Immigration activists rally April 23 outside the Supreme Court in Washington as the justices hear arguments about the Trump administration's plan to ask about citizenship on the 2020 Census. (J. Scott Applewhite/AP)

To hear President Trump’s administration tell it, the 2020 Census needs to include a question about residents’ citizenship because it would aid enforcement of the Voting Rights Act. Few experts think that such a question is needed for that purpose, mind you, and few people think that’s the administration’s motivation. Nonetheless, the Supreme Court appears to be poised to determine that the question can remain on next year’s national survey, accepting the administration’s rationale.

That likelihood will probably not be affected by the discovery of remarkable documents that kneecap the White House’s argument. On Thursday, we learned that a Republican consultant who had argued for such a question had done so explicitly because he understood that it would help his party politically.

The Washington Post’s Tara Bahrampour explained the discovery:

The evidence, found in the files of the prominent Republican redistricting strategist Thomas Hofeller after his death last August, reveal that Hofeller “played a significant role in orchestrating the addition of the citizenship question to the 2020 Decennial Census in order to create a structural electoral advantage for, in his own words, ‘Republicans and Non-Hispanic Whites,’” and that Trump administration officials purposely obscured Hofeller’s role in court proceedings, lawyers for plaintiffs challenging the question wrote in a letter to U.S. District Judge Jesse Furman.

Hofeller had been evaluating Texas’s legislative districts and determined that narrowing the population that was considered for the purposes of districting to voting-age citizens would aid Republicans in the Lone Star State. In other words, instead of drawing district boundaries based on the population of the state, he advocated using a narrower pool of individuals that would effectively force Democratic districts to be drawn larger to include enough residents — reducing the number of districts in heavily Democratic areas.

To effectively implement that plan, though, he needed better data about the number of citizens. Ergo: a question on the census.

The question itself probably will have much the same effect on Democratic districts, some experts say. It is likely to dampen responses from largely Hispanic and immigrant communities, places that often vote more heavily Democratic.

Limiting the considered population pool to adult citizens would sharply reduce the number of people included in federal and state districting efforts. On average, about a quarter of the residents of each state would no longer be included in the population used for such analysis. Census Bureau citizen estimates suggest that the percentage of residents excluded would range from a 35.8 percent drop in Texas to a 20.3 percent drop in Maine.


(Philip Bump/The Washington Post)

Why such a large difference? Because Texas has a much larger density of Hispanic residents. About 57 percent of those who would be excluded from Texas’s count are Hispanic, both because the Hispanic population in the United States tends to be younger than other groups and because the state has many Hispanic noncitizens. California would experience a similar drop.


(Philip Bump/The Washington Post)

The percentage of non-Hispanic whites that would be excluded nationally is 20.2 percent. The percentage of Hispanics? More than half, 51.1 percent.

That lopsidedness is one reason the political effect in Texas would be uneven.

Hispanics vote more consistently Democratic than do whites. Most of the congressional districts nationally in which a shift to counting only adult citizens would yield the largest drop are ones that Hillary Clinton won in 2016. Generally, that’s a function of the size of the Hispanic population in the district.

Of Texas’s 36 districts, the 10 that would be most affected by a shift to counting only adult citizens backed Clinton. The 15 that would be affected the least by that change voted for Trump.


(Philip Bump/The Washington Post)

We can look at the political effects of this in another way. If this concept were expanded nationally to the extent that electoral votes were allocated solely on the basis of the adult citizen population in each state, the effect would be a decline of four votes in blue states — driven by a decline of five votes in California — and a gain of four in red states. That’s a swing of eight votes to the GOP.


(Philip Bump/The Washington Post)

Shifts like these were apparently the point of Hofeller’s work. He recognized that by identifying the number of citizens across the country, he and his party could advocate for a rules change that benefited Republicans politically.

For understandable reasons, that’s not the message the Trump administration wanted to promote.