The Census and the Constitution

This essay is prompted by indications that knowledgeable constitutionalists are misinterpreting Chief Justice Roberts’s ruling in Department of Commerce v. New York, plus the Law & Liberty editors’ request for a what-it-all-means retrospective on the case. What follows is analysis of factors underlying the New York decision; paths not taken in the wake of the ruling; and, finally but briefly, what New York means for future political litigation brought against this Administration. 

New York presented a multi-pronged, state-of-the-art challenge brought by New York State and NGOs to Commerce Secretary Wilbur Ross’s decision to include a citizenship question in the 2020 census. The case was expedited to the Supreme Court and, on June 27th, the Court held on the merits that: (1) including the question was lawful under the Constitution; (2) including it was lawful under Section 6(c) of the Census Act; (3) including it was lawful under the Section 141(f) of the Act; (4) there was a reasonable, objective basis for including the question; and (5) it was not unlawful for Secretary Ross to act, in part, for “unstated reasons,” such as “unstated considerations of politics, the legislative process, public relations, interest group relations, foreign relations, and national security concerns (among others).”

Having chalked up five significant wins, the Government lost. Speaking through the Chief Justice, the Court held as to a sixth issue that Secretary Ross had furnished “contrived” reasons for including the question in the census; hence, the case would be returned to the Department so Department officials could supply “genuine justifications” for their decision. 

It is this sixth merits ruling, the one adverse to the Government, that has drawn constitutionalist ire down on the Chief Justice. An example is a recent posting by Professor Stephen Presser, a respected scholar, charging the Chief Justice with legislating from the bench. Professor Presser recalls that Chief Justice Roberts once “boldly claimed” that Supreme Court justices should act as “apolitical ‘umpires,’” and he asserts that New York “makes clear that Roberts engage[s] in judicial legislation [that], in effect, put[s] the federal courts in position to frustrate countless policies of the executive.” 

A Fateful Concession

Professor Presser is right to worry about courts frustrating legitimate executive decision-making. And he is right to worry about disastrous consequences if judges are allowed to review administrative actions for “genuineness.” But is he right about New York? Consider the following measured but powerful argument that the Chief Justice declined to embrace:

[The district court decision] makes a mistake that is both contrary to law and dangerous in its potential effects on governance by seeking to plumb the motivation of administrative decision-makers, rather than evaluating the consistency of their actions with legal standards. [The] Court has made clear that, in general, for a court reviewing agency action, it is “not the function of the court to probe the mental processes” of the administrator. Morgan v. United States, 304 U.S. 1, 18 (1938) (Morgan II). The Court has warned that delving into the motives and thought processes of a decision-maker in a co-equal branch of government would be “destructive” of the responsibility of administrators and would undermine “the integrity of the administrative process.” United States v. Morgan, 313 U.S. 409, 422 (1941) (Morgan IV). 

[Except] in extraordinary circumstances, [inquiring into decision-makers’ motives] is not appropriate. It is especially inappropriate if it relies on extrinsic evidence, and most emphatically if it is based on queries to or examination of decision-makers. Such inquiries will chill discussion of potential government actions among a wider circle of officials—even though discussion among a broader set of officials frequently improves decisions. Pursuit of extra-record evidence of official motives in court should not be countenanced for the same reason that calling judges before the dock to answer questions about the motives behind their decisions is not permitted. See, e.g., Morgan IV, 313 U.S. at 422. 

Finally, changing the traditional [standards] of review to accommodate inquiries into official motives encourages use of judicial review not strictly as a means for keeping official actions within legal bounds but as extensions of political disputes into the judicial domain. This undermines the perceived legitimacy of the courts and intrudes on decisions committed to other branches. [Changing] the rules of judicial review to accommodate concerns about motives exacerbates problems associated with the political use of judicial fora.

The above analysis convinces me; I suspect it also convinces Professor Presser. Unfortunately for both of us, this powerful argument, drawn from amicus briefing by Ronald Cass and Christopher DeMuth, may well have been off-limits as grounds for the Court’s decision. 

Professor Presser, like other constitutionalists, overlooks the two most important words in the Roberts opinion: “conceded below” – as in the Government “conceded” in proceedings “below” that if it could be shown that the Secretary of Commerce’s decision “rested on a pretextual basis” that showing would “warrant a remand to the agency.” 

It was this unwise concession that left the Court with little choice but to rule in New York’s favor. To see how the Court was hemmed in, consider the following passage comprising effectively all of the Government’s advocacy on the crucial sixth merits issue. This 312-word composite, which fuses an extract from the Government’s principal brief with two from its reply, with emphasis added and minor changes, is the whole of the Government’s legal stand on the dispositive question: 

[T]o set aside an agency action that is supported by a rational justification, a court must find that the decisionmaker did not believe the stated grounds on which he ultimately based his decision, irreversibly prejudged the decision, or otherwise acted on a legally forbidden basis. See Mississippi Comm’n on Envtl. Quality v. EPA, 790 F.3d 138, 183 (D.C. Cir. 2015) (per curiam); Jagers, 758 F.3d at 1185; Air Transp. Ass’n, 663 F.3d at 488. 

[Here,] the district court found that the Secretary did not in fact believe his stated rationale for reinstating a citizenship question. Pet. App. 320a. Yet the court cited no evidence (much less “solid” evidence, ibid.) that the Secretary disbelieved DOJ’s letter and, instead, secretly thought that reinstating the citizenship question to the census would not be useful for [Voting Rights Act (“VRA”)] enforcement. The court’s finding thus has no basis in the record, let alone the compelling support necessary for a court to overcome the presumption of regularity and level a charge of deceit against a Cabinet Secretary who has taken an oath to obey the law. See United States v. Armstrong, 517 U.S. 456, 464 (1996).

[Specifically,] respondents did not show that the Secretary disbelieved his stated reasons, had an unalterably closed mind, or otherwise acted on a legally forbidden basis. Indeed, respondents have not identified any evidence suggesting that the Secretary thought DOJ’s analysis in its formal request for citizenship data was anything but genuine. 

[New York asserts] that the Secretary’s decision was pretextual because “DOJ did not exercise independent judgment” in stating its VRA rationale. There is no basis for that assertionNothing in the administrative record—or for that matter in the extra-record evidence—supports the contention that DOJ did not independently analyze the issue and independently conclude that census citizenship data would improve VRA enforcement for the four reasons identified in its letter. 

In sum, the Government demurred from contending, as do Dean Cass and Mr. DeMuth, that it is “contrary to law and dangerous” to allow courts “to plumb the motivation of administrative decision-makers.” Instead, it unwisely conceded that courts may “set aside an agency action” where they find “a decisionmaker did not believe the stated grounds on which he ultimately based his decision,” then weakly contended this evidentiary standard had not been met. As a result, the Court was forced to turn something of a blind eye to the Government’s own litigation choices to the extent it was determined to adhere to John Roberts’s famous pledge that, as Chief Justice of the United States, he would strive to the best of his ability to call legal balls and strikes as an impartial umpire. In New York, the Government’s pitch on the dispositive question was wide of the zone, and the Chief Justice called it accordingly. 

Signs of the Times

In retrospect, constitutionalists should have read the signs of the times and prepared themselves for New York. True, the Government’s most consequential decision was its the game-changing concession (see New York slip opinion, page 23). But even aside from that, other signs pointed to trouble for the Government.

First, in light of its concession about pretext, the Government might have anticipated the Chief Justice’s skepticism. Six days before the Court’s census decision, in Knick v. Township of Scott, the Chief Justice subtly admonished (footnote 5) the Solicitor General against pressing in the Supreme Court positions not advanced in lower courts. Fourteen years into his tenure, litigants ought to realize, for better or worse, that this Chief Justice is more committed than some to orderly proceedings and candor in litigation. Against this backdrop, and in light of a momentous concession, the Government’s effort – beginning with the second paragraph of its opening brief and continuing through the end of its reply – to sell the Court on an idea that Secretary Ross’s decision turned solely and exclusively on a weighing of the benefits of “census citizenship data” for “enforcement of the Voting Rights Act of 1965” risked falling on deaf ears. 

Second, the Government appears to have given insufficient attention to framing the issue. The Government stated the Question Presented as follows: “Whether the district court erred in enjoining the Secretary of Commerce from reinstating a question about citizenship to the 2020 decennial census” (emphasis added). But this formulation – “reinstating a question” – impliedly accepts a baseline in which the question in question is omitted. A more neutral articulation might have asked, “Whether the district court erred in enjoining the Secretary of Commerce from including a question about citizenship [in] the 2020 decennial census, as had been done, in one form or another, in all but two censuses since the beginning of the republic.” 

Third, the Government in all likelihood should have invoked the United States v. Morgan precedents. Under those decisions, litigants enjoy only limited ability to challenge administrative action by making claims about administrators’ mental processes. As noted above, those New Deal-era rulings were highlighted by Ronald Cass and Christopher DeMuth; they culminate in a respected 1941 opinion by Justice Felix Frankfurter: and they are cited to this day, including in the administrative-law casebook co-authored by Justice Breyer. The Government declined to mention the Morgan cases, as it made surprisingly restrained efforts to push back against a 1971 case said to stand for a proposition that intrusive discovery may be obtained merely on “strong showing” of “improper behavior.”

Finally, the Government should have recognized that the Justices likely appreciate that the Trump Administration is committed to drawing distinctions that make a real difference between citizens and non-citizens. In response to this recognition, the Government might have emphasized the helpful fact that citizenship is a status distinction approved by and enshrined in our Constitution. A better alternative to sole reliance on the benefits of collecting Voting Rights Act data would have been to say, straightforwardly, that the Secretary was determined to ask a customary question, included in practically every census for more than 200 years, about a status distinction mentioned in many constitutional provisions, including, most importantly, the Citizenship Clause of the Fourteenth Amendment. Tellingly, in all, or at least most all, other cases where the Constitution mentions statuses – slavery; titles of nobility; willingness to take religious oaths; attainders; corruption of blood – it is for purposes of outlawing them. 

To its credit, the Government came close (principal brief, page 28) to making this better sort of argument:

At the threshold, it simply cannot be arbitrary and capricious—or “irrational,” as the district court put it—to reinstate to the decennial census a question whose pedigree dates back nearly 200 years. Indeed, 2010 was the first time in 170 years that a question about citizenship or birthplace did not appear on any decennial census form. As the Secretary observed, “other major democracies inquire about citizenship on their census, including Australia, Canada, France, Germany, Indonesia, Ireland, Mexico, Spain, and the United Kingdom, to name a few.” The United Nations also recommends asking about citizenship on a census. Ibid. 

But while this passage adds color, it fails to advance a defensible litigation position. It suggests, if anything, that the Court ought to rule American administrative actions in- and out-of-bounds based on what the United Nations approves and other countries do, or else based on practices followed in the more distant – as opposed to more recent – American past. Good luck with that. 

Now consider these slight modifications under the heading of what might have been:

At the threshold, it simply cannot be arbitrary and capricious—or “irrational,” as the district court put it—to include in the decennial census a question whose pedigree is rooted in the Constitution and whose answer depends on a hotly contested definition forged in the crucible of the bloodiest war in American history. Indeed, only once in more than 150 years since ratification of the Fourteenth Amendment, and entrenchment of its Citizenship Clause, has a question about citizenship or birthplace not appeared on any decennial census form. Notwithstanding the impugning of the Secretary’s motives for asking such a question, other democracies similarly inquire about citizenship as part of their censuses, including Australia, Canada, France, Germany, Indonesia, Ireland, Mexico, Spain, and the United Kingdom. Even the United Nations—never the keenest friend of member-state sovereignty—recommends this unexceptionable practice. 

Ouch, Now What?

Although the Administration chose not to pursue the matter, asking the census question remained a viable option in the wake of the Court’s decision, so long as the Administration recognized that it could fashion a bespoke process for considering whether to include such a question in the census and then justifying any decision in favor of its inclusion. 

Under this alternative scenario, the Government might have assigned responsibility for crafting a decisional recommendation – and only a recommendation – to a panel of high-level, interagency decisionmakers untainted by the original decision. The panel’s essential goal would have been to articulate, independently of Secretary Ross and better than Secretary Ross, Secretary’s Ross’s own intuitions about citizenship and civic responsibility as they relate to including a citizenship question in the census. On this understanding, and working in complete isolation from the Secretary and everyone else who participated in the initial decision, the panel could have crafted an advisory recommendation, which the Secretary might then have accepted or rejected in a short, prompt, written memorandum. 

Such a panel likely would have noted that Article I, Section 8, Clause 4 of the Constitution assigns exclusive responsibility to the federal government for “establish[ing] a uniform Rule of Naturalization.” It likely would have stressed the affirmative rights, responsibilities, and opportunities associated with citizenship, including mandatory military service in the event of a draft, voting, service on juries, and holding federal office. Undoubtedly, it would have distinguished between immigration and naturalization and have observed that an effective naturalization process, under the express authority of Article I, permits and promotes higher levels of immigration. Likely, this civics lesson would not have omitted listing the constitutional references to citizenship; also likely, it would have stressed that citizenship is constitutionally approved, in contrast to other statuses that are constitutionally prohibited, even demonized. Quite possibly, the panel’s recommendation would have concluded with reference to the constitutional oath taken by officials and a contention that asking a question going to the success of the government’s naturalization programs, in light of the constitutional importance of citizenship, cannot possibly be an arbitrary, capricious, irrational, unconstitutional, or unlawful way for a Commerce Secretary to carry out his solemn obligations. 

Finally, a pro-tip enhancement to the process would have the advisory panel make identical (and simultaneous) up-or-down recommendations to both Secretary Ross and a second official who was uninvolved with the earlier proceedings. (This second decisionmaker would have been duly pre-designated by the Secretary to act is the Secretary’s stead in the event the Secretary were adjudged irredeemably compromised by his earlier decision.) True, running parallel processes would risk fatally conflicting decisions. But assuming the type of civics lesson described above could be composed, both decision-makers could be expected, independently, to arrive at the same decisional destination. In that event, the upshot would be, not fatal conflict, but mutual reinforcement.

What It Means

Academic constitutionalists unfamiliar with Supreme Court litigation are apt to attribute New York too much to the Court’s political proclivities—and too little to the Government’s advocacy activities. By the same token, appellate litigators are likely to overlook that the cleanest path to victory in the wake of the decision lay in a simple civics lesson, not in complex legal reasoning. 

The vital point is that the Government’s five hard-fought Supreme Court wins took technical issues of construing the Census Act and constructing a nationwide canvass off the table, thus opening previously unavailable pathways on remand. Because of these five victories, the Government could have convened an untainted advisory panel—with no expertise in census statutes or censuses—solely for purposes of assessing whether Secretary Ross, a businessman sworn to uphold the Constitution, had justifiable constitutional intuitions for including a citizenship question in the census. I, for one, suspect that such a committee, working independently, might readily pronounce heretofore unarticulated constitutional reasoning that the Secretary himself would embrace as a wordsmith’s reformulation of his own intuitive thinking. 

The Supreme Court’s New York decision is indicative of the high tide of litigation flowing against executive decisionmakers. The Government lawyers in New York prevailed on five merits arguments; fought a sixth to a near draw; met demanding deadlines; and demonstrated enviable mastery of, if not all, then almost all aspects of a brutally complicated and expedited case. They are to be congratulated. That said, the question of the hour is what lessons will be gleaned from the ups and downs of New York, as a litigation swell continues to break against this Administration.