There Is More At Stake In The Census’s Citizenship Question Than Response Rates


One way or another, the Trump administration’s decision to add a citizenship question to the census has the potential to change the math our democracy is built on. That new math could shift political representation towards eligible voters and away from the country’s overall population — even if more non-citizens respond to the census than in years past.

By now you’ve likely heard concerns that adding the question could lower response rates among non-citizens during the 2020 population count. As my colleague Anna Maria Barry-Jester’s reporting suggests, this could happen due to concerns about privacy and possible deportations.

If response rates decline, states with larger proportions of non-citizens could lose political power. That’s because Congressional seats are doled out to states based on their total populations — including non-citizens, children and prisoners. Estimates suggest this would also mean a slight shift in power to Republican-leaning states. We won’t know whether this shift will occur until we see response rates. And if the changes are small, they may not be enough to affect congressional seat apportionment.

A lot of coverage of the citizenship question has focused on this dynamic. And it clearly matters for both federal funding and congressional apportionment, but there’s more to the story. “I think that’s probably not the most important potential effect (of the citizenship question),” said Rick Pildes, a constitutional law professor at New York University School of Law.

The addition of the question plays into an ongoing legal debate over how political districts should be drawn within states. Should we base our maps on the number of people or the number of citizens? Because once states have the data for the latter, there’s a good chance someone will use it.

As things stand today, states use census data to draw congressional and state legislative districts with approximately equal total populations, including people who are ineligible to vote. But the Supreme Court has not said whether that’s necessarily how it has to happen.

In the 2016 U.S. Supreme Court case Evenwel v. Abbott, plaintiffs from Texas challenged the way states draw districts, relying on the one person, one vote principle. They wanted states to draw districts with equal voter-eligible populations, instead of total populations. The former would exclude non-citizens and children. The argument was that voters living in districts with large non-citizen populations are given undue political influence.

For example, Sue Evenwel, the lead plaintiff in the case, lived in a state Senate district with approximately 584,000 citizens eligible to vote, while a neighboring district had only about 372,000 eligible voters. Both districts were about equal in total population.

It’s basically a fight over what the denominator of our democracy should be, and therefore what the goal is: electoral or representational equality. Should a representative be accountable to the same number of voters or the same number of total people?

The court unanimously rejected Evenwel’s argument that states must redistrict based on “electoral equality.” In her majority opinion, Justice Ruth Bader Ginsburg wrote, “Nonvoters have an important stake in many policy debates… By ensuring that each representative is subject to requests and suggestions from the same number of constituents, total population apportionment promotes equitable and effective representation.”

But while the court declined to force states to draw districts according to the voter-eligible population, it did not say whether states could if they wanted to. It saved that question for another day, and that is where the citizenship question comes into play.

“The Supreme Court was mindful in deciding the Evenwel case that the data collected by the Census Department may be insufficient to accomplish citizen-population redistricting,” said Edward Blum, the director of the Project on Fair Representation, a conservative advocacy group that represented Evenwel. “Therefore this decision to add this [citizenship] question should alleviate the court’s concern should this [legal] question ever return to the high court.”

Blum isn’t the only one thinking along these lines:

If adding the citizenship question is Step 1 in this question returning to the high court, Step 2 would be a state or municipality deciding to redistrict according to voter-eligible population during the next round of redistricting, in 2021. Ilya Shapiro of the Cato Institute views that as a distinct possibility. “I could see maybe Alabama or municipalities there or elsewhere deciding to redraw on those lines and starting up a new case that way,” he said.

Justice Clarence Thomas posited that such a move would be legal in his concurring opinion in the Evenwel case. “The Constitution does not prescribe any one basis for apportionment within states,” he wrote. “It instead leaves states significant leeway in apportioning their own districts to equalize total population, to equalize eligible voters or to promote any other principle consistent with a republican form of government.” Of course, there are eight other justices who would have to weigh in.

In the first scenario we discussed — lower response rates among non-citizens — political representation would shift in favor of states with larger proportions of eligible voters. Likewise, a court decision along the lines of Thomas’s opinion would also pave the way to shift political representation to regions within states that have larger proportions of eligible voters. American Community Survey data shows that those areas tend to be rural. According to the ACS data, rural counties have far smaller proportions of foreign-born people than urban counties. The measure is imperfect, because foreigners can become U.S. citizens, but it gives us some indication of the trend. If representation within states were to shift toward eligible voters — and therefore rural areas, it would also likely benefit Republicans.

So, say all the concerns about response rates are for naught, and even more non-citizens respond to this coming census. The new citizenship data could still lay the groundwork for other changes in how we are represented. Or to put it more directly, who is represented.