Cass R. Sunstein (Harvard University - Harvard Law School) has posted Is OSHA Unconstitutional? (Virginia Law Review, Forthcoming) on SSRN. Here is the abstract:
Under the Occupational Safety and Health Act, the Secretary of Labor is authorized to issue whatever standards are reasonably necessary or appropriate to provide safe or healthful places of employment. More than any other provision in federal regulatory law, this language is subject to a plausible nondelegation challenge, because it seems to ask the Secretary to choose among a wide array of intelligible principles for standard-setting. The constitutional challenge raises serious and unresolved questions for both regulatory policy and administrative law. In answering those questions, courts have three principal alternatives. The most aggressive approach would be to invalidate the statute in the hopes of encouraging, for the first time, sustained legislative deliberation about the proper content of occupational safety and health policy. The most modest approach, rooted in the Avoidance Canon, would be to construe the statutory language to produce floors and ceilings on agency action; that approach would require the Secretary to ban significant risks while forbidding the Secretary from regulating trivial or de minimis risks and also requiring the Secretary to show that any regulations are feasible. The third and preferable approach, also rooted in the Avoidance Canon, would be to construe the statute so as to require the agency to engage in a form of cost-benefit balancing. Such a construction would have the advantage of promoting greater transparency and accountability at the agency level. At the same time, it would raise difficult questions about the precise nature of such balancing in the context of occupational safety policy and also about legal constraints on agency assessment of both costs and benefits. Because of the distinctive nature of workplace safety, the best approach would give the agency considerable flexibility on questions of valuation while also permitting serious attention to distributional factors.
And from the paper:
There is a final puzzle, and it raises a large issue with respect to the relationship between courts and the administrative state. That issue involves the status of narrowing constructions, designed to avoid nondelegation challenges by construing agency discretion to be narrow rather than broad.147 Such narrowing constructions are not uncommon,148 and I have suggested that the cost-benefit approach is best justified as an example. But the whole approach raises a serious question. The problem, in sum, is that if, as American Trucking teaches, agencies are not permitted to rescue open-ended delegations through subsidiary policymaking in the guise of interpretation, courts should not be permitted to do so either.149 The question therefore arises: What is the status of the Avoidance Canon, in the specific context of a nondelegation challenge, in the aftermath of American Trucking?
At first glance, nothing in American Trucking should endanger the use of the Avoidance Canon. The Court’s suggestion was merely that if a statute does confer open- ended authority on an agency, the agency cannot eliminate that problem by deciding how much discretion to exercise. “The idea that an agency can cure an unconstitutionally standardless delegation of power by declining to exercise some of that power seems to us internally contradictory. The very choice of which portion of the power to exercise–that is to say, the prescription of the standard that Congress had omitted–would itself be an exercise of the forbidden legislative authority.”150 The new nondelegation doctrine, repudiated by American Trucking, asked agencies to develop “subsidiary policy” by which to discipline their discretion under open-ended statutes. The reason is that the development of subsidiary policy counts as an exercise of discretion. It is not “interpretation.”
If American Trucking is understood in this way, it certainly suggests that courts cannot rescue a statute from a nondelegation challenge if they are themselves making subsidiary policy. But when a court is legitimately selecting an interpretation that narrows agency discretion, it is not really making subsidiary policy, certainly not in the sense that the court of appeals deemed sufficient, and the Supreme Court irrelevant, in American Trucking. Instead a court that properly uses the Avoidance Canon is relying on standard legal materials to hold that of two or more plausible interpretations of a text, the agency is bound by the one that gives it limited rather than open-ended authority. An approach of this kind would not be legitimate if the standard legal materials left both court and agency at sea – if the narrower interpretation does not qualify as an interpretation at all. But if courts can fairly insist on that interpretation, as a reasonable way of coming to terms with what Congress has actually said, American Trucking creates no obstacle.
Here, then, is a possible problem with the approaches I have outlined. Suppose that the relevant interpretation is really an exercise in policymaking – that courts are choosing an intelligible principle not on the basis of anything that Congress has done, but as a means of implementing what judges see as the best way to come to terms with a difficult problem. Under American Trucking, the “floors and ceilings” approach, and the cost-benefit approach, would not cure the nondelegation problem if they amounted to judicial policymaking. But I have argued that on the basis of the standard legal materials, both approaches are legitimate readings of the legal materials in light of the Avoidance Canon. If this argument is correct, then judicial insistence on one or another should not be taken as running afoul of the Court’s rejection of the new nondelegation doctrine.
A deeply interesting treatment of an important question--very highly recommended.
I am not fully convinced by Sunstein's argument at the end of the paper. Forcing "cost-benefit analysis" as a narrowing construction requires courts to make the crucial policy choice (between cost-benefit analysis on the one hand, and approaches that would privilege worker safety, on the other hand) in a way that is permitted (but not required) by the statutory text. That is, the Avoidance Canon may require an "intelligible principle" but judicially-imposed cost-benefit analysis is only one of several possible "intelligible principles" that courts might impose. Does it make sense to have courts (as opposed to the agency) make the choice as to which of the available "intelligible principles" should be chosen? One way to think about this issue is via the following hypothetical--suppose that the Court were to mandate cost-benefit analysis and the agency were to reject that choice and adopt an alternative principle. (Chevron immediately comes to mind.)
But I am really just thinking out loud here. Read Sunstein!