Supreme Court and the 2020 census citizenship question – Part B-1

t B-1

Part B-1 of this post discusses one of two key decision points in the Supreme Court’s decision on the citizenship question in the 2020 census, the Constitution’s Enumeration clause. Part B-2 will cover the second decision point, the Administrative Procedure Act. Part A covered the Court’s opinions on both these points.

Enumeration Clause

The history of the census – citizenship nexus

My lay opinion says that the New York District Court and the Supreme Court probably erred in finding that the Enumeration Clause was not relevant to the 2020 Census citizenship question issue. As highlighted below, I say this for two reasons. One focuses on the administration’s misinterpretation/misrepresentation of census history. The second covers the difference in context between the Court’s decision on the Wisconsin case and the citizenship question case in light of Utah v. Evans. An earlier post addressed both these two Court cases.

The New York district court dismissed the Enumeration Clause complaint largely due to the administration’s claim about the near-continuous role a citizenship question had played in the census since the initial census. This history was frequently mentioned in the conservative Supreme Court justices as a significant rationale for their decisions.

In its motion to dismiss in the New York District Court the administration included three pages about the history of the citizenship question in the decennial census. This included two key summary sentences: “With the exception of 1840, censuses from 1820 to 1880 asked for citizenship or birthplace in some form. Decennial censuses from 1890 through 1950 specifically requested citizenship information more consistently.”

However, the information provided to the Court, in the words of Thomas Wolfe’s and Brianna Cea’s Georgetown Law Journal article, “is misleading where it is not outright false.” Given its directness and simplicity, the following provides the summary of the article’s point regarding the historical narrative provided by the administration.

“First, the census has never asked for the citizenship status of everyone in the country. Secretary Ross’s proposal is therefore historically unprecedented.

“Second, the administration relies on an impoverished view of census history to suggest that Secretary Ross can find a historical warrant for his decision in citizenship questions that were posed only to small subsets of the population at various points in American history. Viewed in context, these citizenship questions originated as sporadic components of an approach to census-taking that the Census Bureau long ago rejected as incomparable with its foundational constitutional goal of actual enumeration. . . .  In the wake of the 1950 Census, the Census Bureau rejected this older paradigm of census practice in favor of a radically different model. Indeed, social science techniques like sampling granted the Bureau the technical ability to identify and remedy substantial problems in its approach to enumeration. The Bureau overhauled its approach dramatically. As part of this overhaul, the Census Bureau rebuffed citizenship questions as viable items for any census survey designed to obtain a complete count of the population. Due to intervening developments in the American immigration environment, these questions have never been deemed fit to return to the complete-count form: they have been confined to solely sample surveys sent only to subsets of the population.

“This Article will show that under these circumstances, the Administration cannot plausibly invoke census history to justify its current decision to add a new, untested citizenship question to the 2020 Census either under the Enumeration Clause or the Administrative Procedure Act. History instead creates a broad presumption against Secretary Ross’s proposal, one which the Administration has not succeeded in rebutting.”

One may also want to read Justin Levitt’s article in the Columbia Law Review. Levitt also understands citizenship and census history like Wolf and Cea.

In sum, seeing the history of the decennial census as having a long connection with citizenship is flatly erroneous. To the extent that this false history of the historical nexus between citizenship and the census significantly influenced the decisions of the New York District Court and the Supreme Court, these decisions rested on a very weak foundation.

The real-life impact of a small change in the census population count

In Wisconsin v. City of New York, the Court determined that the decision of the Secretary of Commerce not to use an adjustment (then called the postenumeration survey) to the decennial census’s actual enumeration conformed to applicable constitutional and statutory requirements. New York City had argued that the use of the postenumeration survey would produce a more accurate count of the population. The Secretary of Commerce had decided that the use of the postnumeration survey might produce a more accurate count but determined that the core constitutional use of the decennial census was “distributive accuracy.” He concluded that the use of a postenumeration survey would not alter the distributive accuracy of the 1990 census. The essential constitutional purpose of the decennial census was its use to allocate seats of the House of Representatives among the states. He believed there was no compelling evidence that suggested the use of the postenumeration survey would alter the allocation of House seats among the states.

In Utah vs. Evans, the Court upheld the Census Bureau’s decision to use “hot-deck imputation” in the 2000 census. The use of hot-deck imputation added 1.2 million people to the census’s enumeration count, an increase of 0.4% of the total population count. But this very small percentage increase was not spread evenly across the country. North Carolina’s population increased by the national average of 0.4% but Utah’s increase in population was only 0.2%. This set of changes resulted in North Carolina gaining one seat in the House. Utah lost a seat that it would have otherwise gained if hot-deck imputation had not been used. Roughly, Utah’s population increased by 45,000 compared to North Carolina’s increase of about 321,000. This differential cost Utah a House seat. Essentially, a more accurate distribution of the population resulted in a decision that better matches the constitutional purpose of the decennial census.

In sum, the Wisconsin decision suggests that distributive accuracy is the essential constitutional rationale for the decennial census. The Utah case demonstrates how small numbers in the change in the population distribution can result in a loss or a gain or both of a House seat.

The detailed material available through the California and New York Districts well indicates that the professional and technical positions almost invariably concluded that adding a citizenship question would result in a lower quality of census results than would otherwise be the case. It suggested that the Latino and related undercount would be significant and affect the distribution of the population estimates. Most analyses suggested that three states were at risk of losing a House seat (or not gaining a seat they were otherwise entitled to) if the census included a citizenship question: California, Florida, or Texas.

Levitt puts it this way: “Secretary Ross made the determination to ask this question despite his own contemporaneous admission that the career staff of ‘the Census Bureau and many stakeholders expressed concern [that the decision] would negatively impact the response rate’ for enumeration, and despite the absence of any opportunity to test that impact before implementing the change. He did so despite the direct warning of six former Directors of the Census Bureau, whose collective 25 years of service as Director spanned eight Presidential Administrations (of both political parties). That adding the question so late in the planning process would put the accuracy of the enumeration ‘at grave risk.'”

In 1980, a time in which the issues of immigration and citizenship were much less salient and volatile than they are now, the Census Bureau argued in federal court against adding a citizenship question:

“Any effort to ascertain citizenship will inevitably jeopardize the overall accuracy of the population count. Obtaining the cooperation of a suspicious and fearful population would be impossible if the group being counted perceived any possibility of the information being used against them. Questions as to citizenship are particularly sensitive in minority communities and would inevitably trigger hostility, resentment, and refusal to cooperate.”

This consequence of a Latino-related undercount within a state may be is as important as the allocation of House seats, which refers to the population count for each state. The distribution of Latino-related households and immigrant-related households within a state is not evenly distributed. Consequently, intra-state districting in states with high levels of affected populations would result in voter suppression. Latino-related and immigrant-related populations would have their voting power reduced compared to populations not affected by the citizenship question.

Justin Levitt notes that the Census Bureau’s “single indefeasible obligation is to ensure, to the best of its ability, the absolute and inviolate integrity of the constitutionally mandated decennial count.” One might suggest that this line of thought should also apply to the Supreme Court.