The Costs of Aggregating Administrative Claims

Introduction

Aggregation 1 Class actions, in which named parties bring claims on behalf of many different people, are the most famous kind of aggregate proceeding. But courts have found numerous other ways of aggregating disputes, including bankruptcy and trustee suits, statistical sampling, and parens patriae actions by states on behalf of certain citizens. What all these processes share is that a single party or proceeding binds numerous parties to the outcome of the proceeding, even if some of those parties never actually participate in the proceeding. has emerged in the past few years as a critical tool by which agencies can quickly resolve groups of claims that would otherwise languish for years in bureaucratic limbo. 2 See infra Part I. The idea is simple: Consolidating many similar claims in a single proceeding would help agencies process claims more quickly, efficiently, and fairly. 3 See Michael Sant’Ambrogio & Adam Zimmerman, Admin. Conference, Aggregate Agency Adjudication 9-10 (2016) (urging agencies to “avail themselves of the benefits of aggregate adjudication in the service of their statutory mandates”), https://perma.cc/NHY7-B2YY; see also Monk v. Shulkin, 855 F.3d 1312, 1320 (Fed. Cir. 2017) (suggesting that agencies could use class proceedings to “promot[e] efficiency, consistency, and fairness, and improv[e] access to legal and expert assistance by parties with limited resources”).

But aggregation presents unique challenges in the administrative context with respect to agencies that dole out entitlements. Such agencies, of course, have limited resources. An Administrative Law Judge (ALJ) with the authority to aggregate entitlement claims is thereby armed with the power to prioritize how those resources will be spent—whether in line with the agency’s own priorities or not.

The power to aggregate could therefore grant ALJs the ability to undermine agencies’ deliberate policy decisions. Using data reported by the Department of Veterans Affairs (VA) and the Department of Housing and Urban Development (HUD), we show that without bureaucratic delay, those agencies’ abilities to offer critical services could be hampered. From this view, aggregation threatens to prevent an agency from financing its core operations. Scholars and judges have long argued that bureaucratic delay may be an unfortunate effect of agencies’ resource limitations. We suggest that delay may also be a mechanism for allocating those limited resources in the way that best suits the agencies’ goals.

This Essay does not dispute the usefulness of artfully aggregating administrative claims. Rather, we propose that agencies develop procedures to control the costs of group hearings. Specifically, we suggest three ways agencies could control the costs of group claims: First, agencies could disaggregate prohibitively expensive claims. Second, they could delay hearings that might result in unaffordable outlays. Third, they could control the costs of group claims by requiring that certain officials authorize or initiate group proceedings.

Finally, we suggest that these disaggregation procedures would likely incentivize agencies to aggregate claims more frequently. Currently most agencies do not use aggregation techniques at all, and only a few use them regularly. 4 See Michael Sant’Ambrogio & Adam Zimmerman, Inside the Agency Class Action, 126 Yale L.J. 1634, 1658-59 (2017) (“[W]e found that very few agencies use formal class action or other complex litigation procedures.”). The agencies that have considered adopting aggregation procedures and ultimately declined to do so have cited potential costs as a major reason for their decisions. 5 See infra notes 40-45 and accompanying text. If the costs of group claims can be more easily controlled, agencies might be more willing to permit aggregation in the first place.

This Essay proceeds in three parts. Part I details the current shift toward aggregation in administrative proceedings. Part II analyzes potential costs. Part III explains how agencies could authorize group hearings without forfeiting control of agency priorities.

I. The Benefits

According to scholars, judges, and politicians, bureaucratic delay now amounts to a “crisis” in administrative law. 6 See Sant’Ambrogio & Zimmerman, supra note 4, at 1637 (“[T]he number of claims languishing in bureaucratic limbo has become a new crisis . . . .”); id. at 1637 n.5 (listing other scholars, judges, and politicians who have taken note of the crisis). In April 2015, for example, the Office for Medicare Hearings and Appeals (OMHA) had a backlog of more than 800,000 cases. 7 Nancy J. Griswold, U.S. Dep’t Health & Human Servs., Appellant Forum—Update from OMHA 6-8 (2015), https://perma.cc/XFK3-DH7T. The following year, wait times for claims ballooned to almost two years. 8 Id. at 6. The VA, meanwhile, has a backlog of more than 470,000 appeals. 9 See H.R. Rep. No. 115-135, at 5 (2017) (describing this as “a 20% increase in little more than 2 years”). The VA also has a backlog of more than 75,000 claims for benefits. Veterans Benefits Administration Reports—Claims Backlog, U.S. Dep’t of Veterans Affairs, https://perma.cc/Z4AT-LB7T (archived Dec. 27, 2017). These appeals take an average of three years to resolve at the regional-office level and an average of five years to resolve if the veteran chooses to appeal to the Board of Veterans’ Appeals. 10 H.R. Rep. No. 115-135, at 5. The VA has projected that by 2027, veterans will be waiting an average of ten years for final resolution of their appeals. 11 Statement of the Honorable Sloan Gibson Deputy Secretary Department of Veterans Affairs Before the Committee on Veterans’ Affairs U.S. Senate 3 (May 24, 2016), https://perma.cc/7FF8-W8C7.

The cure, it seems, is to aggregate claims to ensure that entitlement beneficiaries actually receive federal benefits. 12 See Student Assistance General Provisions, Federal Perkins Loan Program, Federal Family Education Loan Program, William D. Ford Federal Direct Loan Program, and Teacher Education Assistance for College and Higher Education Grant Program, 81 Fed. Reg. 75,926, 75,927 (Nov. 1, 2016) (to be codified at 34 C.F.R. § 685.222(f)) (adopting new regulations that would permit DOE to aggregate student claims for loan forgiveness); see also Monk v. Shulkin, 855 F.3d 1312, 1320 (Fed. Cir. 2017) (endorsing aggregation techniques for the VA); Kurtz v. Kimberly-Clark Corp., 315 F.R.D. 157, 159 (E.D.N.Y. 2016) (urging the FTC to consider aggregating claims). Complainants are also attempting to use this technique before the Federal Maritime Commission. See, e.g., Notice of Filing of Complaint and Assignment, Landers Bros. Auto Grp., Inc. v. Nippon Yusen Kabushiki Kaisha, Fed. Mar. Comm’n No. 16-11 (May 4, 2016), https://perma.cc/DYD5-VLQL; Notice of Filing of Complaint and Assignment, Cargo Agents, Inc. v. Nippon Yusen Kabushiki Kaisha, Fed. Mar. Comm’n No. 16-01 (Jan. 7, 2016), https://perma.cc/44LD-EL5U. According to the Administrative Conference of the United States (ACUS), agencies should “take advantage of the benefits of aggregation” to “avail themselves of the benefits of aggregate adjudication in the service of their statutory mandates.” 13 Sant’Ambrogio & Zimmerman, supra note 3, at 9-10.

Aggregation seems like a sensible solution because it allows judges to hear multiple claims in a single proceeding. Indeed, the agencies that pioneered administrative aggregation procedures have done so because aggregation would allow them to process claims more efficiently and thereby reduce delays. For example, OMHA justified new aggregation procedures as “offer[ing] a scientifically reliable, time-saving and lower cost approach to addressing large volumes of claim disputes.” 14 Office of Medicare Hearings and Appeals, Statistical Sampling Pilot Program: Fact Sheet (2017), https://perma.cc/BM7R-Q85F. The EEOC gave a similar justification when it promulgated a rule allowing ALJs to certify classes. 15 See Federal Sector Equal Employment Opportunity, 57 Fed. Reg. 12,634, 12,639 (Apr. 10, 1992).

The judges and policymakers who have pushed agencies to adopt aggregation procedures have also cited judicial economy and fairness as reasons to do so. Barton F. Stichman, the Joint Executive Director of the National Veterans Legal Services Program, recently advocated legislation to create a class-action procedure for veterans’ appeals “to conserve the limited resources of the VA and the courts, and to ensure that similarly situated veterans receive the VA benefits to which they are entitled.” 16 See Examining Appellate Processes and Their Impact on Veterans: Hearing Before the Subcomm. on Disability Assistance & Memorial Affairs of the H. Comm. on Veterans’ Affairs, 111th Cong. 15 (2009) (statement of Richard Paul Cohen, Executive Director, Nat’l Org. of Veterans Advocates) (“The Veterans Court should get class action status. That way a problem that affects a vast majority of veterans can be fixed all at once.”). The Court of Appeals for the Federal Circuit and a judge in the Eastern District of New York cited similar reasons in advocating for agency aggregation of claims. 17 See Monk, 855 F.3d at 1320; Kurtz, 315 F.R.D. at 159.

Thus, the general reasons used to justify aggregation in civil litigation—that it increases judicial efficiency, promotes legal access, and makes judicial decisions more consistent—apply part and parcel to administrative proceedings. Aggregation promotes legal access by providing a way to resolve claims for small damages that would be inefficient to bring individually. 18 See Sant’Ambrogio & Zimmerman, supra note 4, at 1649 (“Formal aggregate procedures enable litigation when damages are too small for individuals to justify the high costs of retaining counsel.”). It promotes efficiency because aggregation “avoid[s] the duplicative expenditure of time and money associated with traditional case-by-case adjudication, 19 Id. at 1650 (citing William B. Rubenstein Et Al., Newberg on Class Actions § 1.9 (5th ed. 2015)); see also Rubenstein et al., supra, § 1.9 (“Class actions are particularly efficient when . . . the courts are flooded with repetitive claims involving common issues.”). which may entail months or years of the ‘same witnesses, exhibits and issues from trial to trial.’” 20 Sant’Ambrogio & Zimmerman, supra note 4, at 1650 (quoting Jenkins v. Raymark Indus., Inc., 782 F.2d 468, 473 (5th Cir. 1986)); see also William W. Schwarzer, Settlement of Mass Tort Class Actions: Order Out of Chaos, 80 Cornell L. Rev. 837, 837-38 (1995) (explaining that class actions avoid duplicative litigation). And it promotes uniformity because aggregate proceedings “treat like parties in a like manner.” 21 Sant’Ambrogio & Zimmerman, supra note 4, at 1650; see also id. at 1650 n.60 (“Individual processing leaves open the possibility that one court, or jury, will resolve a factual issue for the plaintiff while the next resolves a seemingly similar issue for the defendant.” (quoting Rubenstein et al., supra note 19, § 1:10)). The alternative is to require judges and parties to engage in redundant fact-finding and legal analysis.

In fact, agencies have adopted a wide variety of aggregation techniques. 22 See generally id. at 1663 (providing some examples of informal and formal aggregation techniques). The EEOC, for example, has developed a class procedure that mimics Federal Rule of Civil Procedure 23. 23 Id. at 1642, 1659 (showing that the EEOC has used class proceedings analogous to Rule 23 proceedings to adjudicate over 700 claims). OMHA has begun using statistical sampling to quickly and effectively process complicated medical claims. 24 Id. at 1676-79. And the Vaccine Court has adopted “Omnibus Proceedings” to centralize similar cases involving the same scientific questions. 25 Id. at 1664. Thus, while most agencies do not frequently use aggregation techniques, agencies can—and have—used group proceedings creatively and in ways that suit their own particular objectives. 26 Id. at 1659.

II. The Costs

But those benefits do not tell the entire story. Aggregation could certainly reduce bureaucratic delay, but it could also hinder an agency’s ability to control its own policy priorities.

Administrative adjudications frequently involve claims for entitlements in which the ALJ determines the fiscal obligations of the very agency for which she works. Congress has been known to pass generous entitlement schemes without earmarking enough resources to honor all claims. 27 See, e.g., Steve Carlson et al., WIC Participation and Costs Are Stable: Have Returned to Pre-Recession Levels 3 (July 19, 2017), https://perma.cc/5DA5-E55W (“For many years after policymakers established WIC in 1974, Congress did not provide enough funding to serve all eligible low-income women and children who sought benefits.”). The effect is that agencies cannot afford to honor all potentially meritorious entitlements. For example, this summer the VA came alarmingly close to being unable to pay for all the health claims it had already agreed to provide. 28 See Leo Shane, VA Choice Funding Problem Looms for Congress, Military Times (July 5, 2017), https://perma.cc/6BQ8-Y4JC. Federal housing assistance is so underfunded that only a quarter of households who qualify for rental assistance actually receive it. 29 Editorial, Show HUD’s Budget Cuts the Door, N.Y. Times (May 30, 2017), https://perma.cc/TV4E-9ETY (“These destructive proposals come at a time when the federal housing effort is already so underfunded that three-quarters of the households that qualify for rental assistance based on income do not receive it.”). And, while nine million people qualify for the Department of Labor’s job training program for low-income, older Americans, budget shortfalls mean that the program can only serve 65,000 Americans each year. 30 Chris Farrell, Trump Should Fix, Not Scrap, Job-Training Program for Older Adults, Forbes (June 12, 2017), https://perma.cc/99X2-YW7Y.

The root issue thus seems to be chronic underfunding, not administrative waste. The VA currently has almost 115,000 pending disability claims. 31 Veterans Benefits Admin., Monday Morning Workload Report (Oct. 16, 2017), https://perma.cc/366J-FGR9. This number—and those in the following two notes—are calculated from information listed on the sheet entitled “Traditional Aggregate (TA).” More than 30% have languished for more than 125 days. 32 Id. It also has more than 275,000 appeals, which have been pending for an average of 420.8 days. 33 Id. The quick resolution of disability claims—each of which currently costs the VA an average of $14,855 per year 34 See U.S. Dep’t of Veterans Affairs, Compensation 5 (2017), https://perma.cc/8J8D-G2MV (providing data for fiscal year 2016). —would have made it significantly more difficult for the agency to honor the other entitlements it is obligated to pay. 35 Payments for disability compensation already account for roughly three-quarters of the VA’s total mandatory benefits budget. See Department of Veterans Affairs, Budget in Brief 2 (2018), https://perma.cc/KV5Z-TZN2.

Rental assistance programs provide another example of the dangers of unsupervised agency aggregation. The local agencies that administer federal rental assistance rely on bureaucratic delay to provide essential services. Funding shortfalls mean that only one-quarter of qualified applicants actually receive housing assistance. 36 Joint Ctr. for Hous. Studies of Harvard Univ., The State of the Nation’s Housing: 2016, at 5, https://perma.cc/726C-MFWZ (“Only one in four income-eligible renters receives assistance of any kind.”). To provide assistance to those most in need, agencies have coupled extreme delays—reportedly to the tune of almost a decade 37 New York City’s Affordable Housing Programs, Metro. Council on Hous., https://perma.cc/L4C3-ACDW (“The wait time for income-eligible households who are not in a priority category is said to average 9 years.”). —with a fast-track process for those most in need of assistance. HUD even provides guidance that explains how to develop waitlist procedures to ensure that those most in need receive fast-track service. 38 U.S. Dep’t of Housing & Urban Dev., HUD Handbook 4350.3: Occupancy Requirements of Subsidized Multifamily Housing Programs 4-11 to 4-12 (2013), https://perma.cc/MST5-L3Y4. Imagine if an ALJ aggregated rental assistance appeals. Absent a method of disaggregating the claims, the agency might be required to provide assistance it cannot afford, and it would not be able to help those most in need.

From this angle, it seems that the procedural hurdles to receiving the benefits of entitlement programs are not random. As in the case of housing assistance, they may actually ensure that those most in need are able to receive assistance more quickly than other qualifying applicants. 39 See id. (recommending that building owners fast track those applicants most in need of housing assistance so that they can provide entitlements to those most in need without being forced to process all pending claims). Preventing ALJs from certifying classes thus begins to look like a mechanism for conserving scarce resources.

Of course, it is unlikely that a given ALJ would certify a class for sinister reasons. But it is not hard to imagine an ALJ seeking to swiftly resolve claims the agency would prefer to keep on the backburner. Even that, however, could pose challenges to administrative efforts to prioritize certain activities. Another possibility is that an ALJ frustrated with the slow pace of administrative proceedings would substitute her own notions of good policy priorities to hasten the process. While well-meaning, such a decision could create financial obligations far beyond what the agency anticipated. Without a mechanism to control those costs, agencies might find their budgets stretched in unexpected ways, and they might lack the resources to provide entitlements to those most in need.

Indeed, all three of the agencies that have invited public comment on the use of aggregation in administrative proceedings have ended up citing the costs of aggregation as a reason not to grant ALJs unfettered license to aggregate claims. First, the Federal Communications Commission (FCC) considered and rejected a proposal by private attorneys to hear class actions in its own adjudications for alleged violations of the Communications Act. 40 In re Solvable Frustrations, Inc., 29 FCC Rcd. 4205 (2014). It observed that aggregation would “needlessly divert” resources from serving the public. 41 Id. The Commodity Futures Trading Commission (CFTC) also considered and rejected a class action mechanism in proceedings involving broker-dealer disputes. 42 See Rules Relating to Reparation Proceedings, 59 Fed. Reg. 9631 (Mar. 1, 1994). Like the FCC, it feared the costs of aggregation. 43 Id. (“The [CFTC] finds that . . . its resources would be used more effectively elsewhere . . . .”). Third, in November 2016, the U.S. Department of Education (DOE) adopted regulations providing for a claim-aggregation process modeled after class actions in civil adjudication. 44 Student Assistance General Provisions, Federal Perkins Loan Program, Federal Family Education Loan Program, William D. Ford Federal Direct Loan Program, and Teacher Education Assistance for College and Higher Education Grant Program, 81 Fed. Reg. 75,926, 76,084 (Nov. 1, 2016) (to be codified at 34 C.F.R. § 685.222(f)); see also Adam S. Zimmerman & Michael Sant’Ambrogio, Should the Education Department Hear Class Actions When Colleges Collapse?, Reg. Rev. (Oct. 10, 2016), https://perma.cc/D3PS-78ET (discussing the proposal of the “innovative new rule to streamline student claims for loan forgiveness” by offering a “‘group process’ modeled on class action rules in federal court”). The effective date of these regulations has been delayed until July 1, 2018. See 34 C.F.R. 685.222 (2017). The process would allow the DOE to hear large numbers of students’ claims for debt relief when those students attend schools that go bankrupt or commit fraud. Under the final rule, however, only designated DOE officials may commence the class proceeding, and there are no formal procedures to petition the agency to aggregate claims. 45 Student Assistance General Provisions, Federal Perkins Loan Program, Federal Family Education Loan Program, William D. Ford Federal Direct Loan Program, and Teacher Education Assistance for College and Higher Education Grant Program, 81 Fed. Reg. at 75,968 (“We disagree that a formal right of petition for entities such as State AGs, advocacy groups, or legal aid organizations should be included in the regulations.”); see also Luke Herrine, The Dark Side of Departmental Discretion, Reg. Rev. (Jan. 5, 2017), https://perma.cc/ZWW6-UVGA. The DOE thereby prevented ALJs from determining its priorities by not allowing them to unilaterally determine whether to aggregate claims.

It is noteworthy that the FCC, CFTC, and DOE considered aggregation in the context of administrative adjudications in which claims are brought solely against private parties. They did not consider the use of aggregation procedures in cases where the agency itself would disburse the funds. Those three agencies, therefore, felt the costs of managing group adjudication alone were too high to countenance. And such costs pale in comparison to those agencies who dole out benefits themselves might face. 46 See Federal Spending: Where Does the Money Go, Nat’l Priorities Project, https://perma.cc/Y9FZ-6XSH (archived Dec. 27, 2017) (explaining that Social Security comprises about 23% of the total federal budget, Medicare comprises another 15%, and veterans’ benefits comprise another 4%). It is therefore unsurprising that the only agencies to use aggregation procedures with any regularity only incur management expenses. 47 Sant’Ambrogio & Zimmerman, supra note 4, at 1660 n.97 (listing the agencies with “more than fifteen reported administrative decisions referencing the agency’s consolidation rule,” none of which pays out entitlements itself). And, while courts have promoted aggregation for agencies that would themselves have to pay for entitlements, 48 See Monk v. Shulkin, 855 F.3d 1312, 1320 (Fed. Cir. 2017) (recommending aggregation as a way for the VA to “promot[e] efficiency, consistency, and fairness, and improve[e] access to legal and expert assistance by parties with limited resources”). those agencies have fought tooth and nail to avoid developing aggregation procedures. 49 Id. at 1320 (documenting the reluctance of the Court of Appeals for Veterans’ Claims to provide aggregation techniques).

The problem of unfettered aggregation is all the more acute given the standard of review applied to most entitlement claims. Generally, the head of an agency is permitted to interpret law without regard to the decision below, but may only overturn the ALJ’s findings of fact for abuse of discretion or clear error. 50 Most agency findings of fact are given substantial deference. See, e.g., Hodges v. Sec’y of Dep’t of Health & Human Servs., 9 F.3d 958, 961 (Fed. Cir. 1993) (noting that the special master’s scientific and medical findings may only be set aside on appeal if found to be “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law”); Sugule v. Frazier, 639 F.3d 406, 411 (8th Cir. 2011) (“By contrast to legal questions, we review the agency’s findings of fact under a deferential substantial-evidence standard.”). Entitlement claims are often fact-intensive decisions. 51 See Joseph W. Mead & Nicholas A. Fromherz, Choosing a Court to Review the Executive, 67 Admin. L. Rev. 1, 48 (2015) (describing “fact-based adjudication[s] about an individual’s entitlement to benefits or other relief” as “fact-intensive cases”). The agency head may thus have little to no power to stop ALJs from advancing policies that the agency head would rather place on the backburner.

III. A Solution

Fortunately, there is an easy solution to the potential costs of administrative aggregation. In fact, the same authority that permits agencies to aggregate claims could also be used to develop procedures to disaggregate claims an agency regards as contrary to its mission.

When the Federal Circuit found that the Court of Appeals for Veterans’ Claims had authority to certify classes, it located the VA’s authority in its power to craft hearing procedures to help it carry out its statutory mission. 52 See Monk, 855 F.3d at 1319 (“[T]he Veterans Court’s jurisdiction extends to ‘compel action of the Secretary unlawfully withheld or unreasonably delayed.’ We see no principled reason why the Veterans Court cannot rely on the All Writs Act to aggregate claims in aid of that jurisdiction.” (citation omitted) (quoting 38 U.S.C. § 7261(a)(2) (2015))); see also 38 U.S.C. § 501(a) (2015) (“The Secretary [of Veterans Affairs] has authority to prescribe all rules and regulations which are necessary or appropriate to carry out the laws administered by the Department . . . including . . . the manner and form of adjudications and awards.”). That decision is consistent with numerous other cases finding that agencies have sweeping authority to implement rules in furtherance of their statutory mandates. 53 See Sant’Ambrogio & Zimmerman, supra note 4, at 1653-54. For example, the Supreme Court has reasoned that agencies “should be free to fashion their own rules of procedure and to pursue methods of inquiry capable of permitting them to discharge their multitudinous duties.” 54 FCC v. Pottsville Broad. Co., 309 U.S. 134, 143 (1940). Therefore, agencies can consolidate cases and decide “subordinate questions of procedure,” such as “the scope of the inquiry, whether [cases] should be heard contemporaneously or successively, whether parties should be allowed to intervene in one another’s proceedings, and similar questions.” 55 Id. at 138. Indeed, prohibiting aggregation would be at odds with the substantial flexibility the Supreme Court has granted agencies when choosing the best procedural format for decisions that affect large groups of people. 56 See, e.g., Vt. Yankee Nuclear Power Corp. v. Nat. Res. Def. Council, Inc., 435 U.S. 519, 524 (1978) (“[T]his Court has for more than four decades emphasized that the formulation of procedures [is] basically to be left within the discretion of the agencies to which Congress had confided the responsibility for substantive judgments.”).

If the power to aggregate flows out of the agency’s authority to craft procedural rules, so too does the power to limit the effects of aggregation. The DOE’s recent use of aggregation techniques indicates that agencies have a variety of options to control the costs of aggregation. When the DOE adopted an opt-out class action proceeding to resolve the common claims of thousands of student borrowers, it emphasized that only designated DOE officials could commence class proceedings. The agency hoped to avoid creating a “‘cottage industry’ of opportunistic attorneys and agents attempting to capitalize on victimized students . . . and unleashing a torrent of frivolous lawsuits.” 57 Student Assistance General Provisions, Federal Perkins Loan Program, Federal Family Education Loan Program, William D. Ford Federal Direct Loan Program, and Teacher Education Assistance for College and Higher Education Grant Program, 81 Fed. Reg. 75,926, 75,928 (Nov. 1, 2016); see also id. (“[B]y providing that only a designated Department official may present group borrower claims in the group processes . . . the Department believes that the potential for frivolous suits in the borrower defense process will be limited.”). Thus, the DOE created a mechanism that would allow it to aggregate claims on behalf of student borrowers but kept for itself exclusive authority to determine whether to commence aggregated proceedings. 58 Student Assistance General Provisions, Federal Perkins Loan Program, Federal Family Education Loan Program, William D. Ford Federal Direct Loan Program, and Teacher Education Assistance for College and Higher Education Grant Program, 81 Fed. Reg. 39,330, 39,347-48 (proposed June 16, 2016). The Notice of Proposed Rulemaking goes on to state that “the Secretary is best positioned to make a determination as to whether a group process is appropriate since the Secretary is likely to have the most information regarding the circumstances that warrant use of a group process.” Id. at 39,348. This initiation limitation reduces the potential for group proceedings to frustrate the DOE’s priorities by allowing DOE officials to opt out of offering a group process if the claim ends up running at cross purposes with the agency’s own priorities.

Despite their promise to more efficiently dispose of student borrower claims, the DOE’s procedures have been criticized. Adam Zimmermann and Michael Sant’Ambrogio praised the DOE rule but urged the agency to “permit individual student borrowers, their legal representatives, or state attorneys general to petition the Department to commence a group proceeding.” 59 Zimmerman & Sant’Ambrogio, supra note 44. Luke Herrine argued thast “[a] deep problem underlies the new group-based process: the Department, in its sole discretion, retains exclusive authority to initiate a group process. The new regulations lack any formal procedure for students, advocates, enforcement agencies, or other interested parties to argue for group treatment.” 60 Herrine, supra note 45.

Yet the DOE’s Notice of Proposed Rulemaking cites fiscal health as a reason to retain control over the aggregation process. 61 Student Assistance General Provisions, Federal Perkins Loan Program, Federal Family Education Loan Program, William D. Ford Federal Direct Loan Program, and Teacher Education Assistance for College and Higher Education Grant Program, 81 Fed. Reg. at 75,967 (citing fiscal health as one of the reasons that the agency should retain sole discretion to aggregate claims). Herrine claims that those provisions show that the DOE is “more focused on protecting its own bottom line than protecting the students it exists to serve.” 62 Herrine, supra note 45. But the DOE itself defended its decision as a way of ensuring that it can further its statutory mandate of serving these students as it sees fit.

There are, of course, other ways for agencies to control the costs of aggregation. One could simply grant the agency head the authority to disaggregate claims. In this way ALJs rather than high-level agency officials could determine whether to aggregate claims, but the Secretary would be able to undo an ALJ decision if she determined that the ALJ was out of line. Similarly, the agency head could delay the group hearing until a time when the agency is financially prepared. Claimants in bureaucratic limbo would be able to try their cases in a class proceeding, but only after a certain period of time has elapsed. That time delay would enable the agency to prospectively account for the costs of a large judgment, and the agency would not find itself constrained by spending decisions it might have made years ago.

While critics might complain that these proposed disaggregation tools would incentivize agencies to disaggregate all claims, this critique is unconvincing for a few reasons. First, agencies today rarely aggregate claims. Any move toward aggregation should therefore be favorably received by claimants whose claims may be more quickly resolved as a result. Disaggregation procedures would prompt such a move, as they would allow agencies to control the costs that are preventing them from allowing aggregation in the first place. Second, under the Administrative Procedure Act, all agency actions are subject to judicial review for abuse of discretion. 63 See 5 U.S.C. §§ 701-706 (2015) (providing for judicial review of agency actions). To withstand this review, an agency deciding to disaggregate would have to “examine the relevant data and articulate a satisfactory explanation.” 64 Motor Vehicle Mfrs. Ass’n of U.S, Inc. v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43 (1983). Third, agencies are overworked. The time it takes to examine ALJ aggregation decisions—especially for relatively inexpensive claims—would likely be seen as time better spent elsewhere. Finally, it is simply unrealistic to regard agencies as working against the interests of their beneficiaries. 65 Of course, there is a seemingly interminable debate about whether the government can be trusted to regulate effectively. Even those most skeptical of government interference, however, tend to view the government as inept—not as somehow acting to undermine the public interest. See, e.g., Joshua C. Macey, Note, Playing Nicely: How Judges Can Improve Dodd-Frank and Foster Interagency Collaboration, 126 Yale L. J. 806, 812-32 (2017) (criticizing the SEC and CFTC for failing to coordinate swap regulations). Note, however, that one of the bedrocks of modern administrative law—Chevron deference—takes for granted that agencies promote the public good. See, e.g., Stephen Breyer, Judicial Review of Questions of Law and Policy, 38 Admin. L. Rev. 363, 371 (1986) (arguing that deference to agency decisions “allows courts to allocate the law-interpreting function between court and agency in a way likely to work best within any particular statutory scheme”). Agencies may be understaffed and underfunded, but only a cynic would think that agencies are intent on actively preventing beneficiaries from receiving entitlements. A narrative that portrays agencies as ruthlessly set out to “protect their bottom line” ignores the fact that money not spent on entitlements simply goes back to taxpayers.

Conclusion

Proponents of administrative aggregation are right that aggregation is a powerful tool with tremendous promise. But if unchecked, it could cause more harm than good. Giving ALJs the unbridled power to aggregate claims would allow them to reroute agencies’ limited resources away from the mission-critical purposes the agencies have identified and toward paying massive group claims in one fell swoop. Such harm can be mitigated, however, by providing agencies a disaggregation safety valve to use when the costs of aggregation are insurmountable. Agencies empowered to manage the costs of administrative aggregation will at the same time be empowered to harness its benefits.

* Shannon M. Grammel received her J.D. from Stanford Law School in 2017. Joshua C. Macey received his J.D. from Yale Law School in 2017. Both authors wish to thank the editors of the Stanford Law Review, particularly Shannon Galvin, Dennis Martin, Fiona Theobald, Andrew Jones, and Adam Hersh.