Column: The Coming Storm Over The 2020 Census

By Chris Dunn

Last week’s announcement by the Trump administration that it intends to include in the 2020 Census a question about citizenship marks the opening salvo in what promises to be a pitched battle over the upcoming decennial “Enumeration.” Within hours of the announcement, California filed suit challenging the citizenship question, New York announced it would lead a coalition of states in a similar challenge, and various non-governmental advocacy groups scrambled to respond.

Because of the highly polarized dynamic created by the election of Donald Trump, the upcoming Census promises to be a particularly contentious one, and the citizenship question is just the first of what undoubtedly will be many flashpoints. While these flashpoints likely will come in the form of technical methodological disputes, their resolution will have enormous consequences for our democracy, as the final Census results determine the distribution amongst the states of the 435 seats in the House of Representatives (and therefore in the Electoral College) and also determine vast amounts of federal funding flowing to states and localities.

Given these consequences, the Census has been the subject of extensive litigation, much of which has made its way to the Supreme Court. While it’s too early to predict the cases the Court might see over the 2020 Census, an examination of three of its most recent decisions provides important insights into what types of disputes might come to the Court, who can bring those disputes and when, and what types of constitutional and statutory claims the cases may offer.

Census Basics

The Census is a creature of the Constitution, with the so-called “Census Clause” being amongst the very first of constitutional provisions. Specifically, Article I’s section 2, which focuses on the apportionment of members of the House of Representatives (and which the Fourteenth Amendment amended as it relates to slavery), includes the constitutional mandate concerning the Census:

Representatives and direct Taxes shall be apportioned among the several States which may be included within this Union, according to their respective Numbers, which shall be determined by adding to the whole Number of free Persons, including those bound to Service for a Term of Years, and excluding Indians not taxed, three fifths of all other Persons. The actual Enumeration shall be made within three Years after the first Meeting of the Congress of the United States, and within every subsequent Term of ten Years, in such Manner as they shall by Law direct.

Pursuant to the broad mandate the Census Clause confers on Congress, federal legislation has regulated the conduct of the Census since 1790. And as a result of a congressional reapportionment stalemate following the 1920 Census, Congress adopted a scheme still in place today that prescribes a process by which reapportionment happens more or less automatically following completion of the Census and certification of the results by the President.

What is not automatic, however, is the conduct of the Census itself or the Presidential certification. Therein lies the room for controversy, as three recent Supreme Court decisions reveal.

Whom to Count

The most instructive of the Court’s recent Census decisions is the 1992 ruling in Franklin v. Massachusetts,[1] which involved a post-Census dispute about whether to count people physically outside the United States. Specifically, following the 1990 Census the Census Bureau decided to include in its final tally federal employees serving overseas and to count those employees as being residents of states designated in their personnel records as their “home of record.” Under this adjustment, Washington gained a seat in the House at the expense of Massachusetts, which sued contending inclusion of the overseas federal employees violated the Administrative Procedure Act and the Census Clause.

Starting with the APA, it serves as an important check on executive branch authority and promises to be central to efforts to resist many efforts by the Trump administration. Two key aspects of APA litigation are that the challenged action must be “final” and must be attributable to a federal “agency.” Unexciting as disputes about these terms may be, the Supreme Court’s interpretation of these terms in Franklin may have significant implications for coming disputes over Trump’s handling of the 2020 Census.

On the issue of “final” action, the Supreme Court in Franklin construed the relevant provision of what it had characterized as the “automatic reapportionment statute” to afford the President considerable discretion in how he handles final figures provided to him by the Census Bureau. And as a result of this discretion, the Court held that in an apportionment challenge the Census Bureau’s decision to include the overseas employees was not “final” action for purposes of the APA:

Section 2a does not expressly require the President to use the data in the Secretary’s report, but, rather, the data from the “decennial census.” There is no statute forbidding amendment of the “decennial census” itself after the Secretary submits the report to the President. For potential litigants, therefore, the “decennial census” still presents a moving target, even after the Secretary reports to the President. . . . Moreover, there is no statute that rules out an instruction by the President to the Secretary to reform the census, even after the data are submitted to him. It is not until the President submits the information to Congress that the target stops moving, because only then are the States entitled by § 2a to a particular number of Representatives.

The notion that the President may have considerable discretion to modify Census figures in determining apportionment is an alarming one when one thinks of Donald Trump as being the one with that discretion. That concern, however, could be allayed if the APA were available to address action he took that then would qualify as “final” action, but Franklin forecloses this by holding that the President is not an “agency” within the meaning of the APA:

The President is not explicitly excluded from the APA’s purview, but he is not explicitly included, either. Out of respect for the separation of powers and the unique constitutional position of the President, we find that textual silence is not enough to subject the President to the provisions of the APA.

Having dispensed with the APA claim, the Court turned to the constitutional claim under the Census Clause. As a threshold matter, only four Justices agreed the claim was justiciable, with the main sticking point being whether the courts could order necessary relief in light of the unsettled law concerning whether the President is subject to injunctive relief. On this extremely complicated point, the four-Justice plurality avoided the issue by concluding that an alternative form of meaningful relief was available to Massachusetts: declaratory relief against the Commerce Secretary (who has authority over the Census Bureau). The Justices based this conclusion on a proposition that looks questionable now: “[W]e may assume it is substantially likely that the President and other executive and congressional officials would abide by an authoritative interpretation of the census statute and constitutional provision by the District Court, even though they would not be directly bound by such a determination.”

On the merits of the Census Clause claim, the Court explained that “appellees argue that the Secretary’s allocation of overseas federal employees to the States violated the command of Article I, § 2, cl. 3, that the number of Representatives per State be determined by an ‘actual Enumeration’ of ‘their respective Numbers,’ that is, a count of the persons ‘in’ each State.” Massachusetts further argued that exclusion of overseas employees was required because the First Congress – to which the Supreme Court gives particular weight when construing constitutional provisions – specified in the original Census Act that those absent from home at the time of the Census should be counted as their “usual residence,” which for overseas employees would be their overseas location, not any particular state.

The Court rejected this argument with an analysis devoid of doctrinal standards and strongly suggesting the Census Clause imposes minimal constraints on Census methodology:

            In this case, the Secretary of Commerce made a judgment, consonant with, though not dictated by, the text and history of the Constitution, that many federal employees temporarily stationed overseas had retained their ties to the States and could and should be counted toward their States’ representation in Congress . . . . The Secretary’s judgment does not hamper the underlying constitutional goal of equal representation, but, assuming that employees temporarily stationed abroad have indeed retained their ties to their home States, actually promotes equality. . . .

How to Count

While Franklin dealt with whom to count during the Census, two other recent Supreme Court cases arising out of the 2000 Census dealt with how to count. How-to-count disputes are extremely important because so many people do not return Census forms or even receive them in the first place, thus leaving large data gaps that if not filled render Census results substantially inaccurate.

The first case was Department of Commerce v. U.S. House of Representatives,[2] which arose out of a decision by the Census Bureau to use statistical sampling to remedy “undercounting,” a phenomenon that historically has affected certain minority groups, children, and renters. After concluding the plaintiffs had standing because they had established the likelihood they would lose a seat under the sampling procedure, the Supreme Court held that the procedure violated a long-standing Census Act provision that on its face barred the Census Bureau from using sampling in counting persons. In light of this holding, the Court did not reach the constitutional claim under the Census Clause.

Department of Commerce cast considerable doubt on various Census Bureau techniques for addressing the undercount problem, but three years later the Supreme Court pulled back. In Utah v. Evans,[3] the Court considered a Census Bureau decision to deploy “hot-deck imputation” in determining the final 2000 Census numbers. Under this technique, the Census Bureau used an address list from the most recent prior Census (thus “hot” deck) to identify addresses that did not complete forms that nonetheless likely had residents and then imputed to those addresses characteristics exhibited by similar households for which actual information had been collected. Following this imputation, the 2000 Census was increased by about 1.2 million people and resulted in North Carolina gaining a House seat at the expense of Utah, which challenged the procedure.

As a threshold matter, the Supreme Court dealt with the critical issue of whether Utah could even challenge the reapportionment result given that the Clerk of the House of Representatives had issued a final certificate of reapportionment based on the imputation-based Census results. In Franklin only four Justices had found that a state could bring a post-certification challenge, but in Utah a majority held such a challenge was appropriate given that no congressional elections had yet taken place based on the reapportionment.

The Court then turned to and rejected Utah’s claim that imputation was indistinguishable from the sampling invalidated in Department of Commerce under the Census Act. Without delving into the intricacies of the two techniques, suffice it to say that the Supreme Court ultimately concluded that the Census Act’s ban on sampling had to be read relatively narrowly and that various methodological techniques for addressing undercounting could be distinguished from sampling.

Finally, the Court rejected Utah’s challenge to imputation as violating the Census Clause requirement that there be an “actual Enumeration.” While recognizing that imputation necessarily meant that not every person included in the final Census figures would have provided information, the Court held that this was entirely permissible: “[T]he text uses a general word, ‘enumeration,’ that refers to a counting process without describing the count’s methodological details.”

Looking Forward

Though still two years away, the 2020 Census promises to be highly contentious and is certain to spawn wide-ranging litigation. Under existing Supreme Court precedent, the Census Clause appears to pose only minimal limits on how the Census is conducted, but federal statutory provisions may be a richer source of restraint on the Trump administration. Equally importantly, just as they have proven pivotal in challenges to the administration’s attempted Muslim travel bans, federal constitutional limits on discriminatory and arbitrary government behavior may loom large in upcoming Census disputes.

This column originally appeared in The New York Law Journal.