Jonathan P. Baird

One of the most consequential cases coming before the U.S. Supreme Court this spring term is the Commerce Department’s effort to add a citizenship question to the 2020 census.

When you clear away the fog, the case is really about the Trump Administration’s effort to depress response levels by immigrant communities, particularly Latinos, to the census. They will, of course, say the citizenship question was not politically motivated.

In the trial on the case held in federal court in New York, the Census Bureau’s own experts estimated that 6.5 million people were unlikely to respond to the census citizenship question.

Dale Ho, the ACLU lawyer who litigated this case, has said that 6.5 million would be equal to our 18th largest state. If these people lived in one state,  they would be represented by nine seats in Congress and eleven electoral votes.

Immigrant communities rightly fear that the Trump Administration will use census data to track people down. While it is illegal for the Census Bureau to release data on specific households, the unprecedented hostility to immigrants demonstrated by the Trump Administration has led many immigrant households to worry that their disclosures could be handed over to ICE.

Unfortunately there is precedent: during the Japanese-American internment, the government used census data to identify people. That is why, among others, some Japanese-Americans have been speaking out against the inclusion of a citizenship question in the census. In 1943, the Census Bureau gave the Secret Service a list of name, addresses, citizenship status and other personal information of people of Japanese ancestry in the Washington DC area. Japanese-American citizens were then rounded up.

If there were a significant undercount of U.S. residents, the census would have an enormous effect. The census is used to determine congressional apportionment for the next decade. California, New York, Arizona, and Texas could lose seats in the U.S. House of Representatives, as well as lose federal dollars.

Census data could influence the allocation of more than $800 billion in federal government resources to states, localities and families every year. Census data helps state and local officials identify both current and future needs for the communities they serve.

It is not hyperbole to recognize that the stakes are immense.

The U.S. Constitution requires a census every ten years of all persons living in the country for the purpose of apportioning states in the U.S. House of Representatives. The Constitution explicitly requires an “actual enumeration” of “all persons”. That means the actual number of persons who reside in each state.

In adopting the Fourteenth Amendment, Congress rejected proposals to allocate seats in the House based on voter-eligible population, rather than total population.

Congress has the authority to conduct the census and Congress has delegated the task to the Secretary of Commerce. That delegation is subject to further legal requirements, including compliance with the Census Act and the Administrative Procedures Act or APA. The APA does require that the Secretary be truthful about the reasons for his or her inclusions in the census.

In January of this year, after a two week trial, Federal Court Judge Jesse M. Furman blocked the Commerce Department’s effort to add a citizenship question to the 2020 census. In a 277 page decision. Judge Furman found that the government and Commerce Secretary Wilbur Ross had acted improperly. The Court found the inclusion of the citizenship question was arbitrary and made in spite of overwhelming evidence the question would depress census response.

Judge Furman noted that the Census Bureau itself had strenuously objected to Secretary Ross’s plan because they believed that adding the question would harm the quality of census data and would increase costs significantly.

In his decision, Judge Furman specifically ruled that Ross had violated the APA, which governs how agencies exercise their powers. The APA outlines the requirement to be truthful in the reasons behind a rule. Here is how Judge Furman summarized Ross’s role:

“[Ross] failed to consider several important aspects of the problem, alternately ignored, cherry-picked, or badly misconstrued the evidence in the record before him; acted irrationally both in light of that evidence and his own stated decisional criteria; and failed to justify significant departures from past policies and practices.”

The Court ruled that Ross’s purported reason for adding the citizenship question – to enforce the Voting Rights Act – was “pretextual”, a made-up excuse. In the record there was much evidence of bad faith by Ross. He had denied to Congress in testimony that he had had conversations with the White House about the citizenship census question. That was a lie.

The records show Ross had spoken to Steve Bannon, White House advisor at the time, and Kris Kobach, Kansas Secretary of State. Kobach has been a player in schemes around voter suppression.

Emails showed Ross had been asking around for months about adding a citizenship question. He pushed for it and then added a rationale.

Unfortunately, the plaintiffs in the case never had the opportunity to depose Secretary Ross as the government successfully fought off the deposition.

The case is now heading to the Supreme Court where the Trump Administration hopes it has five friends. Because census forms need to be printed in summer or fall at the latest, the case is on a fast track, skipping the normal appellate procedure which would typically have required a circuit court decision after the trial court. Oral argument is scheduled for late April and a written opinion by June.

There is a risk of a failed census in 2020. Or if not a failed census, an inaccurate and unreliable one, with a variety of harmful consequences. Accuracy goes hand in hand with fairness.

Congress did not help things when it decreed in 2014 that the 2020 census should cost no more than the 2010 count, without adjusting for inflation. The General Accountability Office already designated the 2020 census as one of those few federal programs at high risk of failure.

An undercount in the census would likely harm immigrant (particularly Latino) populations as they would face reduced funding and less political representation. Already marginalized populations would become further under-represented and under-resourced.

Under the cloud surrounding the 2020 Census, it is easy to imagine people freaking out at the sight of census correspondence. The arrival of census takers at the front door would also not be a welcome sight for many.

There will be no way to fix a census undercount. Whatever the Supreme Court ultimately decides, the accuracy of the census remains a critical question that will determine its credibility and the degree to which there will be political and economic fairness over the next decade across the U.S..