In March of 2018, Secretary of Commerce Wilbur Ross announced his intention to add a citizenship question to the 2020 census. A collective group of eighteen states, the District of Columbia, cities and local governments, and several non-governmental organizations filed suit, claiming that the Secretary’s decision violated the Enumeration Clause and the Equal Protection Clause of the Constitution, and certain provisions of the Census Act and the Administrative Procedure Act. On Thursday, June 27, on the final day of its October Term of 2018, the Supreme Court issued its long-awaited decision in the case. Here is my discussion of Chief Justice John Roberts’ opinion for the Court in Department of Commerce v. New York.
Background of the Case
Each state has members in the U.S. House of Representatives. Some—like California and Texas—have a large number, and some—like Alaska and Wyoming—have a small number; the exact tallies depend on a state’s population. To determine the number of Representatives per state, the Constitution requires an “Enumeration” of the population every ten years (Art. 1, §2, cl. 3; Amend. XIV, §2). This is what we call the census, and it is conducted “in such Manner” as Congress “shall by law direct.” In 1790, Congress in the Census Act gave the Secretary of Commerce the authority to carry out the census “in such form and content as he may determine.” The Census Bureau—a federal agency within the Department of Commerce—also aids the Commerce Secretary in conducting the census.
The census does a plethora of other things in addition to simply counting the population of each state and the country itself. Historically, one of its objectives has been to tally the number of U.S. citizens. Indeed, from 1820 to 1950 (excluding 1840), the census asked every household in the U.S. about their citizenship status. From 1950 to 2000, the census scaled back its citizenship question, asking about only 20% of the population. To make this change, the Census Bureau split the census into two versions: the “short form”, which asked only basic demographic questions like sex, age, race, marital status, etc.; and the “long form”, which asked a greater array of demographic questions, including citizenship. In 2010, the Census Bureau changed the format of the census yet again. This time, the “short form” census was asked of all households. The “long form” census was scrapped, and its more-detailed questionnaire—which included citizenship status—was asked in the American Community Survey (ACS), sent to less than 3% of households on a random basis.
What gave rise to Dep’t. of Commerce v. New York was current Commerce Secretary Wilbur Ross’ push to reinstate the citizenship question on the 2020 census, a decision he announced in March of 2018. His reason? To better enforce the 1965 Voting Rights Act by improving data on the population of citizens at or above the voting age. The Justice Department (DOJ) had concluded the existing data on such persons—taken from the ACS—was insufficient and not ideal. Consequently, Secretary Ross determined the best course of action was to restore the citizenship question to the 2020 census.
Shortly thereafter, two groups (“Respondents”) sued Secretary Ross and the Commerce Department in the U.S. District Court for the Southern District of New York (“the District Court”). The first group—consisting of, inter alia, eighteen states and the District of Columbia—challenged the Secretary’s decision under the Constitution’s Enumeration Clause and the Administrative Procedure Act (APA). The second group—composed of non-governmental organizations—cited the Equal Protection Clause. In a decision published in January of this year, the District Court agreed with the first group, holding that the Secretary’s decision was arbitrary and capricious, was grounded in pretext, and was also a violation of the Census Act. The District Court disagreed with the second group, however, holding that the plaintiffs had not met their burden of proving that discriminatory intent underlaid the Secretary’s decision.
The Commerce Department appealed both to the Second Circuit Court of Appeals and to the U.S. Supreme Court. In this latter appeal, the Commerce Department urged the Supreme Court to jump the gun and take up the case before the appellate court considered it, stressing that the case was an “issue of imperative public importance” and also noting that the census forms needed to be printed before the end of June 2019. On February 15, the Supreme Court granted the petition. The questions the Court agreed to hear were essentially twofold: One, is a citizenship question on the census constitutional? And two, was Secretary Ross’ reason for adding the question sufficient, meeting the statutory requirements under the APA?
Chief Justice Roberts’ Opinion for the Court
Chief Justice John Roberts, in answering those questions, walks a tightrope, stretching out his hands to both polar ends of the Court’s ideological spectrum. To the first question, Roberts answers “yes,” where he is joined by the Court’s more-conservative bloc of Justices Clarence Thomas, Samuel Alito, Neil Gorsuch, and Brett Kavanaugh. But as to the second, Roberts answers “no,” siding with the more-liberal group of Justices Ruth Bader Ginsburg, Stephen Breyer, Sonia Sotomayor, and Elena Kagan.
After recounting the history of the census and the case itself, Roberts first addresses whether the Court has jurisdiction to rule. Article III of the Constitution bestows on the Judiciary the power to resolve genuine “Cases” and “Controversies.” For a matter to qualify as a real case or controversy, a plaintiff must establish standing. To have standing, a plaintiff must “present an injury that is concrete, particularized, and actual or imminent; fairly traceable to the defendant’s challenged behavior; and likely to be redressed by a favorable ruling” (quoting Davis v. Federal Election Commission (2008)). The District Court determined that the addition of the citizenship question to the 2020 census would result in a number of sufficient injuries and concluded some of the Respondents have standing to sue. Chief Justice Roberts, for a unanimous court, agreed with the District Court’s conclusion about standing, writing that the District Court’s “findings of fact were not so suspect as to be clearly erroneous.” Thus, he declares that “at least some [R]espondents have Article III standing,” and the Court therefore may rule.
Roberts then turns to Respondents’ first claim—that the Enumeration Clause bars the Secretary’s attempt to add the citizenship question. This is where the unanimity breaks down; only Justices Thomas, Alito, Gorsuch, and Kavanaugh join Roberts here.
Without any delay, Roberts makes clear: “The Enumeration Clause of the Constitution does not provide a basis to set aside the Secretary’s decision.” The Clause’s language, he continues, “‘vests Congress with virtually unlimited discretion in conducting the decennial actual Enumeration, and Congress ‘has delegated [that] broad authority . . . to the Secretary'” (quoting Wisconsin v. City of New York (1996)). Since the Census Act in 1790, every branch of government has understood that the Secretary (via Congress) may conduct the census “in such form and content as he may determine,” including asking questions beyond a simple population count. Recall also that a citizenship question has been asked of at least some households on 18 of the last 20 censuses, and has been asked of all households in 13 of the 23 total censuses throughout our nation’s history. 2010 was the first census since 1840 to not ask the citizenship question (even though that question was present on the 2010 ACS form sent to a fraction of households). “That history matters,” says Roberts. “In light of the early understanding of and long practice under the Enumeration Clause,” Roberts holds that the Enumeration Clause “permits Congress, and by extension the Secretary, to inquire about citizenship on the census questionnaire.”
After first disposing of the government’s contention that the Secretary’s decision is unreviewable under the Administrative Procedure Act, Roberts then turns to the second claim—that the Secretary, by adding the citizenship question, “abused his discretion” under the Constitution and the Census Act. The District Court determined he did so for two reasons: First, that his decision “was not supported by the evidence before him”; and second, that “his stated rationale was pretextual.”
But Roberts disagrees with the District Court’s first finding for a number of reasons. Roberts says that the Secretary “examined the [Census] Bureau’s analysis of various ways to collect improved citizenship data”; “considered but rejected the Bureau’s recommendation to use administrative records alone”; and “reasoned” that “[a]sking a citizenship question of everyone . . . would eliminate the need to estimate citizenship for many of th[e] [voting-age] people.” After “consider[ing] the relevant factors, weigh[ing] risks and benefits, and articulat[ing] a satisfactory explanation,” the Secretary made his decision. So, in Roberts’ mind, when the District Court overrode the Secretary’s decision, it “improperly substituted its [own] judgment for that of the agency”—a discretionary judgment the Census Act gives to the Secretary and the Bureau, not the courts.
After dismissing a pair of backup arguments that the Secretary violated twin provisions of the Census Act, Roberts returns to the District Court’s determination that the Secretary abused his discretion—specifically its second reason, that the Secretary’s “stated rationale” for adding the question was “pretextual.”
Here is where Roberts—and the Court—does his about-face. Justices Thomas, Alito, Gorsuch, and Kavanaugh no longer join Roberts in his opinion. Roaring into their place come Justices Ginsburg, Breyer, Sotomayor, and Kagan. Until this point, Roberts has sided with the government—disposing of Respondents’ arguments pertaining to the Enumeration Clause and the Census Act. But now, Roberts sides against the government, agreeing with the District Court and holding that the Secretary’s decision “rested on a pretextual basis.”
The Secretary’s offered reason for adding the citizenship question was that it would strengthen a key provision of the Voting Rights Act (VRA), at the request of the DOJ. But “viewing the evidence as a whole,” Roberts “share[s] the District Court’s conviction that the decision to reinstate a citizenship question cannot be adequately explained in terms of the DOJ’s request for improved citizenship data to better enforce the VRA.” For one, the administrative record shows that the Secretary moved to add the citizenship question “about a week into his tenure,” Roberts writes, “but it contains no hint that he was considering VRA enforcement in connection with that project” at that time. The Secretary then attempted to “elicit requests for citizenship data” from the Department of Homeland Security and the DOJ’s Executive Office for Immigration Review, but none of those requests made any mention of the VRA. Nor are either of those departments responsible for enforcing the VRA. Next, the Secretary “asked Commerce [Department] staff to look into whether [he] could reinstate the question without receiving a request from another agency” (emphasis added). At long last, the DOJ’s Civil Rights Decision suggested it “might be willing to request citizenship data for VRA enforcement purposes.”
“Altogether,” Roberts says, “the evidence tells a story that does not match the explanation the Secretary gave for his decision.” Specifically, the Secretary opined that the Commerce Department “was simply acting on a routine data request from another agency.” But the record says otherwise, “indicat[ing] that Commerce went to great lengths to elicit the request from DOJ (or any other willing agency).” So, to Roberts and the four more-liberal Justices, the “VRA enforcement rationale—the [Secretary’s] sole stated reason—seems to have been contrived.”
With some rhetorical flourish, Roberts explains the Court’s conclusion in the administrative law context:
The reasoned explanation requirement of administrative law, after all, is meant to ensure that agencies offer genuine justifications for important decisions, reasons that can be scrutinized by courts and the interested public. Accepting contrived reasons would defeat the purpose of the enterprise. If judicial review is to be more than an empty ritual, it must demand something better than the explanation offered for the action taken in this case.
Lastly, Roberts points out that the decision to add a citizenship question to the census was not itself “substantively invalid.” It was only so invalid because the Secretary’s reasoning was “more of a distraction,” and the APA calls for a “[r]eason[able] . . . explanation for administrative action.”
In sum, then, the disposition of the Court can be laid out as follows: The Enumeration Clause of the Constitution does not prohibit the addition of a citizenship question to the census. That said, because there was no citizenship question on the 2010 census, adding one to the 2020 census requires a reasonable, credible explanation under the APA. To date, Secretary Ross’ only explanation—obtaining citizenship data to better enforce the VRA, per the DOJ—was contrived. Therefore, there will be no citizenship question on the 2020 census until (or unless) the Secretary presents a more reasonable and more credible explanation. In other words, Roberts orders a do-over.
Justice Thomas’ Opinion
Justice Thomas, joined by Justices Gorsuch and Kavanaugh, penned an opinion concurring in part and dissenting in part. Thomas joins all of Roberts’ majority opinion up until Part V, when Roberts invalidates the Secretary’s decision due to his pretextual rationale. To Thomas, the Court’s holding there “reflects an unprecedented departure from our deferential review of discretionary agency decisions.”
After recounting the factual history of the case, Thomas cuts right to the chase: “Part IV–B of the opinion of the Court correctly . . . conclude[s] that the Secretary’s decision survives ordinary arbitrary-and-capricious review,” and that holding “should end our inquiry.”
“But the opinion continues,” Thomas writes. “According to the Court, something just ‘seems’ wrong” with the Secretary’s stated reason to add the citizenship question (namely, to better enforce the VRA), and as a result, the majority sides with the District Court and overturns the Secretary’s decision. That conclusion, Thomas argues, is “extraordinary,” and the majority reaches it via an “unauthorized inquiry into evidence not properly before us.”
Thomas attempts to directly undermine the majority’s reliance on “pretext.” The APA permits a court to canvass the administrative record in order to decide whether an agency’s “action, findings, and conclusions” satisfy a number of standards. But “none” of those standards, Thomas maintains, “instructs the Court to inquire into pretext.” Moreover, if a federal agency’s public findings and determinations withstand a court’s APA-based scrutiny, “the APA does not permit a court to set aside the decision solely because the agency had ‘other unstated reasons’ for its decision, such as ‘political considerations’ or the ‘Administration’s priorities'” (quoting Roberts’ majority opinion). To Thomas, it is “unsurprising, then, [that] this Court has never held an agency decision arbitrary and capricious on the ground that its supporting rationale was ‘pretextual.’ Nor has it previously suggested that this was even a possibility.” Quite the opposite, in fact, says Thomas: “Under settled provisions of administrative law . . . pretext is virtually never an appropriate or relevant inquiry for a reviewing court to undertake.”
Thomas also makes a few more points in the administrative law context. First, “[u]ndergirding our arbitrary-and-capricious analysis is our longstanding precedent affording the Executive a ‘presumption of regularity'” (quoting Citizens to Preserve Overton Park, Inc. v. Volpe (1971)). That presumption, then, means the Court “give[s] the benefit of the doubt to the agency.” Second, Thomas asserts that the majority “errs at the outset by proceeding beyond the administrative record to evaluate pretext.”
Thomas then reverts back to attacking the majority’s finding of pretext itself, arguing that “[e]ven if it were appropriate for the Court to rely on evidence outside the administrative record, that evidence still fails to establish pretext.” As we know, the Secretary’s rationale for adding the citizenship question was to better enforce the VRA via better citizenship data, at the request of the DOJ. To Thomas, there is a fundamental distinction being missed by the majority. There is a difference between that rationale being untrue or not factoring into the Secretary’s decision at all (as the majority concludes), and that rationale not being the principal reason for the addition (as Thomas would conclude). Put differently, “the evidence cited by the Court suggests at most that the Secretary had ‘other unstated reasons’ for reinstating the citizenship question,” Thomas says (quoting Roberts’ majority opinion). But none of that evidence “suggests, much less proves, that the Secretary harbored an unstated belief that adding the citizenship question would not help enforce the VRA.” Nor does that evidence suggest or prove that “the VRA rationale otherwise did not factor at all into his decision.” Instead, it “simply suggests that a number of executive officials agreed that adding a citizenship question would support VRA enforcement.”
Finally, Thomas addresses the lasting implications of Roberts’ decision. To Thomas, “the Court has opened a Pandora’s box of pretext-based challenges in administrative law.” This is because Dep’t. of Commerce v. New York is the “first time the Court has ever invalidated an agency action as ‘pretextual.'” Taken together, Thomas gathers, Roberts’ opinion has three forseeable consequences:
Now that the Court has opened up this avenue of attack, opponents of executive actions have strong incentives to craft narratives that would derail them. Moreover, even if the effort to invalidate the action is ultimately unsuccessful, the Court’s decision enables partisans to use the courts to harangue executive officers through depositions, discovery, delay, and distraction. The Court’s decision could even implicate separation-of-powers concerns insofar as it enables judicial interference with the enforcement of the laws.
“In short, Thomas concludes, “today’s decision is a departure from traditional principles of administrative law. Hopefully it comes to be understood as an aberration—a ticket good for this day and this train only.”
Justice Breyer’s Opinion
Justice Breyer also wrote an opinion concurring in part and dissenting in part. Justices Ginsburg, Sotomayor, and Kagan joined his opinion in full. Breyer agrees with Roberts in that the Secretary’s decision to add the citizenship question was based on a pretextual rationale. However, Breyer writes mostly to dissent from the conservative-joined majority’s conclusion that the Secretary’s decision did not violate twin provisions of the Administrative Procedure Act (APA) (see supra, ¶13).
Breyer’s argument on this point is laid out in one of his opening paragraphs:
[T]he Secretary’s decision to add a citizenship question created a severe risk of harmful consequences, yet he did not adequately consider whether the question was necessary or whether it was an appropriate means of achieving his stated goal. The Secretary thus failed to articulate a satisfactory explanation for his decision, failed to consider important aspects of the problem, and offered an explanation for his decision that runs counter to the evidence, all in violation of the APA (internal quotation marks and citations omitted).
Based on the administrative record, it seems to Breyer that the Secretary’s decision to reinstate the citizenship question “rests upon a weighing of potentially adverse consequences (diminished responses and a less accurate census count) against potentially offsetting advantages (better citizenship data).” But “[i]n my view,” Breyer writes, “the Secretary did not make reasonable decisions about these potential costs and benefits in light of the administrative record.”
First, the Secretary determined he was “not able to determine definitively how inclusion of a citizenship question on the decennial census will impact responsiveness.” But if such a conclusion implies that adding a citizenship question is doubtful to impact responsiveness one way or another, Breyer argues, the administrative record “indicates the contrary.” For example: In Breyer’s view, the administrative record includes “repeated Census Bureau statements”, “studies”, and “recommendations” from “[s]everal [s]tates”, all of which declare that adding the question would produce a less accurate count because noncitizens and Hispanics would be less likely to respond to the questionnaire.” But when the Secretary looked at this administrative data, he “concluded that neither the Census Bureau nor any stakeholders had provided ‘definitive, empirical support’ for the proposition that the citizenship question would reduce response rates.”
Second, ponder the Secretary’s next determination: Even if the citizenship question were to depress response rates, “the value of more complete and accurate data derived from surveying the entire population outweighs . . . concerns” about diminished accuracy (emphasis added by Breyer). “That conclusion,” Breyer states tersely, “was also arbitrary.” If anything, the “administrative record indicates that adding a citizenship question to the short form would produce less ‘complete and accurate data,’ not more” (emphasis added). That indication comes from a Census Bureau recommendation to the Secretary. That recommendation, Breyer says, contained the following: First, “for about 90% of the population, accurate citizenship data is available from administrative records” in two federal agencies. Second, the Census Bureau was “high[ly] confiden[t]” it could construct a “statistical model that would accurately impute citizenship status for the remaining 10% of the population.” Therefore, the Census Bureau advised the Secretary that using those two instruments “alone . . . would yield more accurate citizenship data than also asking a citizenship question” on the census.
Based on this reading of the administrative record, Breyer sees plain as day “a serious legal problem”: “How can an agency support the decision to add a question to the short form, thereby risking a significant undercount of the population, on the ground that it will improve the accuracy of citizenship data, when in fact the evidence indicates that adding the question will harm the accuracy of citizenship data” (emphasis in original)?
“Of course it cannot,” Breyer answers. “But as I have just said,” nowhere does the evidence in the administrative record indicate that more accurate citizenship data follows from adding the citizenship question—and the District Court came to the same conclusion. Yet the Secretary put the shoe on the other foot and did a 180, “stat[ing] simply that ‘[a]sking the citizenship question of 100 percent of the population gives each respondent the opportunity to provide an answer,’ which ‘may eliminate the need for the Census Bureau to have to impute an answer for millions of people.'” To Breyer, the Secretary “therefore must have assumed, sub silentio, exactly what the Census Bureau experts urged him not to assume—that answers to the citizenship question would be more accurate than statistical modeling” (emphasis added).
As a result, Breyer writes that in addition to the conclusion that the better-enforcement-of-the-VRA rationale was “unconvincing,” the Secretary “failed to consider important aspects of the problem and offered an explanation for his decision that runs counter to the evidence before the agency” (internal quotation marks and citation omitted). Thus, Breyer would hold that the Secretary’s decision to add the citizenship question was unconstitutional not only because it was grounded in pretext, but because it also was “arbitrary and capricious, and an abuse of his lawfully delegated discretion.”
Justice Alito’s Opinion
Justice Samuel Alito brings up the caboose, writing only for himself in an opinion concurring in a smidgen and dissenting in a ream. First, Alito would side with Thomas, Gorsuch, and Kavanaugh—and against the majority—and agree that the Secretary’s decision to add a citizenship question was not based on a pretextual motive. But Alito goes further. Even Thomas, Gorsuch, and Kavanaugh would at least maintain that the Secretary’s decision was reviewable under the APA. Alito, however, would not. “To put it bluntly,” Alito states, “the Federal Judiciary has no authority to stick its nose into the question whether it is good policy to include a citizenship question on the census or whether the reasons given by Secretary Ross for that decision were his only reasons or his real reasons.”
The APA “authorizes judicial review of ‘agency action’ taken in violation of law” (5 U.S.C. §§706(2)(A)–(D)). However, §701(a)(2) prohibits judicial review of agency action that is “committed to agency discretion by law.” Although the Court “has characterized the scope of §701(a)(2) as ‘narrow’, . . . there are circumstances in which it applies” (citing Heckler v. Chaney (1985)). In determining its applicability in administrative law cases, the Court has looked to a number of factors, including:
- “[W]hether the text and structure of the relevant statutes leave a court with any meaningful standard against which to judge the agency’s exercise of discretion” (internal quotation marks and citation omitted);
- And “whether the matter at hand has traditionally been viewed as committed to agency discretion.”
Alito then applies those two factors to the decision of the Secretary to add a citizenship question to the census. First, are there any statutes that lay out a standard by which a court can review the Secretary’s decision? Alito says no, adducing 13 U.S.C. §141(a). That statute gives the Secretary the sweeping authority to “take a decennial census of population . . . in such form and content as he may determine, including the use of sampling procedures and special surveys” (emphasis added by Alito). In other words, Alito argues, §141(a) gives the Secretary “unfettered discretion to include on the census questions about basic demographic characteristics like citizenship.” Hence, no standard by which to review that decision.
Second, have courts historically viewed the Secretary’s decision to add a citizenship question “as committed to agency discretion?” Yes, says Alito. In such cases where courts have found an agency’s action to be committed to agency discretion by law, “an important factor has been the absence of an established record of judicial review prior to the adoption of the APA.” That is exactly the case here. The first census in 1790 “went beyond a mere headcount and gathered additional demographic information, and during virtually the entire period prior to the enactment of the APA, a citizenship question was asked of everyone.” “Notably absent from that long record,” Alito stresses, “is any practice of judicial review of the content of the census.”
In Alito’s eyes, then, “neither respondents nor my colleagues have been able to identify any relevant, judicially manageable limits on the Secretary’s decision to put a core demographic question back on the census.”
Thus, Alito concludes with the following:
Throughout our Nation’s history, the Executive Branch has decided without judicial supervision or interference whether and, if so, in what form the decennial census should inquire about the citizenship of the inhabitants of this country. Whether to put a citizenship question on the 2020 census questionnaire is a question that is committed by law to the discretion of the Secretary of Commerce and is therefore exempt from APA review. The District Court had the authority to decide respondents’ constitutional claims, but the remainder of their complaint should have been dismissed.