Prime Only Members-Only Article
reporter's notebook

The DOJ Is Hanging Its Entire Census Case On One Paragraph

on May 10, 2017 in Washington, DC.
<> on May 10, 2017 in Washington, DC.
April 23, 2019 4:15 p.m.

Solicitor General Noel Francisco’s defense in front of the Supreme Court Tuesday of Secretary Wilbur Ross’ addition of a citizenship question to the census primarily came down to one paragraph buried deep in a Census Bureau analysis — one of several that Bureau experts put together advising Ross against including the question on decennial survey.

The oral arguments, as I previously wrote, were surprisingly technical, given the big-picture issues at play, which include the apportionment of political power nationwide, the adequate representation of immigrant communities and the scope of executive power.

But Francisco and his allies on the bench sought to focus his argument on one narrow issue: whether Ross acted unreasonably — or, in the language of the Administrative Procedure Act, arbitrarily or capriciously — in overriding career experts’ advice on the question. Three trial courts so far say that he did.

In four separate memos and several pages of other supporting documents and testimony, the Bureau’s top scientists warned Ross that adding the question would risk the accuracy of the survey’s total count and that they had a proposal that would produce just as accurate — if not more accurate — citizenship data without that risk.

That proposal involved the Census Bureau looking at its existing data, as well the records held by other agencies regarding citizenship, to produce the kind of  citizenship data the Justice Department was ostensibly seeking for voting rights enforcement in its request that Ross add the question. (Ross, internal records released in the case make clear, had in fact engineered the DOJ’s request.) In the Bureau’s analysis of the citizenship question’s performance on a smaller scale survey — the American Community Survey (ACS) — its experts found that about one-in-three noncitizens said falsely that they were citizens, meaning that looking at the legal records held by other agencies was more likely to produce an accurate citizen count.

In the last of those memos, issued on March 1, 2018, the Bureau analyzed its proposal against a new “hybrid” proposal offered by Ross that would have the Bureau use a combination of the existing records approach and looking at the responses it received in asking the citizenship question on the Census. The Bureau still concluded that its proposal (known as Alternative C) would produce more accurate data than Ross’s hybrid one (Alternative D) but hedged just a bit on just how much more accurate Alternative C’s model would be, because it could not “quantify the relative magnitude of the errors” of the modeling in each proposal at the time:

Francisco pointed to this paragraph repeatedly to defend Ross’ decision, even as the Bureau had explicitly and strongly made the assessment that its proposal was still the better option. Not only did Alternative C produce better citizenship data, it did so without the harm to the overall accuracy brought by adding the question, the Bureau said.

However, in Francisco’s telling, “what the Secretary concluded was, in the face of uncertainty, he’d rather go with the bird in a hand and ask the question at 98 percent accuracy than an unknown and untested statistical model.”

This line of argument clearly irked the liberal justices, who showed themselves to be well-versed in the technical aspects of the case.

Justice Sonia Sotomayor accused Francisco of mischaracterizing the Bureau’s use of term “comparative errors.”

“How do you take or pluck out of what they say in one sentence, if you’re the Secretary, and rely on that one sentence and ignore the wealth of statistics, graphs, testimony, proof, control studies of how — how these response rates came about and decide that that one sentence is enough?” Sotomayor said.

Justice Elena Kagan pointed out Francisco’s invocation of 98 percent accuracy was misleading because it was a relatively small segment of noncitizens the two proposals dealt with, and the comparison point of the Bureau’s model could have been 99.5 percent accuracy.

She also noted that this robust of a defense of Ross’ decision was not in Ross’ memo itself announcing the decision.

“The fact that SG lawyers can come up with 60 pages of explanation for a decision, that’s all post hoc rationalization,” she said.

However, Francisco’s argument gave the conservative justices the limb they needed to ignore the dicier elements of the case, which included a record that clearly showed that DOJ request was pretextual, in violation of the Administrative Procedure Act.

Justices John Roberts, Brett Kavanaugh and Samuel Alito all asked the challengers’ lawyers questions that expressed sympathy for the idea that Ross was simply going with the proposal that represented a “a known quantity,” in Alito’s words, of accuracy.

New York Solicitor General Barbara Underwood argued that only the Ross idea came with the risk to the total accuracy of the census, while ACLU lawyer Dale Ho told Alito he was oversimplifying what the Bureau had told Ross about its proposal.

“Justice Alito, respectfully, I think the Census Bureau said a little bit more than ‘trust us,'” Ho said, referencing the phrase (“trust us”) that Alito repeatedly ascribed to the Bureau. “What the Census Bureau said was we can develop a highly accurate model for this that’s going to be better than getting the question wrong one-third of the time.”

Latest Prime
Masthead Masthead
Founder & Editor-in-Chief:
Executive Editor:
Managing Editor:
Senior Editor:
Special Projects Editor:
Editor at Large:
General Counsel:
Head of Product:
Director of Technology:
Associate Publisher:
Front-End Developer:
Senior Designer: