« Federalism & Marijuana: Externalities vs. Minority Protection as Justifications for Federal Law | Main | Adrian Vermeule’s Deliberately Distorted Understanding of “Liberalism”: Why Liberalism and Secular Rationalism are, historically speaking, more antonyms than synonyms. »

Thursday, January 11, 2018

The Legality of Rescinding DACA With Minimal Reason-Giving: When are incomplete legal arguments "Arbitrary and Capricious"?

Adam Cox, Marty Lederman, and Cristina Rodriguez have offered four paragraphs over at Balkinization on why Secretary Duke's rescission of DACA might be illegal because Duke's reasons for the decision were "arbitrary and capricious" under APA section 706. (To be clear, the post is much longer than four paragraphs, but most of it is devoted to a primer on the legal meaning of DACA; only the last four paragraphs discuss the legality of rescission). Judge Alsup relied on such a theory in his opinion preliminarily enjoining this rescission. The indefatigable Josh Blackman wrote up a quick legal analysis for the the National Review in which he characterized Judge Alsup's opinion as "ludicrous," an analysis that I endorsed over Twitter.

An itchy twitter finger can lead one hastily to endorse views that one later regrets, but, even after reading Adam's, Marty's, and Cristina's typically measured and careful argument, I stand by my initial view that the DACA rescission is not arbitrary and capricious. The gist of the Cox-Lederman-Rodriguez (CLR) "substantive" (as opposed to Daniel Hemel's "procedural") argument against Duke's DACA rescission is that the Secretary relied exclusively on a legally insufficient reason contained in a one-page letter from Attorney General Sessions. Sessions, in turn, essentially relied on a simple syllogism that piggy-backed off of Judge Smith's opinion for the Fifth Circuit in United States v. Texas striking down the analogous deferred action program for parents of U.S. citizens ("DAPA"): (1) As per Judge Smith, DAPA was unauthorized by the Immigration and Naturalization Act ("INA"), (2) DACA is not legally distinguishable from DAPA, so (3) DACA is also unauthorized by the INA. CLR argue that this argument is legally insufficient, because DACA is, in fact, authorized by an obscure definitional provision of IRCA (8 U.S.C. section 1324h(a)(3)) allegedly giving the Secretary of DHS blanket authority to confer work authorization on any alien who is the beneficiary of deferred action, even if that deferred action stems entirely from an exercise of the Secretary's own enforcement discretion.

After the jump, I will take a deepish dive into the murky waters of section 1324h(a)(3) to offer my own assessment of this argument for statutory authorization. There is, however, a larger and more interesting point floating in this swamp of statutory detail: When an official rescinds an informal guidance of its predecessor on the grounds of wanting more fully to enforce a statute, how thorough must that official's reasons be? I am inclined to say, "not thorough at all." Cursory reference to roughly correct legal authority is good enough for government work here. My general reason -- more below -- is that an official's decision ought not to be vacated and remanded for more reason-giving when that official's reasons provided adequate notice to everyone about a legally sufficient basis for the decision. Where popular feelings run as high as they do in this area of deferred action, judges should not stall controversial policies with gratuitous demands for extra-precise reason-giving that, in substance, add nothing to the quality of the decision.. That sort of stalling tactic creates too great a risk of the judge's appearing to join #LegalResistance rather than writing a principled opinion.

1. Is DACA illegal?

I am not sure. The critical question is whether the Secretary of DHS is statutorily authorized to extend work authorization to anyone on whom she grants deferred action. CLR argue that section 1324h(a)(3) "expressly provides that the [Secretary of DHS] may 'authorize[]' aliens to be eligible for employment, even in cases where the statute itself does not directly authorize them to be employed." Well, maybe -- but keep in mind that 1324h(a)(3) merely uses seven words to define "unauthorized alien." One might respond that treating this seven-word definition as an "express" grant of authority to grant work authorization to anyone whatsoever just so long as the Secretary is entitled to defer deportation against them is hiding an elephant in a mouse hole. One might even go so far as Judge Smith to argue that, because such a delegation of power is unconfined by any intelligible principle, it transforms the Secretary into a mini-legislature under Indus. Union Dept. v. Amer. Petroleum Inst. and FDA v. Brown & Williamson Tobacco Corp. (See Texas v. United States at page 59 and note 178).

I do not wish to endorse Judge Smith's opinion or a narrow reading of section 1324h(a)(3), mostly because other people have already provided pretty thorough arguments for a narrower reading of section 1324h(a)(3). I recommend Peter Margulies' The Boundaries of Executive Discretion. CLR can respond that the broad reading of section 1324h(a)(3) is further supported by 8 CFR 274a.12, a 1987 rule that has been used to confer work authorization by several Presidents on classes of unlawfully present aliens not eligible for such authorization directly under the statute. As Marty has argued, this long pattern of administrative interpretation constitutes a gloss on the statute entitled to some deference from the courts. As Margulies notes, however, both DAPA and DACA cover more people for a more extensive period of time than prior exercises of discretion like the 1990 Family Fairness Program.

Of course, even if one endorses Texas v. United States's attack on DAPA, it does not automatically follow that narrower grant of deferred action and accompanying work authorization in DACA is also illegal. DACA covers fewer people: Perhaps this distinction places DACA closer to, say, the Family Fairness precedent than an open-ended dispensation from the INA. But once one goes down the path of counting beneficiaries and measuring length of authorization, one is outside the bounds of any "express" statutory authorization. Glossed with non-delegation canons, the authorization contained in section 1324h(a)(3) and 8 CFR section 274a.12 is at least ambiguous. As I noted in an earlier post), the opponents of Trump's various "travel bans" tend to read the President's authority under section 1182(f) narrowly, out of respect for canons like Brown v. Williamson's non-delegation principle. It is a little incongruous to take the position that the much more obscure authorization in section 1324h(a)(3) is somehow an "express" delegation of power.

I should be plain about my own views here. Far from endorsing either United States v. Texas or Margulies' argument for a narrow reading of section 1324h(a)(3), I am inclined to support the view that, if the Secretary as executive discretion to defer deportation, then she must have a corollary implied power to provide work authorization for those covered by the deferral. IMHO, as I have suggested elsewhere, it is ridiculous to allow the Secretary to refrain from deporting someone yet deny businesses the right to employ them: That practically requires the people who have been given a reprieve from deportation to violate the law on their employment. I am inclined, therefore, to read the ambiguous statutory power of the Secretary broadly to avoid such an odd policy consequence. The argument, however, that section 1324h(a)(3) is plain on this point strikes me as underwhelming. (BTW Marty and I tangled over this issue a bit earlier in the comments to the post just cited).

2. When is a minimal legal justification good enough for government work?

Suppose, however, that one agreed with CLR's reading of the law. After wading through this swamp of an ambiguous definitional phrase in IRCA, executive precedents under a 1987 rule, and a bunch of murky canons like Brown & Williamson, one finally concludes that the Secretary actually did have the legal power to confer work authorization on DREAMers contrary to General Session's letter opinion.

What then? Would such a conclusion entail the further conclusion that the Secretary's rescission is "arbitrary and capricious"? I do not think so.

Here's my defense. The point of reason-giving is to aid judicial review and improve the political process by informing judges and citizens of the basis for a decision so that they can challenge that basis either in an opinion or at the polls. Demanding that agencies give more detailed reasons for their decisions when those extra details will not perform either of these judicial or political functions is an admin law version of "Mother May I?" in which players who inadvertently omit some magic words get sent back to the start for no reason beyond the amusement of lawyers.

Secretary Duke's rescission is a case in point. Yes, she relied on a very spare letter that could have done a much better job of jockeying the legal sources and acknowledging rival arguments. Yes, General Sessions' argument would have been more persuasive had he argued that the legal basis for DACA was merely uncertain rather than unquestionably non-existent. Such a new, improved letter would cite the sorts of considerations canvassed above and discussed by Margulies, Judge Smith, and others -- murky statutory language, murky agency precedents, murky canons -- and would conclude that, because DACA is a controversial program with uncertain statutory justification, the Department should rescind the program and revert to "case-by-case" decisions about DREAMers. Note that such an argument rooted in legal uncertainty would run exactly parallel to the reason actually given by Duke/Sessions: It would declare that (1) the DHS should not offer a broad dispensation from statutory requirements unless the statutory basis for such dispensation is plain and (2) the statutory basis here is not plain here. Substitute the words "exists" and "does not exist" for "is plain" and "is not plain," and you have precisely the reasons actually offered by Duke/Sessions.

What possible practical function is served by forcing Sessions to produce such a new, improved letter? Does anyone have any doubt that, with a few rather modest amendments in the reasons, DACA's rescission should be upheld? Does anyone suffer from any uncertainty right now about what those modest amendments should be? Does not Sessions' actual letter pretty much suggest the gist of the revised letter that Duke/Sessions could provide that would adequately justify the rescission? If the answer to these three questions is "no," then it seems to me that forcing General Sessions or Secretary Duke to add a few sentences to the reasons they actually gave would be an empty exercise in Admin Law "Mother May I?"

Yes, I know that Chenery I seems to impose such a "Mother May I?" duty on agencies to specify precisely the legal grounds of the agency's decision. In Chenery I, however, one could make the argument that the SEC was hiding behind prior judicial decisions about corporate law rather than taking full responsibility for its own policy on the meaning of "fiduciary duties" under the Public Utility Companies Holding Act. Here, there is no doubt that Duke/Sessions are staking out the position that their statutory authority to grant blanket work authorization is constrained. Whether they stake out that position with unwarranted confidence or a more becoming diffidence surely makes no practical difference insofar as political or judicial accountability is concerned.

3. The political dangers of gratuitous demands for better legal reasons in polarized times

There is a deeper danger to demanding extra-precise legal justifications for repeals of guidance documents when popular feelings run as high as they do regarding DACA. Inevitably, these demands lead to accusations that the judge is throwing up technical speed bumps to an agency action that the judge knows should and will ultimately be upheld, simply because the judge dislikes the agency's action for policy reasons. If an agency's failure to give reasons leads to genuine uncertainty about why the agency acted or whether the agency's decision ultimately should be upheld, then, of course, by all means, vacate and remand for clarification. If the judge, however, can easily infer a sufficient legal basis for an agency's decision from the actual reasons submitted by that agency, then the judge's demanding more precise reason-giving looks suspiciously like the judge is running down the clock in hopes that, by stalling, the politics will shift and prevent the agency's action.

I am afraid that Judge Alsup's opinion reads a bit like such a stalling tactic. At page 32, Judge Alsup asks (rhetorically) "what exactly is the part of DACA that oversteps the authority of the agency?" and then proceeds to ask an answer a series of rhetorical questions about DACA's legality that has the sound of an orator's ringing peroration rather than a measured opinion resolving a close case. Judge Alsup's answers to one of his own questions has a question-begging quality that could lead an observer to wonder whether he had ideological priors. "Is [the illegality in DACA's] granting work authorizations coextensive with the two-year period of deferred action?" he asks. "No, aliens receiving deferred action have been able to apply for work authorization for decades." Well, yes, they have -- but, as Peter Margulies notes, there really has not been a program that covered such a large number of people for such a certain period of time. Why not at least acknowledge the uncertainty here rather than adopt a soap-box orator's indignant tone? Along with what sound to my ear like Judge Alsup's snide digs at the Trump Administration's ambivalence over repealing the DACA, this sort of rhetoric lends aid and comfort to the belief by conservatives that judges are simply inveterate opponents of Trump's hostility to immigrants rather than honest judicial brokers.

I should conclude this over-long post with my own admission of inveterate opposition to Trump's immigration policies. I am an immigration libertarian who believes that the repeal of both DAPA and DACA is a colossal mistake. I would go further to say that the entire Trump immigration agenda, from the various travel bans to his apparent limits on H1B visas, is IMHO destructive to our nation's best interests and rooted in economically obtuse predictions about the wage and productivity effect of foreign labor. But I also believe that we anti-Trumpistas should fight fair. To say that the DHS Secretary cannot disclaim discretion based on doubts about her statutory authority simply because she described such doubts as certainties rather than doubts looks like pointless pettifoggery to me -- a pettifoggery explicable only by Judge Alsup's hostility to the policy outcome reached, not the sufficiency of the legal reasons given.

(Just for the record, the "arbitrary and capricious" argument is much stronger, IMHO, than the procedural argument advanced by Daniel Hemel that DACA may be rescinded only through notice-and-comment procedure: To my mind, Hemel's procedural argument creates an indefensible one-way ratchet in which a guidance document can be promulgated without N&C yet only repealed through section 553 procedures. My confidence in my intuition that such a one-way ratchet cannot be the law is strengthened by William Funk's analysis, quoted by Jonathan Adler here).

Posted by Rick Hills on January 11, 2018 at 02:51 PM | Permalink


Thanks for this great discussion. I wanted to add one wrinkle. Rick's original post starts with the idea that an official's decision "ought not to be vacated and remanded for more reason-giving when that official's reasons provided adequate notice to everyone about a legally sufficient basis for the decision." Isn't there a question about whether or not that happened in this case?

As you all note, when DACA was rescinded, the Acting Secretary of DHS provided only a single reason: that DACA was illegal and DHS lacked power to grant deferred action to DACA recipients. This narrow justification had a practical impact on the record in the case challenging the DACA revocation. DHS argued that the record consisted only of a few hundred pages of publicly available documents -- all of which related to whether or not the program was legal.

The plaintiffs sought to supplement the record to include all of the material that DHS had considered in making its decision to revoke DACA. The government alleged it would have to review 21,000 documents—many of which the government claimed was privileged. The District Court granted the request to supplement the record and the government sought mandamus to overturn it. The Ninth Circuit denied mandamus. But the Supreme Court reversed, and held that decision be deferred until the district court determined whether or not DACA was reviewable. In re United States, 138 S.Ct. 371 (2017) and 138 S.Ct. 443 (2017).

The advantage for the government in relying on a purely legal justification for reversing policy decisions is that it could avoid the time and rigor expended in developing an administrative record. And the government doesn't seem to want to release all of the documents it considered when revoking DACA. Finally, the government's resistance to developing an administrative record, and reversing Obama-era policies, isn't limited to DACA. Jack Beermann similarly explores the risk of this approach in the administrations decision to reverse the Clean Power Plan. https://blog.harvardlawreview.org/the-deregulatory-moment-and-the-clean-power-plan-repeal/. The administration has similarly done so to justify reversals by the DOL on class action waivers in arbitration, as well as delays in rules designed to expedite student debt relief.

I'm ambivalent about the theoretical questions raised here about the ability of the government to reverse course on DACA. And I'm persuaded by Rick and other that it may not take much more than saying "we think the better way to go is for Congress to get involved," as the Obama administration once said. But given this background, aren't there very practical reasons -- reasons that go to the heart of State Farm -- to justify a more fulsome administrative record for the administrative reversals in these cases?

Posted by: Adam Zimmerman | Jan 12, 2018 8:00:23 PM

Does the APA recognize "guidance?" I know there are regs and rulings; where does guidance fit in?

Posted by: Biff | Jan 12, 2018 4:31:55 PM

There is a ton of great analysis here. I’d add just one small remark to try to capture what I take to be the central points of difference that I've seen in various takes by folks:

A basic question here is whether the Chenery, the APA, or administrative common law in general requires that an agency’s decision always be based on a “valid” ground. This may or may not be the best understanding of administrative common law. But if this is a requirement, then I think three things follow:

1. Concluding that agencies should not have to say much to justify a guidance doc would be orthogonal to the question whether a guidance can be upheld if the only reason given in defense of the guidance happens to be wrong as a matter of law. Saying little is different than saying something that is legally incorrect.

2. If the agency’s only defense of a decision is an interpretation of law that is wrong, it shouldn’t matter whether that wrong interpretation is “reasonable.” Here I think there may be confusion about what it means for an interpretation to be “reasonable.” If the agency has authority to adopt reasonable interpretations of an ambiguous statute, a la Chevron, then their “reasonable” interpretation is in fact *not wrong as a matter of law.* But that isn’t the way in which “reasonable” is being used here. The claim is not that there is deference for the agency to decide whether this deferred action policy is within the agency’s authority, and that the agency has acted within the zone of that delegated interpretive discretion. The district court concluded (properly in my view) that the deferred action policy is within the agency’s authority, full stop.

3. In any case where the agency had acted in excess of its authority on the basis of a legally incorrect understanding of its regulatory authority, I think everyone would agree that the agency’s decision must be vacated and remanded. Even if it were clear to the reviewing court that, on remand, the agency would be able point to a different statute that did in fact confer authority to undertake the action, I don’t think that would affect the reviewing court’s judgment.

So the difference between (A) a case where an agency wrongly concludes that it has authority under statute X, and (B) this case, where the agency wrongly concludes that it lacks authority, is not necessarily that in case A but not B the agency actually has acted in excess of its authority. In case A the agency may well not have acted in excess of its authority. Yet courts still seem to think that vacatur is always appropriate here, and would refuse to consider affirmance on the alternative ground that some other statute confers authority to the agency. I think that must be because—presumably under the influence of a Chenery-like principle—the agency proffered only a legally incorrect justification.

Those are the points, put together, are what lead me to think that administrative law supports the notion that if an agency rescinds a guidance based on the exclusive justification that the agency lacks the legal authority to issue the guidance, and if a court concludes that the agency’s conclusion is legally incorrect, then the court should send the decision back to the agency.

As Rick notes, this might mean little more than a bit of annoying hoop-jumping by the agency. And whether it is a good use of judicial oversight in general, or in this particular case, turns on lots of difficult judgments about how often agencies will be inclined to try to avoid political responsibility for difficult judgements by hiding behind arguments that the law has tied the agency’s hands, about how effective this strategy is likely to be at actually insulating the relevant political actors, about how much we worry about providing lower court federal judges with more opportunities to intervene in politically fraught cases like this one, etc. But it does seem to me a relatively straightforward application of some general, pretty widely accepted, administrative law principles.

Posted by: Adam Cox | Jan 12, 2018 2:46:00 PM

In reply to Professor Hills' comment, what if - and I'm not sure the district court has said this - the district court not only said that DACA was not certainly unlawful, but said it was certainly lawful? At that point, wouldn't the district court have to remand for the Secretary to exercise her enforcement discretion?

Second, are you sure you're not trading on what you know about this administration's policy preferences when you say you're sure that a remand will elicit reasoning that looks a lot like the reasoning we've got? What's different about this remand from your regular plain meaning remand, which you seem to think is generally appropriate? I vaguely recall Obama once saying that he lacked the discretion to do something like DACA (though I believe the meaning of these remarks is disputed and I only pose their existence as part of a hypothetical). Now, if a court somehow reviewed that view on the part of his Homeland Security secretary and said it wasn't so clear that the agency lacked discretion, would we be confident that on remand the agency would say, "well, we think the best reading of the statute is that we lack discretion"? I think we would probably expect the opposite. So how much assumption that trades on our knowledge of an administration's preferences is appropriate in deciding whether to plain-meaning remand or not?

On Proefssor Lederman's point, the way you describe what happened here does sound A&C in a non-colloquial sense, under, say, Encino Motorcars.

Posted by: Asher Steinberg | Jan 12, 2018 1:48:21 AM

I think it might help to distinguish two separate but related questions here:

1. Was it arbitrary and/or capricious for Sessions and the Secretary to rely upon Sessions's unelaborated assertion on Sept. 4 that DACA is "without proper statutory authority"?

2. Even if the answer to that is "no," now that the court has held that DACA *is* authorized by statute (or if the CTA9 or SCOTUS were to so hold), can the Secretary nevertheless continue to act upon the AG's contrary view, without any reason for thinking she'd rescind DACA if that were not legally compelled--or would that be either A&C or, perhaps, "not in accordance with law"?

On the first question, I think Rick and some commenters are right that the 09/04 legal conclusion would not necessarily be A&C, even if it turns out to be wrong--after all, Jerre Smith and *Senator* Jeff Sessions both embraced theories (of the work authorization statute, anyway) that at least raise questions about DACA's authority (well, on work authorization, at least--but not deferral of removal). Don't forget, however, that AG Sessions was thereby repudiating the *very* carefully and thoroughly articulated, official views of the United States, as elaborated in an OLC opinion and SCOTUS briefs by the SG. To perform such a volte-face *without any explanation whatsoever* might well be deemed A&C, at least in a colloquial sense (I'm not sure about under the APA).

Posted by: Marty Lederman | Jan 12, 2018 12:03:13 AM

Richard Fallon's recent essay in the Columbia Law Review about judicial candor has relevance here. He asks, "consider whether judges breach an obligation of candor when they provide reasons for their decisions but give little or no explanation of why they view those reasons as sufficient. Suppose a judge purports to justify a result by writing simply “Plaintiff wins because Marbury v. Madison so dictates,” with no supporting analysis of how or why Marbury bears on the case, when informed observers might think Marbury’s pertinence debatable. Viewed from one perspective, the judge’s reasoning is honest and transparent: She truly believes (let us assume) that Marbury controls. From another perspective, however, the opinion is opaque. Certainly the judge has provided no “candid recogni­tion of the difficulties of decision and the strength of competing arguments.” We might even suspect that she has willfully refused to do so. Perhaps she thought the legal arguments in the case so nearly in equipoise that she was entitled to choose both an outcome and a rationale from among a set of legally plausible alternatives and that, in her mind, Marbury controls only because she opted to say so in her opinion. What should we conclude about whether the judge has satisfied her obligation of candor under these imagined circumstances?"

He then explains that -- though such an opinion might fall short of fully achieving the ideal level of judicial candor -- it still satisfies the duty of judicial candor: "judges’ failures otherwise to disclose even outcome-dispositive aspects of their actual thinking normally breach no duty. ... Failures fully to achieve the ideal are common, breaches of judicial obligation rarer. We should ask our judges and Justices to do better, but with an awareness of why they may be unlikely ever to realize the ideal of judicial candor fully."

This is worth thinking about here, because Judge Alsup appears to be demanding a level of candor and a degree of nuance and rationale from the agency (in rescinding a guidance document(!)) that far outstrips what we demand of courts. In Fallon's terminology, the court is enforcing against the agency an *ideal* of candor that even courts rarely achieve. But as Rick Hills's original post explains, though, the agency has provided enough -- a Fifth Circuit opinion -- to have met the *duty* of candor--and that ought to be "good enough for government work."

Finally, it is also worth recalling that -- while Alsup's opinion may have been detailed and articulated -- courts don't *have* to write at such length when they write opinions; at least, they needn't in order to satisfy the *duty* of candor. Alsup *could* have just said, "The Administration loses, because the INA and its history of implementation shows they are wrong." Full stop. Had *that* been the opinion, many of those who welcomed it today would *still* have welcomed it--and would have provided post hoc rationales that the court *might* have articulated, but didn't, and urged its affirmance on those (non-court-supplied) rationales. If you fall in that camp, you might ask yourself why you are so comfortable with a judicial failure to supply a nuanced set of reasons, but yet so uncomfortable with an agency saying, "We can't do this, because we don't think the statute authorizes us to do it."

Posted by: Candid | Jan 11, 2018 10:38:00 PM

Asher writes:

"I thought Roberts also says later in his opinion that even if Randolph were right, a remand of the kind the majority ordered only makes sense if there's real doubt about whether the agency would reach the same result once disabused of its error."

Maybe Roberts says this -- but such a position makes no sense to me. Even if the agency would reach the same position on different grounds, there is value to the agency's providing the real basis for its opinion, insofar as democratic accountability is concerned. As "Slippery Slope" puts it, "[t]he Administration might be able to reach the same result by giving its real reasons for the repeal (whatever they are), but it can't rely on the fiction that it didn't have a choice."

My view is that, if the court knows that (a) the agency will rely on essentially the same sorts of reasons on remand on which it initially relied to justify its first decision and (b) the court would uphold the agency based on those slightly revised reasons, then there is little point to a remand.

So, yes, I agree with Marty that "a remand is proper, to see whether the Secretary would rescind DACA on policy grounds even once she knows she's not legally compelled to do so." I just do not see the rival "legal" and "policy" positions here as distinct enough to justify the remand.

To reiterate, I read Duke/Sessions' initial reasons as consistent with the view that, even if Jerre Smith's opinion for the Fifth Circuit is not the only possible correct understanding of executive power, it is still the best understanding, such that the Secretary should endorse this understanding as a matter of enforcement policy even if it is not compelled as a matter of law. I do not see a significant gains in political accountability from forcing the Secretary to acknowledge that there are rival views of what it means to exercise executive discretion: The "distance' between the "legal" theory announced in Sessions' letter and the "policy" position that might be elicited after remand is just too trivial to be worth the effort.

Posted by: Rick Hills | Jan 11, 2018 8:33:28 PM

I'm a bit confused by the comment thread: Are we all, in fact, agreeing that a remand is proper, to see whether the Secretary would rescind DACA on policy grounds even once she knows she's not legally compelled to do so?

Posted by: Marty Lederman | Jan 11, 2018 7:40:25 PM

I thought Roberts also says later in his opinion that even if Randolph were right, a remand of the kind the majority ordered only makes sense if there's real doubt about whether the agency would reach the same result once disabused of its error. That's the argument I took you to be making. Your post definitely doesn't suggest - and how could it? - that the Secretary didn't declare her hands were tied or exercised policy-making expertise. She relied on a letter stating that the policy was unconstitutional. So I thought you were saying, "why remand, this time the agency will just say the law is unclear but best read to forbid DACA."

Also, I'm not sure you're saying that remands of this kind are only appropriate if an agency's reasons are bad-faith evasions -- probably not -- but just in case you are, I don't agree; if the agency makes a good-faith error in believing its hands are tied, that's just as good a reason for a remand in my view. Indeed, an even better reason; if the agency's legal reasoning is a bad-faith front for its real policy judgment, at least there you know what will happen on remand and can argue the remand's a waste of time (though it isn't because we have to know what the agency's policy reasons *are* to sustain them, but at least it's a more arguable position). If the agency's mistaken in good faith, then you don't know what will happen on remand.

Posted by: Asher Steinberg | Jan 11, 2018 6:26:46 PM

PDK Labs is perfectly on point here, Asher. I think that CLR and I disagree on how to characterize Duke/Sessions’ reasons, not on the appropriateness of vacating and remanding if those reasons were bad-faith evasions of a statutory duty to make a a policy judgment. As Roberts puts the matter, “when an agency reads a statute in a particular way based on the erroneous belief that the reading was mandated by the statute (and thus the agency had no latitude to adopt a different interpretation), the case will be remanded so that the agency — now freed from its confined view of its own discretion — can reconsider its interpretation of the statute.”. Of course, that is correct. But Roberts and Randolph disagreed about whether the agency resolved the case at Chevron Step # 1 by declaring that its hands were tied by the statute (Randolph’s reason for vacating and remanding) or, instead, simply exercised its policy-making expertise to reach a reasonable view of the law (Roberts’ view).

I think that CLR and I might be having a similar disagreement about how to characterize Duke/Sessions’ reasoning here.

Posted by: Rick Hills | Jan 11, 2018 6:04:02 PM

Very reminiscent of then-Judge Roberts' arguments in PDK Labs:


Posted by: Asher Steinberg | Jan 11, 2018 5:38:58 PM

It seems to me that there's a huge difference between saying (1) we are repealing DACA because it's an illegal program; and (2) we are repealing DACA because we disagree with the policy choices reflected in it.

In this sense, Chenery I is a truth-in-government doctrine. The Administration might be able to reach the same result by giving it's real reasons for the repeal (whatever they are), but it can't rely on the fiction that it didn't have a choice.

More generally, I wonder whether risk of losing litigation (or more generally risk of litigation) is a legitimate ground to adopt (or change) a policy. It's not as if any court has held DACA unlawful. If the mere risk of losing is sufficient to justify a change in policy, an Administration could arrange for its friends to bring suit (or at least threaten to bring suit) and then use that suit (or threat) as the ground for changing the policy.

Posted by: Slippery Slope | Jan 11, 2018 5:08:35 PM

The comments to this entry are closed.