A New Private Delegation Doctrine?

D.C. Circuit Strikes Down Delegation of Regulatory Authority to Amtrak

On July 2, 2013, in Ass’n of American Railroads v. DOT, the D.C. Circuit struck down a delegation of authority to Amtrak in § 207 of the Passenger Rail Investment and Improvement Act of 2008, holding that the statute unconstitutionally delegated regulatory power to a private party. This is a significant case for several reasons.

First, it’s potentially significant in terms of constitutional doctrine. In holding that private delegations of regulatory authority are illegitimate, the case seems to go against the conventional wisdom, which is that there is no special doctrine for private delegations by Congress: the Nondelegation Doctrine applies equally to public and private recipients of delegated congressional authority by Congress. Moreover, this conventional wisdom is probably right. The D.C. Circuit’s decision may yet be correct under the Due Process Clause, but the D.C. Circuit deliberately refused to choose whether this delegation implicated the Nondelegation Doctrine or the Due Process Clause.

Second, it’s potentially significant in terms of its real-world effect on delegations to private parties-though, again, much depends on precisely why the delegation is unconstitutional. If the decision rests on the Nondelegation Doctrine, it only affects federal delegations; but if it rests on the Due Process Clause, it also affects the much broader set of state delegations.

Third, in holding, based on a multi-factor analysis, that Amtrak is a private actor, it provides yet another example of how the public-private distinction is fuzzy, and an entity that is public for one reason might be private for another.

* * *

Amtrak, formally called the National Railroad Passenger Corporation, was created by statute in 1970. Faced with competition from other modes of transport, railroads that offered passenger service had been incurring heavy losses; many of these railroads had petitioned the Interstate Commerce Commission, which at the time regulated railroads, for permission to stop providing passenger service. With the passage of the statute, a railroad could transfer its passenger service responsibilities to Amtrak if it agreed to a number of conditions, one of which was to grant Amtrak the use of its tracks and other facilities. The statute provides that, except in an emergency, an Amtrak passenger car has precedence over another railroad’s freight car when they both need to use the same facilities. Most railroads agreed to these conditions, which were enshrined in a series of bilateral operating agreements.

Fast forward a few decades, to when Congress passed the Passenger Rail Investment and Improvement Act of 2008. One section of the new statute, § 207, required the Federal Railroad Administration (FRA) and Amtrak to “jointly . . . develop new or improve existing metrics and minimum standards for measuring the performance and service quality of intercity passenger train operations, including cost recovery, on-time performance and minutes of delay, ridership, on-board services, stations, facilities, equipment, and other services.” These performance measures aren’t of merely academic interest. Amtrak and its contractual partners are required to incorporate the measures into their operating agreements “[t]o the extent practical.” Perhaps more seriously, if “on-time performance” or “service quality” is substandard for two consecutive quarters, the Surface Transportation Board (STB), an independent agency housed in the Department of Transportation, is allowed to start an investigation (and is required to do so, if a complaint is filed) to check whose fault it is, and can assess damages against the host railroad if the problems are due to the railroad’s failure to grant preference to Amtrak trains.

These metrics and standards are supposed to be developed “jointly” by Amtrak and the FRA. If they can’t agree, they can petition the STB to appoint an arbitrator, whose decision will be binding. Amtrak thus has equal authority with the FRA on this issue; the FRA has to get Amtrak’s consent in developing the metrics and standards (or it has to abide by the decision of an arbitrator, who might also end up being private). The Association of American Railroads sued, charging that this sort of private delegation is invalid; and the D.C. Circuit agreed.

* * *

First, the D.C. Circuit noted, “[f]ederal lawmakers cannot delegate regulatory authority to a private entity.” The Nondelegation Doctrine states that when Congress delegates power to a government agency, all that’s required is that Congress provide an “intelligible principle” to limit the agency’s discretion. But, said the D.C. Circuit, “[e]ven an intelligible principle cannot rescue a statute empowering private parties to wield regulatory authority.” To illustrate this point, the D.C. Circuit cited Carter v. Carter Coal Co. (1936). The New Deal-era Congress had established a National Bituminous Coal Commission and required the organization of 23 “coal districts.” Within each coal district, all coal producers would be bound by any collective bargaining agreements agreed to by a majority of producers (representing two-thirds of total tonnage) and representatives of a majority of workers. In other words, 2/3 of the coal industry could, by its collective bargaining activity, legally bind the remaining 1/3 of the industry. (This reliance on massive binding industry self-regulation was classic New Deal procedure.) The Supreme Court struck this down: “The power conferred upon the majority is, in effect, the power to regulate the affairs of an unwilling minority. This is legislative delegation in its most obnoxious form; for it is not even delegation to an official or an official body, presumptively disinterested, but to private persons whose interests may be and often are adverse to the interests of others in the same business. . . . The delegation is . . . clearly arbitrary, and . . . clearly a denial of rights safeguarded by the due process clause of the Fifth Amendment . . . .”

To the D.C. Circuit, the question thus became: “precisely how much involvement may a private entity have in the administrative process before its advisory role trespasses into an unconstitutional delegation?” The court distinguished two relatively old cases where the Supreme Court had upheld private delegations: Currin v. Wallace (1939) and Sunshine Anthracite Coal Co. v. Adkins (1940). In Currin, an agency wrote regulations for tobacco auction markets, but they wouldn’t take effect unless two-thirds of the industry approved. The Supreme Court held that this was unobjectionable, because (unlike in Carter Coal) private parties did no more than hold the on-off switch for regulations that were written by the government. In Adkins, the fact pattern was reminiscent of Carter Coal: in fact, Congress had reenacted the same statute that the Supreme Court had struck down in Carter Coal, with the exception that now the government agency wrote the regulations for the industry based on the industry’s recommendation. The Supreme Court upheld this scheme, since now the industry was merely subordinate to the government agency. (The possible reality that the agency might just rubber-stamp the industry’s recommendations was irrelevant to the analysis: what was important was that the agency had the legal authority to modify the recommendations if it wanted to.)

The Amtrak case, though, went far beyond Currin or Adkins and was more similar to Carter Coal: Amtrak had an “effective veto” over FRA regulations and, in fact, enjoyed “authority equal to the FRA.” This really was a case where a private actor could control the regulations that governed the rest of the railroad industry, choosing a set of performance measures that would tend to make it look good relative to its competitors-and if the FRA refused to accede to Amtrak’s demands, the regulations would be written by an arbitrator chosen by the STB who could, for all we know, also be a private party.

* * *

Not so fast, though; one might legitimately argue that Amtrak isn’t private. The hazard with all such doctrines that draw a bright line between public and private is that, in reality, the line is somewhat fuzzy, especially in an age where contracting out of government services and pervasive regulation of the private sector are widespread. (See these two recent posts for examples of how antitrust, labor law, and other doctrinal areas deal with the distinction.)

So let’s tally up the indicia of privateness and publicness. On the public side, Amtrak’s Board of Directors has nine members, one of whom is the Secretary of Transportation and seven of whom are presidential appointees; the ninth, the President of Amtrak, is elected by the other eight. Amtrak has some private shareholders, but almost all its stock is preferred stock held by the federal government. The D.C. Circuit noted that Amtrak gets substantial subsidies from the federal government-though the amount of government money one gets generally isn’t relevant to whether one is public or private.

Notably, still on the public side, there’s a Supreme Court case, Lebron v. National Railroad Passenger Corp. (1995). In Lebron, Amtrak was sued, mostly on First Amendment grounds, for refusing to display a political advertisement in New York’s Penn Station. First Amendment rights, like many other constitutional rights, only apply against “state actors,” so the question was whether Amtrak was a state actor. The Supreme Court held that “where, as here, the Government creates a corporation by special law, for the furtherance of governmental objectives, and retains for itself permanent authority to appoint a majority of the directors of that corporation, the corporation is part of the Government for purposes of the First Amendment.”

On the private side, the 1970 statute specifies that Amtrak “is not a department, agency, or instrumentality of the United States Government.” The statute also commands that Amtrak “shall be operated and managed as a for-profit corporation.” Relatedly, by statute, “Amtrak is encouraged to make agreements with the private sector and undertake initiatives that are consistent with good business judgment and designed to maximize its revenues and minimize Government subsidies.” Amtrak itself announces that it’s “not a government agency or establishment [but] a private corporation operated for profit.” The D.C. Circuit attaches some significance (“somewhat tellingly”) to the fact that Amtrak’s URL is this doesn’t really seem all that telling, as one could make a similar claim about the U.S. Postal Service at

To decide the issue, the D.C. Circuit looked to “what functional purposes the public-private distinction serves when it comes to delegating regulatory power.” One purpose is accountability: a private delegation dilutes democratic accountability, because when power is delegated to a private organization, the government is no longer blamed for that organization’s decisions. (Perhaps; but if something goes wrong, why can’t the voters blame the government for the initial decision to delegate?) Another purpose is the distinction between the public good and private gain: public recipients of delegated power are “presumptively disinterested” and are bound by “official duty,” whereas private recipients may act “for selfish reasons or arbitrarily.” (Perhaps; but doesn’t this display an overly optimistic view of the motivations of public employees?) In the D.C. Circuit’s view, these considerations cut in favor of treating Amtrak as private: the statutory command that it be “managed as a for-profit corporation” requires that it seek its private good, not the public good, and Congress’s and Amtrak’s consistent labeling of Amtrak as private tends to distance Amtrak’s decisions from democratic accountability. (The court distinguished Lebron on the grounds that being a state actor for First Amendment purposes doesn’t mean one is a state actor for all purposes.) Section 207 thus delegates regulatory power to a private party, and is thus invalid.

* * *

One interesting aspect of Ass’n of American Railroads v. DOT is the idea that Amtrak can be private for the purposes of structural provisions like the Nondelegation Doctrine, even though it is public for the purposes of individual-rights provisions like the First Amendment. This is possible, though the D.C. Circuit’s multi-factor analysis isn’t exactly overwhelming.

A more interesting aspect of the case is what the court relegates to a footnote and refuses to decide. Recall the Carter Coal precedent from 1936, where giving businesses the power to regulate their competitors was characterized as “legislative delegation in its most obnoxious form.” Carter Coal is, unfortunately, less than crystal-clear on the precise source of the unconstitutionality. Since it mentions “legislative delegation,” one could think of it (as the D.C. Circuit did) as a Nondelegation Doctrine decision. Nondelegation challenges rest on the Vesting Clause of Article I of the federal constitution, which vests all legislative power in Congress (“All legislative powers herein granted shall be vested in a Congress of the United States . . . .”). Congress must exercise its own legislative power, but it’s allowed to delegate limited authority as long as the delegation is accompanied by some “intelligible principle” to limit the agency’s discretion.

But Carter Coal also says that the delegation is “clearly a denial of rights safeguarded by the due process clause of the Fifth Amendment” (“[N]or shall any person . . . be deprived of life, liberty, or property, without due process of law . . . .”).

So is this a Nondelegation Doctrine decision or a Due Process Clause decision? The D.C. Circuit wasn’t that interested in precisely what part of the Constitution was being violated. It wrote, in a footnote, that “the distinction evokes scholarly interest,” but the parties in this case didn’t press the point, and “neither court nor scholar has suggested a change in the label would effect a change in the inquiry.”

But reading Carter Coal as a Due Process opinion, and likewise grounding the Amtrak challenge in the Due Process Clause, makes more sense. The focus of a nondelegation challenge should be on how much power Congress has given up-has it actually given up legislative power, or has it merely allowed someone to fill gaps and ambiguities?-not on the identity of the recipient of the delegation. The Due Process Clause, on the other hand, is concerned with fair treatment, and the idea that financially biased decisionmakers (whether public or private) are illegitimate has been a staple of Due Process doctrine for a long time. Claims of bias, whether it’s a public official who has prejudged an issue or a private organization that can lose money depending on how it wields its power, thus fit more naturally into a Due Process framework than into a nondelegation framework.

Moreover, the distinction matters for future cases. A nondelegation holding based on Article I’s Vesting Clause would enforce the federal separation of powers by preventing Congress from getting rid of some of its legislative authority, and so it would only govern federal delegations. These separation of powers constraints are irrelevant for the states. The federal constitution doesn’t require that states have the same separation of powers as the federal government: states could adopt parliamentary democracy or engage in any number of structural experiments forbidden to the federal government, provided they comply with certain minimal guarantees like “one-person, one-vote” or having a “republican form of government.”

A Due Process challenge would be quite different: like the First Amendment and most other Bill of Rights provisions, the Due Process Clause now applies against the states, so a Due Process holding would also constrain state delegations and would therefore have a much wider sweep. Under the Due Process Clause, the court wouldn’t do the public/private inquiry that was on display here; rather, it would look to whether Amtrak was a state actor using the substantial body of state action doctrine. But this question was already resolved in 1995 by the Lebron case: yes, Amtrak is a state actor. (Lebron arose in a First Amendment context, but it turns out that the Due Process Clause, as well as various other individual-rights provisions, turns on the same state action question, so any finding of state action for First Amendment purposes carries over directly to the Due Process Clause.) Finding that Amtrak is a state actor doesn’t mean there’s a Due Proces violation; it’s only a threshold step that means that Due Process protections apply. To find out whether the Due Process Clause was violated, the relevant inquiry would be the extent of Amtrak’s financial bias. And given the statutory command that Amtrak act to maximize its profits, one could legitimately conclude that it couldn’t, without an unconstitutional conflict of interest, regulate the rest of the railroad industry-thus arriving at the same bottom line as the D.C. Circuit.

The D.C. Circuit thus seems incorrect when it says that public delegations of regulatory authority are merely evaluated by the “intelligible framework” test while private delegations are per se illegitimate. Rather, all federal delegations (public or private) are evaluated by the “intelligible principle” test, while all delegations of any kind (state or federal, public or private) are scrutinized for conflicts of interest under the Due Process Clause, and perhaps private delegations might be more vulnerable because conflicts of interest are more likely to arise there.

This, then, is the open issue at the heart of the D.C. Circuit’s opinion. The true reach of this decision is yet to be determined.

Alexander “Sasha” Volokh is an associate professor of law at Emory Law School. An archive of his previous articles is available here.