On Tuesday, the Supreme Court will hear the Trump administration's appeal in the case over its nefarious push to add a citizenship question to the 2020 census, the outcome of which could have profoundly disastrous effects for democracy and fair political representation. If the Republican majority on the court sides with Trump, experts fear that the question would have a chilling effect that intimidates millions of people in immigrant communities into not participating. That could in turn turbocharge a new wave of hyperpartisan Republican gerrymandering nationwide, since the census is the bedrock foundation of redistricting.
The Constitution mandates the counting of everyone in the U.S. without regard to their legal status, and a question on citizenship hasn't been included on the decennial census since 1950. Multiple courts have ruled that Trump's attempt to add the question for 2020 violates both federal laws and the Constitution. The lower courts have consistently held that Commerce Secretary Wilbur Ross' claim that the citizenship question was needed for the Justice Department to enforce the Voting Rights Act was a bogus pretext that masked the administration's true motive, which was likely to intimidate immigrant communities.
Adding a citizenship question on the census could lead to a dramatic drop in participation among not only undocumented immigrants but also legal residents and naturalized citizens—all out of fear that Trump's administration could illegally use the data to expand his brutal campaign of human rights abuses directed toward immigrants and asylum-seekers that has shocked the nation over the past year. Indeed, a new review of scholarly analysis in the Washington Post estimates that the question could deter 6 million Latinos from participating, roughly one out of every eight Latinos in America.
Undercounting millions of people by itself would mean their communities would lose out on their rightful political representation in redistricting and they would lose federal funding. But the citizenship question provides a double blow to fair representation in redistricting because it opens the door to Republican efforts to draw districts by only counting eligible voters rather than all citizens. The Supreme Court hasn't foreclosed the possibility of permitting this, even though it has been a longstanding norm to count every person in a district for redistricting, since elected officials represent voters and nonvoters alike.
The citizenship question would be a crippling blow to Democrats and voters of color in redistricting, further entrenching Republican minority rule even when Democrats receive more votes overall. Below, we'll look at how this would work in perhaps the most critical battleground over citizenship and changing demographics: Texas.
In 2018, Texas finally emerged as a battleground state: Even though Democrat Beto O'Rourke lost the Senate race to GOP incumbent Ted Cruz, his 2.6% margin was the closest Democrats have come to winning a Senate seat in the Lone Star State in 30 years. What’s more, voters elected a new wave of Democrats and candidates of color in former GOP strongholds across the state. Given that voters of color and college-educated white voters are rapidly growing as a share of Texas' population just as Republicans do ever worse with these demographics, the state should continue to trend toward Democrats—if elections are fair.
However, while demographics may drive statewide election results, those who draw the district lines have the power to turn future Democratic popular vote victories into Republican majorities at the congressional and legislative level. Texas already is home to some of the most extreme Republican gerrymanders anywhere, but as shown below, even nonpartisan redistricting could end up enforcing a Republican advantage if Trump gets his way and adds a citizenship question to the census.
Daily Kos Elections has drawn a hypothetical nonpartisan congressional map, illustrated below, using traditional nonpartisan redistricting criteria such as following the Voting Rights Act, keeping communities and local governments whole, and aiming for compactness. Unlike the existing GOP gerrymander, this map would be fair in terms of both race and partisanship: Hillary Clinton and O'Rourke both won 17 of 36 districts while losing by single digits statewide, and Latino voters would have been able to elect their preferred candidates in three additional districts in 2018 (redrawn Districts 23, 33, and 35) for a more proportional share of the seats.
However, that fairness takes a serious hit once you throw in the citizenship question. While the census itself hasn't asked about citizenship since 1950, the Census Bureau has included the question on its voluntary American Community Survey, which is a large statistical sampling of Americans that is far more detailed than the decennial census.
By calculating the 2008-2012 American Community Survey numbers on adult citizens (CVAP) in each of our hypothetical districts and comparing them with the 2010 decennial census' total population, as shown on the graph below, you can see a troubling trend emerge: Democratic districts typically have lower eligible voter populations because they include more immigrants and citizen children who are ineligible to vote. On average, only 58% of the population in the 17 Democratic districts were adult citizens, but a higher 68% were in the 19 Republican districts.
What that disparity means is that if districts were drawn solely based on the eligible voter population of adult citizens—a population that is much whiter than the total population—Democratic districts in communities of color and major cities would have to expand by packing in more voters as rural Republican districts shrink. The likely outcome would be that the remaining majority of Republican districts under this map would have become even more conservative as the Democratic districts expand further into the blue-trending suburbs of major cities such as Houston, Dallas, and Austin.
Worse, this map is only a hypothetical: A bona fide Republican gerrymander, which the GOP is poised to draw after 2020, could make the existing districts even more extreme than they already are. And of course, this exercise doesn't even begin to account for the impact of an actual undercount compounded with redistricting based on adult citizens, which would disproportionately affect Democratic-leaning areas, causing those districts to have to expand and further push Republican districts into more conservative turf, solidifying a GOP majority.
This redistricting impact wouldn't just be limited to Texas, and the citizenship question would have ripple effects across the country wherever there are large immigrant populations, other hard-to-count communities, or even simply where Democratic districts have disproportionately more citizen children than Republican ones do. Adding the citizenship question to the census would be a steroid injection for maps that give white Republicans an extreme undue advantage, and it's a monumental threat to democracy and fair elections.