See Assoc. Compl. ¶ 90; id. ¶¶ 89–99. And Count III (arbitrary and capricious) focuses on the “attempt[ ] to render an executive agency incapable of carrying out ․ statutory functions” without directly mentioning remote access. Assoc. Compl. ¶ 102. When viewed in context, the remote access termination is being used as evidence of the Association's bigger contention that “violation of the law is rampant within” the attempt to dismantle the Institute. Lujan, 497 U.S. at 891, 110 S.Ct. 3177.

True, the Association has “identified some specific [decisions] in [its] pleadings that [it] argue[s] are final agency actions.” Peterson, 228 F.3d at 567. And true, the APA lets the Association “challenge a specific final agency action which has an actual or immediately threatened effect.” Id. (cleaned up). But daisy-chaining those premises together still does not let the Association “challenge an entire program by simply identifying specific allegedly-improper final agency actions within that program, which is precisely what [it] did here.” Id.

3.

It should come as no surprise that Lujan bars the Association's way. The Court forecasted these concerns at oral argument and asked the Association for its best cases to show why Lujan does not prevent its APA challenges. Hr'g Tr. at 30:17–19. The Association gave two cases: Meina Xie v. Kerry, 780 F.3d 405 (D.C. Cir. 2015) and American Fuel & Petrochemical Manufacturers v. Environmental Protection Agency, 937 F.3d 559 (D.C. Cir. 2019). Hr'g Tr. at 30:20–31:8. Neither does much for it.

Xie was a visa mandamus case involving an attempt to compel agency action under the APA. 780 F.3d at 405. The case does not cite Lujan or consider its ban on programmatic challenges. Xie does, however, briefly discuss Southern Utah. Id. at 407–08. Its takeaway from Southern Utah is that courts can compel agencies to act if action is legally required, but plaintiffs cannot ask for a general “compliance” order that requires judicial management of daily agency tasks. See id. at 408. Nothing in that conclusion conflicts with the Court's decision today.

American Fuel at least mentions Lujan and programmatic challenges, albeit briefly. The petitioners challenged an Environmental Protection Agency final rule involving renewable energy production targets for the fuel market. Am. Fuel, 937 F.3d at 568. They protested that the agency did not “comply with the requirements” in the Endangered Species Act because it failed to consult another agency to ensure the rule would not harm endangered species. Id. at 591. Although the petitioners criticized a broader renewable energy program the final rule was part of, “their actual challenge [was] to the ․ Rule.” Id. So the challenge was not programmatic because it targeted one agency action—the promulgation of a final rule—and alleged that the agency's procedure in making the rule violated one provision in the Endangered Species Act. See Pet. ¶ 2, No. 18-1040 (D.C. Cir. Feb. 9, 2018). That contrasts with the Association's contention that the Department violated at least a dozen different statutory provisions by canceling one-hundred-plus contracts, instituting a RIF, terminating ten separate Regional Educational Laboratories, and modifying data access procedures. The Association's case bears far more resemblance to Lujan than to American Fuel. So neither of the Association's two strongest cases save it.

4.

The Association also tries its luck at a separation of powers claim grounded in the same conduct as its APA challenges.⁵ But the Supreme Court has rejected the notion “that every action by the President, or by another executive official, in excess of his statutory authority is ipso facto in violation of the Constitution.” Dalton v. Specter, 511 U.S. 462, 472, 114 S.Ct. 1719, 128 L.Ed.2d 497 (1994). To the contrary, courts “often distinguish[ ] between claims of constitutional violations and claims that an official has acted in excess of his statutory authority.” Id. The Constitution is implicated when the challenged executive acts are taken by “inherent constitutional” authority or when the statute the executive relies on for authority is itself unconstitutional. See id. at 473 & n.5, 114 S.Ct. 1719. That is not the case here because the Association's claim is based on executive officials violating statutory requirements relating to the Institute's mission.

The Association still contends that the overall decision to shut down an agency exists outside of statutory considerations, because only Congress has the “power to make a law” that shutters an agency. See Assoc. Mot. Prelim. Inj. at 49 (quoting Wilkerson v. Rahrer, 140 U.S. 545, 562, 11 S.Ct. 865, 35 L.Ed. 572 (1891)). But in the same breath, the Association says the Department acted “in direct contravention of both the statutes that establish [the Institute] and its functions, and of 20 U.S.C. § 3473(a), which prohibits the Secretary from ‘aboli[shing]’ any statutory entity within [the Education Department].” Id. (emphasis added). So it is truly alleging a statutory violation, not a constitutional issue. Plus, the Association expressly disavows any challenge to the Executive Order, the only thing resembling a shutdown order that it has flagged. Assoc. Reply at 11; Opp'n to Assoc. Mot. at 33–34. The Association is left instead to rely on the same concern that animates its APA claims: the Institute is allegedly not performing functions that statutes require it to perform. The attempt at a constitutional claim is just a repackaging of the APA challenge. Dalton forecloses that option, as “claims simply alleging that the [executive branch] has exceeded [its] statutory authority are not ‘constitutional’ claims.” 511 U.S. at 473, 114 S.Ct. 1719. Thus for the Association's so-called constitutional claim, it is unlikely to succeed on the merits because it fails to state a claim for relief altogether.

* * *

At this early stage, the Court is unconvinced the APA gets the Association where it wants to go. The Association's plea for the Court to step in and stop the “dismantling” of the Institute is a request for the sort of programmatic relief reserved for “the offices of the Department or the halls of Congress.” Lujan, 497 U.S. at 891, 110 S.Ct. 3177. True, some of the individual actions the Association identifies may be ripe for standalone challenges, such as the end of the remote access program. But as currently structured, these individual claims are impermissibly bundled as part of a larger programmatic challenge. The Association has chosen to swing for the fences, in hopes it will get everything it wants. But that is a losing gambit. “The ‘dismantling’ that plaintiffs allege is a collection of ‘many individual actions’ that cannot be packaged together and ‘laid before the courts for wholesale correction under the APA.’ ” Widakuswara, 2025 WL 1288817, at *3 (quoting Lujan, 497 U.S. at 893, 110 S.Ct. 3177). At the preliminary injunction stage, it is on the Association to make “a clear showing” that it has a cause of action under the APA. Winter, 555 U.S. at 22, 129 S.Ct. 365. It has not met that burden.⁶

C.

Turn next to the Academy's case (No. 25-cv-1266). The second case encompasses distinct claims and different Plaintiffs, but the analysis follows a similar trajectory. The Academy challenges three things: “(1) the February 10, 2025 announced cancellation of over $900 million in [Institute] contracts, some of which are critical to the basic functioning of [the Institute] and to the realization of many ․ congressional mandates; (2) the March 11, 2025 announced RIF of 90% of [Institute] staff including 97% of NCES staff; and (3) the February 2025 termination of all remote restricted-use data licenses and the end of accepting new applications for these licenses.” Acad. Mot. Prelim. Inj. at 20. But again, the framing and the requested relief give away the game. This is a programmatic challenge in disguise.

The Academy contends that the Department “took action to dismantle [the Institute].” Id. at 25. And each discrete decision was a step toward the “goal of dismantling [the Institute].” Id. at 6. So the Academy says the harms it is experiencing “are traceable to Defendants’ dismantling of [the Institute].” Id. at 19. The Complaint follows the same pattern, focusing on the Department's “dismantlement” of the Institute as the primary issue. See Acad. Compl. ¶¶ 112, 117 (Count II); id. ¶¶ 122, 123 (Count III). The focus of the relief is also general statutory compliance, not the individual actions the Academy purports to challenge.

For instance, the Academy explains that it “only reference[s] contracts because” the Institute relies “on contracts ․ to accomplish [its] statutorily required duties.” Acad. Reply at 4. The “requested relief seeks to ensure that [the Institute] can operate as statutorily required ․ without regard to whether any one contract is in effect or not.” Id. at 3. In fact, the Academy surmises, the contracts would not need to be reinstated if the Institute just “hire[d] sufficient ․ employees to accomplish [its] statutory duties.” Id. at 4. Put another way, the Academy is challenging the decision to cancel contracts, but it does not really care about the decision to cancel contracts. It cares about the Institute's failure to perform a wide array of statutorily mandated functions.

The same holds true for the decision to end remote access for restricted-use data. See Acad. Mot. Prelim. Inj. at 20. It is telling that the Academy's proposed injunction makes no mention at all of the remote-access program. See Acad. Prop. Order, ECF No. 14-34. Instead, it wants the Institute to restore “sufficient employee positions” to “ensure that,” among other things, the Institute “grant[s] and renew[s] restricted-use data licenses.” Id. It would be a surprising thing for the Academy to argue a decision was erroneous but then not request a corresponding remedy if the issue were truly an independent challenge. But these purportedly individual challenges are merely alter egos of the Academy's holistic dissatisfaction with daily agency operations.

Then there is the Tucker Act. When a dispute involves a government contract, district courts must undertake a specific analysis to decide whether the Tucker Act divests the court of jurisdiction. See Transohio Sav. Bank v. Dir., Off. of Thrift Supervision, 967 F.2d 598, 609 (D.C. Cir. 1992) (“[T]his Court and others have interpreted the Tucker Act as providing the exclusive remedy for contract claims against the government.”). The Tucker Act requires case-by-case assessment based on the source of a plaintiff's rights and the nature of the relief sought. See Albrecht v. Comm. on Emp. Benefits of Fed. Rsrv. Emp. Benefits Sys., 357 F.3d 62, 68 (D.C. Cir. 2004). For instance, asking a court to reinstate a contract—the relief sought here—can look a lot like a jurisdictionally improper order for specific performance. See Ingersoll-Rand Co. v. United States, 780 F.2d 74, 80 (D.C. Cir. 1985) (“[A] request for specific performance must be resolved by the Claims Court.”); accord U.S. Conf. of Catholic Bishops v. U.S. Dep't of State, ––– F. Supp. 3d ––––, ––––, 2025 WL 763738, at *5 (D.D.C. 2025). Maybe it is, maybe not. But either way, courts at least need to ask these questions when dealing with government contract disputes.

The Tucker Act looms large here because some Academy members perform work under Institute contracts and could likely show harm from the cancellation of those contracts. See Decl. of Andrew Ho, Acad. ECF No. 14-12, ¶ 21; Decl. of Derek Briggs, Acad. ECF No. 14-14, ¶ 9, 13. It is unclear how many other members are party to Institute contracts or whether the organizations themselves have contracts. But the Court would likely lack jurisdiction to hear any individual contract challenge brought by the organizations or their members if it resembled a traditional contract claim. See Albrecht, 357 F.3d at 67–68 (explaining that the Tucker Act extends to contract cases involving “injunctive relief”). In other words, the Tucker Act might bar members from individually seeking the relief that the Academy now seeks for all of them. Allowing potentially prohibited individual claims to move forward simply because the Academy has bundled them together risks vitiating the jurisdictional limitations Congress imposed in the Tucker Act. Given the “extraordinary” and discretionary nature of the relief sought, see Winter, 555 U.S. at 24, 129 S.Ct. 365, this Court is reluctant to take such a step.

Like the Association, the Academy is seeking wholesale correction of general agency operations and bundling together claims to pursue outsized relief. It wants a court order for the Institute to “reinstate and maintain all congressionally mandated data collection, analysis, and dissemination activities.” Acad. Prop. Order ¶ 2. But that relief necessarily involves a muddled “general order[ ] compelling compliance with broad statutory mandates.” S. Utah Wilderness All., 542 U.S. at 66, 124 S.Ct. 2373. Such an order would inevitably lead to enforcement disputes and require the Court “to determine whether compliance was achieved.” Id. The Court cannot scrutinize the minutiae of how the Institute's multitudes of studies are carried out. Such questions are best left to “expert judgment of agencies specifically created to deal with complex and technical issues.” Peterson, 228 F.3d at 566. In short, the relief the Academy desires cannot be found in a courtroom. They are instead the province of “the halls of Congress, where programmatic improvements are normally made.” Lujan, 497 U.S. at 891, 110 S.Ct. 3177. Thus, the Academy has not made “a clear showing” that it has a cause of action under the APA. Winter, 555 U.S. at 22, 129 S.Ct. 365.

* * *

Separate from its APA claims, the Academy also asserts a separation of powers violation.⁷ It tries to season it differently by sprinkling in the Take Care Clause, but it still tastes the same as the APA claims. See Acad. Mot. Prelim. Inj. at 26–29. The focus remains a generalized displeasure over the Institute's failure to comply with its statutory mandates.

At the start, the Court notes that “[w]hether claims brought directly under the Take Care Clause are even justiciable is open to debate.” Citizens for Resp. & Ethics in Wash. v. Trump, 302 F. Supp. 3d 127, 130 (D.D.C. 2018), aff'd, 924 F.3d 602 (D.C. Cir. 2019). Even if such claims are justiciable, the Academy's Complaint does not raise the Take Care Clause—only a separation of powers claim premised on the executive's attempt to “dismantle” the agency. See Acad. Compl. ¶ 129; see also Opp'n to Acad. Mot. at 32 (pointing this out). So the Academy cannot now add a new basis for relief in its motion for a preliminary injunction. See Sai v. Transp. Sec. Admin., 54 F. Supp. 3d 5, 9 (D.D.C. 2014). Plus, both the Take Care Clause and the separation of powers issues are just doppelgängers of the Academy's APA claim. The Academy simply reiterates how the executive is “hamstringing [the Institute's] ability to fulfill its congressional mandates.” Acad. Mot. Prelim. Inj. at 27. As the Department argues, if every statutory compliance issue triggered separation of powers concerns, virtually all APA cases would balloon into disputes of constitutional proportions. See Opp'n to Acad. Mot. at 31. Again, not every statutory violation by an executive agency “is ipso facto in violation of the Constitution.” Dalton, 511 U.S. at 472, 114 S.Ct. 1719. The Court will not rush to read constitutional issues into APA claims simply because a plaintiff invokes constitutional language. And to the extent that the Executive Order might be more susceptible to constitutional claims, the Academy does not challenge it. See Acad. Reply at 2. So the Academy is unlikely to succeed on the merits of a separation of powers or Take Care Clause challenge because it fails to state a plausible claim for relief.

D.

In sum, Plaintiffs’ failure to show that they are likely to succeed on the merits is fatal to their motions. See Arkansas Dairy, 573 F.3d at 832. This deficiency is weighty. Even assuming that Plaintiffs could prevail on the remaining factors, the overall balance still tips sharply against granting a preliminary injunction. See City of Las Vegas v. Lujan, 891 F.2d 927, 931 (D.C. Cir. 1989) (explaining that district courts have “latitude” in balancing the factors, and that balancing receives deference on appeal). After all, without a likelihood of success, there is “no justification for the court's intrusion into the ordinary processes of administration and judicial review.” Am. Bankers Ass'n v. Nat'l Credit Union Admin., 38 F. Supp. 2d 114, 140 (D.D.C. 1999) (quoting Washington Metro. Area Transit Comm'n v. Holiday Tours, Inc., 559 F.2d 841, 843 (D.C. Cir. 1977)).

It would be easier for Plaintiffs if they could resolve their array of concerns through a single APA action with sweeping relief. It would probably drive some measure of judicial efficiency too. But for better or worse, the Court is bound by the scope-of-review limitations Congress has imposed on the APA. These limitations are necessary to “avoid encroaching on the other branches of government.” Peterson, 228 F.3d at 566. Fidelity to the law, not convenience, is the Court's North Star. As Plaintiffs themselves recognize in their briefs, separation of powers is vital. Treating the APA as a panacea for whatever ails society would erode that separation and call up the specter of judicial overreach.

IV.

For these reasons, Plaintiffs’ motions for preliminary injunctions will both be denied. Separate orders will issue in each case.

FOOTNOTES

1.   The Court collectively refers to Plaintiffs in the first case as “the Association” and Plaintiffs in the second case as “the Academy.” The Court also uses the abbreviations “Assoc.” (Case No. 25-cv-999) and “Acad.” (Case No. 25-cv-1266) in record citations to distinguish between the two sets of filings. Citations are to CM/ECF pagination where possible.

2.   To be clear, there is a difference between improperly challenging a “program” that is not itself an action and properly challenging a “specific order or regulation” that “appl[ies] some particular measure across the board to” downstream agency decisions. Lujan, 497 U.S. at 890 n.2, 110 S.Ct. 3177. Plaintiffs identify no such specific order here. The most obvious candidate—the Executive Order—post-dates most of the actions at issue here and has not been challenged by any Plaintiff.

3.   In Widakuswara, an appellate panel stayed two injunction provisions requiring the government to reinstate employees and restore grants. 2025 WL 1288817, at *1. A series of Circuit en banc orders—also non-precedential—declined to reconsider the stay of employee reinstatements but did allow the grant restoration provision to take effect. See Assoc. Not. Suppl. Auth., ECF No. 22. In doing so, the orders did not discuss Lujan or programmatic challenges. See id. at 4–5 (May 22 order on employee reinstatements), 15–17 (May 28 order on grant restorations). Notwithstanding these subsequent unpublished developments, the Court finds the Widakuswara panel's Lujan discussion persuasive and considers it accordingly.

4.   Indeed, it appears unlikely that either organization or their members would have standing to challenge the RIF itself. After all, they have no direct stake in the Institute's employment decisions because “the plaintiffs ․ lack ․ employment relationships with the federal government.” Assoc. Reply at 1. And even if they did have an employment relationship, “federal employees may not use the [APA] to challenge agency employment actions.” Filebark v. U.S. Dep't of Transp., 555 F.3d 1009, 1010 (D.C. Cir. 2009).

5.   The Association makes a passing reference to the Department's actions being ultra vires too. Assoc. Prelim. Inj. Mot. at 49. But it does not develop this argument so the Court does not address it. See New York Rehab. Care Mgmt., LLC v. N.L.R.B., 506 F.3d 1070, 1076 (D.C. Cir. 2007) (declining to consider an argument that was “included [in] a heading” but not discussed further because “[i]t is not enough merely to mention a possible argument in the most skeletal way, leaving the court to do counsel's work.”).

6.   As part of its “contrary to law” argument under the APA, the Association also claims violations of the Impoundment Control Act and the Anti-Deficiency Act. Assoc. Mot. Prelim. Inj. at 48. Having already found that the Association's claims are likely programmatic, the Court does not address these separately because they are part of the same wholesale challenge.

7.   The Complaint and reply brief include an ultra vires claim too, but the Academy never mentioned it in the initial preliminary injunction motion. See Acad. Mot. Prelim. Inj. at 20–29 (discussing the APA and separation of powers only); Acad. Reply at 17 (arguing a likelihood of success on the ultra vires claim). It is a “basic precept that arguments generally are forfeited if raised for the first time in reply” or if a party was initially “obscure on the issue ․ and only warmed to the issue in their reply brief.” Twin Rivers Paper Co. LLC v. Sec. & Exch. Comm'n, 934 F.3d 607, 615 (D.C. Cir. 2019) (cleaned up). So the ultra vires claim is forfeited for purposes of the preliminary injunction because the Court is now ruling only on the Academy's motion, not the Complaint.

TREVOR N. McFADDEN, United States District Judge