Railroads around the time of the mid-twentieth century found themselves in a quandary. Passenger rail service had become unprofitable, yet common carriers were legally obliged to provide it.1 With the Rail Passenger Service Act of 1970,2 Congress struck a compromise. The National Railroad Passenger Corporation, also known as Amtrak, would assume the railroads’ common-carrier obligations; in exchange, Amtrak would use the railroads’ tracks and facilities3 and gain preference over freight traffic.4 In 2008, Congress passed the Passenger Rail Investment and Improvement Act5 (PRIIA), § 207(a) of which directs Amtrak and the Federal Railroad Administration (FRA) jointly to develop metrics and standards with which to evaluate the performance of intercity passenger trains.6 If Amtrak and the FRA cannot agree on metrics and standards, § 207(d) allows the Surface Transportation Board7 (STB) to appoint an arbitrator to resolve the dispute.8 Last Term, in Department of Transportation v. Association of American Railroads9 (Amtrak), the Supreme Court held that Amtrak is a governmental, not a private, entity for purposes of determining the validity of the metrics and standards created under § 207. The Court’s answer to the narrow question presented was unremarkable, but Amtrak illustrates courts’ unsettled approach to analyzing the constitutionality of nontraditional agencies at the public–private border.
In May 2010, Amtrak and the FRA jointly issued a set of metrics and standards.10 The Association of American Railroads (AAR), a trade group whose members would have to modify their operating agreements with Amtrak in order to incorporate the metrics and standards,11 promptly sued the Department of Transportation and the FRA, asking the District Court for the District of Columbia to declare § 207 unconstitutional and to invalidate the metrics and standards.12 AAR claimed two constitutional deficiencies: first, that § 207 violated the nondelegation doctrine by placing legislative power “in the hands of a private entity [Amtrak]”; second, that it violated the Fifth Amendment Due Process Clause by “[v]esting the coercive power of the government” in an “interested private part[y].”13 The district court rejected both constitutional challenges and granted the government’s motion for summary judgment.14 AAR appealed.
The Court of Appeals for the District of Columbia Circuit reversed the nondelegation ruling and declined to decide the due process claim.15 Writing for a unanimous panel,16 Judge Brown set out a simple syllogism. First, she maintained, Congress cannot delegate legislative power to private entities.17 Under the conventional version of the nondelegation doctrine, Congress can assign powers that resemble lawmaking to an agency of the Executive Branch only if it supplies an “intelligible principle” to guide the agency’s discretion.18 Under a precept Judge Brown termed the “lesser-known cousin” of that doctrine, even an intelligible principle cannot save a scheme that assigns such powers to private bodies.19 She acknowledged that private parties may serve an “advisory role” in the lawmaking process, but argued that the PRIIA contemplates more by granting Amtrak “authority equal to the FRA,” along with effective veto power, in issuing metrics and standards.20 Second, based primarily on Congress’s denomination of Amtrak as a “for-profit corporation,”21 Judge Brown found Amtrak to be a private entity.22 Thus, she concluded, § 207 effects an unconstitutional delegation of power to Amtrak.
The Supreme Court reversed and remanded. In a crisp opinion for the Court, Justice Kennedy23 explained that the decision below rested on a flawed premise: that Amtrak is a private entity for purposes of the constitutional questions presented.24 Justice Kennedy began with Amtrak’s distinctive ownership and corporate structure. On Amtrak’s board of directors sit the Secretary of Transportation (who also holds most of Amtrak’s stock) and seven other members appointed by the President, confirmed by the Senate, and removable by the President without cause. The ninth board member, who also serves as Amtrak’s president, is chosen by the other eight.25 The political branches “exercise substantial, statutorily mandated supervision over Amtrak’s priorities and operations”: Amtrak must submit detailed annual operating reports and attend frequent congressional oversight hearings; it is subject to the Freedom of Information Act and is a “designated federal entity” under the Inspector General Act of 1978; and, “rather than advancing its own private economic interests, Amtrak is required to pursue numerous, additional goals defined by statute.”26 Finally, Amtrak depends heavily on federal subsidies.27 In sum: “Amtrak was created by the Government, is controlled by the Government, and operates for the Government’s benefit.”28
Amtrak thus acted as a government entity in its joint issuance of metrics and standards under the PRIIA. Justice Kennedy brushed aside the seemingly contrary statutory declarations that Amtrak “shall be operated and managed as a for profit corporation” and “is not a department, agency, or instrumentality of the United States Government.”29 As the Court had determined previously in Lebron v. National Railroad Passenger Corp.,30 such pronouncements are not dispositive; “for purposes of Amtrak’s status as a federal actor or instrumentality under the Constitution, the practical reality of federal control and supervision prevails over Congress’ disclaimer of Amtrak’s governmental status.”31 Having supplied the lower court with the proper premise — Amtrak is a governmental entity — the Supreme Court remanded the case for a fresh round of consideration.32 The D.C. Circuit would need to consider AAR’s remaining Appointments Clause, nondelegation, and Due Process Clause challenges in light of Amtrak’s governmental status.33
Justice Alito joined the Court’s opinion in full but wrote separately in concurrence to consider more thoroughly the issues on remand. He organized his analysis around a theme: “Liberty requires accountability.”34 Several components of Amtrak’s statutory scheme, he noted, seemingly shield it from constitutionally required accountability mechanisms. First, Justice Alito worried that the arbitration provision of § 207 — authorizing an STB-appointed arbitrator to issue a binding decision on metrics and standards in the event of disagreement between Amtrak and the FRA — might violate the nondelegation doctrine by vesting a private entity with rulemaking power.35 Second, Justice Alito was vexed by the composition of Amtrak’s board. Amtrak’s president is arguably a “principal officer” under the Appointments Clause, yet, unlike the other eight board members, is not a presidential appointee.36 In conclusion, Justice Alito reiterated that although Amtrak is a governmental entity, “it does not by any means necessarily follow that the present structure of Amtrak is consistent with the Constitution.”37
Justice Thomas also wrote separately, concurring only in the judgment and proposing a different framework to “guide [the Court’s] resolution of delegation challenges.”38 He would jettison the “intelligible principle test,” which inadequately protects “the Constitution’s allocation of legislative power.”39 Article I vests legislative power exclusively in Congress. Historically, he explained, the “formulation of generally applicable rules of private conduct” was understood to be an exercise of legislative power.40 If a body other than Congress purports to formulate such rules, it intrudes on Congress’s prerogative — never mind any intelligible principle. Because the metrics and standards issued under § 207 affect the legal duties of private railroads, Justice Thomas continued, they qualify as generally applicable rules of private conduct, and Amtrak’s role in their issuance was an exercise of legislative power. Thus, he concluded, § 207 violates Article I by vesting legislative power in a body other than Congress.41
As decided, Amtrak was a straightforward case. The question presented was narrow — whether to invalidate the metrics and standards on the ground that Amtrak, as a private entity, had exercised unconstitutionally delegated power in their issuance42 — and the Court’s answer followed naturally from Lebron’s earlier holding in combination with the principle that the “structural” and “individual rights” provisions of the Constitution are complementary.43 In these circumstances, there was no need to go beyond invalidating the lower court’s minor premise to address its major premise respecting private nondelegation. Viewed more broadly, Amtrak is an illustration of the judiciary’s unsettled approach to analyzing the constitutionality of “boundary agencies” that sit at the public–private border.44 In navigating between “the Scylla of unlawful delegation on one hand, and the Charybdis of the Appointment Clause, on the other hand,”45 courts have sometimes jumbled separation of powers doctrine with a general unease over the often-fuzzy public–private boundary. In so doing, they have flirted with enforcing a special version of the nondelegation doctrine against private entities, an unwieldy approach that seems to adopt inconsistent premises respecting public and private delegations. Although the Court dodged it pro tem, private nondelegation will resurface on remand, as the AAR contends that § 207’s arbitration provision unconstitutionally delegates rulemaking power to a private party.46 The next round of Amtrak presents a chance to eliminate a troublesome doctrinal gadfly.
A statute that invites private participation in rulemaking presents different concerns from one that confers broad rulemaking authority on an executive agency: the former may countenance self-dealing, while the latter may improperly fuse separated functions. Grants of power to purely private or public–private hybrid entities ostensibly “undermine[] public accountability,”47 “expand[] government power through the use of shadow proxies,”48 and present “abuse of power concerns.”49 In particular, they raise the specters of “arbitrariness, lack of due process, and self-dealing.”50 Separation of powers doctrine, meanwhile, enforces formal partitions between the legislative, executive, and judicial functions of government in order to prevent, for example, the concentration of the lawmaking and law-execution functions in the same hands.51 Yet disputes about the proper relationship between the public and private spheres are frequently cast in separation of powers terms.52 The D.C. Circuit’s “private nondelegation” principle is in this vein.
Private nondelegation, “a second, more muscular version of the nondelegation doctrine,”53 supposedly finds its roots in a passing statement in Carter v. Carter Coal,54 where the Supreme Court invalidated a statute that allowed a portion of the coal industry to set wage and hour standards for the rest of the industry.55 Others have argued that Carter Coal rested on a due process rather than a nondelegation rationale;56 at any rate, the Supreme Court itself has never unambiguously espoused a private nondelegation subdoctrine,57 and it has since upheld statutes giving private parties an effective veto over proposed regulations58 or a role in drafting their own regulations.59 More to the point, the Court devised the “regular” nondelegation doctrine as a way to enforce the formal separation of functions between the legislative and executive branches — to ensure that Congress constrain agencies’ discretion with an “intelligible principle” when giving them expansive authority.60 If the purpose of the nondelegation doctrine is to prevent Congress from utterly relinquishing its legislative function to the executive branch, and if Amtrak is a private entity, then the inquiry is not one of separation of powers, but of something else, perhaps due process; if Amtrak is a government entity, the statute need only pass the intelligible principle test.61 Regardless, abstract separation of powers doctrine, and specifically the nondelegation principle, is not the appropriate lens through which to analyze grants of power to entities straddling the public–private border.
Perhaps that claim is too strong; possibly the D.C. Circuit’s opinion and Justice Alito’s concurrence hint at a way to unite the separation of powers and public–private inquiries. Yet their approach seems to adopt inconsistent premises respecting public and private delegations. The Court famously “does not enforce the nondelegation doctrine with . . . vigilance” because “the other branches of Government have vested powers of their own,” such as executive power, “that can be used in ways that resemble lawmaking.”62 But, the concurrence and decision below suggest, the Court can vigilantly enforce nondelegation against private entities, in which the Constitution vests no power that resembles lawmaking.63 Put another way, the Constitution has an “implicit design principle limiting the federal government to three branches,”64 so even if it is hard to spot an unconstitutional delegation to an executive agency because of the blurring of legislative into executive power, it is easy to spot one to a private entity outside the three branches.
The trouble with this expedient is that it simultaneously rejects and accepts the notion that the Vesting Clause powers can be defined in an essentialist way and policed on that basis.65 On the one hand, it yields to the orthodoxy that courts cannot usually distinguish cession of legislative power from authorization of broad executive discretion — an orthodoxy that implicitly disputes Justice Thomas’s contention that legislative power is usually or always identifiable by its nature.66 On the other hand, it claims to be able to distinguish an unconstitutional grant of legislative power to a private entity from a statute that merely solicits private-party input or conditions enforcement on private-party approval. Perhaps there is some way to reconcile these premises at the periphery — perhaps it is possible to define at least the core essence of legislative power and easier to tell when Congress has given it away to a private party because the lack of a constitutional private-party vesting clause makes the zone of ambiguity smaller. But in the broad run of cases, it is probably equally difficult to determine whether Congress has ceded legislative power or merely specified the terms of law implementation. To be consistent, courts should either enforce the nondelegation doctrine against all power recipients, whether public, private, or hybrid, or else against none of them.
Section 207’s arbitration provision well illustrates the difficulties with satisfactorily invoking a distinct private nondelegation doctrine. First is the problem just mentioned: the difficulty of characterizing the kind of power granted. Justice Alito found that the issuance of metrics and standards is “obviously regulatory,” making the arbitrator’s involvement a question of legislative delegation.67 Yet the arbitrator’s role is fairly removed from the actual issuance of rules: the arbitrator only resolves disputes over standards proposed by government entities for measuring performance under a preexisting statutory scheme that imposes obligations on the railroad industry. Surely it is fair to characterize the issuance of measurement standards, and particularly the arbitrator’s role in their issuance, as merely implementing or effectuating preexisting rules rather than as de facto “legislating.”68
Second is a problem of degree. If the D.C. Circuit is right that “[e]ven an intelligible principle cannot rescue” a delegation of governmental power to a private entity,69 then courts must invalidate all such delegations. But as that court acknowledged, private entities may “help a government agency make its regulatory decisions,” so the question becomes, “precisely how much involvement may a private entity have in the administrative process before its advisory role trespasses into an unconstitutional delegation?”70 There is not much guidance on how to discern that line. The FRA and Amtrak in fact promulgated new standards without resorting to arbitration, so the case for invalidating them depends on the argument that (1) the arbitrator’s role would have crossed the line from advisory to unconstitutional, and (2) any standards developed in the shadow of the mere possibility of private arbitration are tainted.71 On the one hand, the role of the private entity under this scheme seems inconsequential, not obviously any more significant than that of the regulated industries in schemes the Court has approved.72 On the other, as Justice Alito argued, the arbitrator influences what appears in the Federal Register.73 If the intelligible principle standard for public delegations is vague and imprecise, the proposed linedrawing standard for private delegations is at least equally so.74
Congress frequently creates entities that sit at the public–private border and assigns them tasks that may resemble lawmaking, execution, or adjudication.75 To structure such organizations to withstand constitutional challenge, Congress needs clear guidelines ex ante. A judicial approach that mixes abstract separation of powers doctrine with concerns about private influence in the public sphere does not provide clear guidance. Instead, by scrambling distinct issues — the separation of governmental functions with the prevention of private self-dealing — the private nondelegation doctrine engenders confusion in both its rationale and its application. A better approach would be to identify what makes private participation in public functions objectionable and then proceed by matching the doctrine to the concerns. If arbitrariness and self-dealing are the main fears, courts might analyze power grants to private or hybrid organizations in terms of due process.76 If transparency and accountability are wanting, courts might also consider whether the Appointments Clause should apply more broadly.77 It remains to be seen what the D.C. Circuit will do with the case on remand, but § 207(d) — the arbitration provision — presents an opportunity to disentangle two distinct areas of doctrine.