In recent weeks, lawyers, journalists, and concerned citizens have learned more than they ever wanted to about “executive orders.” They’ve come thick and fast: a directive to roll back health reform, the first and second travel bans, and a missive on climate change. Questions about legal limits on executive orders, and the role of the courts in evaluating them, have a newfound and well-deserved prominence.

But most executive orders do not change much on their own; instead, they command or encourage agencies to make changes. And in a provocative post over at Notice & Comment, Adam White argues that the D.C. Circuit’s 2012 decision in Sherley v. Sebelius could create difficulties for parties who challenge agency actions taken pursuant to President Trump’s executive orders.

The post makes some good points, but I think Sherley is so badly reasoned that its holding ought to count for little in future cases. It’s worth explaining why.

Shortly after President Obama was elected, he announced in an executive order that the federal government would expand its support for embryonic stem-cell research. When the National Institutes of Health (NIH) adopted new guidelines to that effect, they refused to consider objections from commenters opposed to such research. The commenters sued, leading to Sherley, where the D.C. Circuit sided with the NIH:

NIH may not simply disregard an Executive Order. To the contrary, as an agency under the direction of the executive branch, it must implement the President’s policy directives to the extent permitted by law . . . Bound as it is to carry out the President’s directives, NIH thus reasonably limited the scope of its Guidelines to implement the Executive Order. And because the Executive Order’s entire thrust was aimed at expanding support of stem-cell research, it was not arbitrary or capricious for NIH to disregard comments that instead called for termination of all [embryonic stem cell] research (including research that the executive branch has permitted since 2001).

Read for all it’s worth, Sherley will allow agencies to point to President Trump’s executive orders as excuses for refusing to consider objections to their policy choices. The effect could be to insulate those choices from a significant form of judicial scrutiny.

But Sherley’s reasoning is screwy. To see why, consider a slight variant of the facts in State Farm, one of the seminal cases about judicial review of agency action.Imagine that President Reagan in 1981 had issued an executive order instructing the National Highway Traffic Safety Administration (NHTSA) to withdraw its rule requiring the installation of airbags in new motor vehicles. During the withdrawal proceedings, comments were submitted objecting to the irrationality of that policy choice, but NHTSA blew them off.

Under Sherley, that’s perfectly OK: because the agency was duty-bound by President Reagan’s executive order to withdraw the airbag rule, it could disregard comments seeking the opposite result. Sherley thus suggests that the outcome in State Farm would have been different if only President Reagan had used an executive order to direct NHTSA to withdraw its airbag rule.

That can’t possibly be right. Whatever the president says, agencies have an independent duty under the Administrative Procedure Act to offer reasons and respond to comments. That duty doesn’t dissolve when the president has made a particular policy choice. That’s the whole point of the Supreme Court’s canonical holding in State Farm: “the president made me do it” is not adequate to justify an agency action. Agencies must stand or fall on their own reasoning.

For much the same reason, it’s imprecise for Sherley to say that NIH was “bound” to carry out the president’s instructions. Congress vested in Secretary of Health and Human Services Kathleen Sebelius the authority to distribute research funding. It did not vest that authority in the president. If President Obama ordered Sebelius to do something arbitrary or irrational, she could’ve refused to carry out the president’s instruction. She would’ve been sacked, but she still had room to exercise her independent discretion.

The Sherley court’s contrary reasoning suggests that a unilateral executive has directive authority over his subordinates. But whether the president has such directive authority is a live debate in administrative law; the mainstream view is that he does not. Sherley seems unaware that such a debate exists, much less that it’s taking sides in it. A drive-by holding like that doesn’t have much value as precedent.

If Sherley doesn’t make sense on its face, is a narrower reading available? Or, if not a narrower reading, is there a more compelling justification for the result?

Possibly. The Sherley court seemed to think it significant that the commenters’ blanket objections to stem-cell research also applied to research that even the Bush administration had supported. Because NIH wasn’t revisiting that Bush-era decision, maybe it didn't have to respond to comments that sought to reopen the broader debate. (This point goes to the scope of the rulemaking, but it could be reframed as a claim about harmlessness.)

Admittedly, this reformulation of Sherley isn’t compelling: moral objections seem germane to a sweeping expansion of federal support for stem-cell research. A better explanation for the result might be that the agency had no obligation to respond to comments to begin with. The new guidelines probably weren’t legislative rules and, in any event, they related to grants. As such, they were exempt from notice and comment under §553 of the Administrative Procedure Act.

Whatever the best way to rehabilitate the case may be, it’d be a mistake to take Sherley too seriously. A confused and poorly reasoned decision shouldn’t be read to shield agencies from judicial review whenever they happen to be following an executive order.

Follow Nick on Twitter: @nicholas_bagley

Tags: