Opinion - Trump’s proposed snap Census won’t get him the seats he wants

Related video: FIGHT FOR POWER: Explaining The Redistricting Wars

President Trump recently announced that he had directed the Department of Commerce to conduct an unprecedented, mid-decade census of the U.S. population. Like his order for the 2020 Census that was rejected by a federal court, Trump has ordered this new census to exclude undocumented immigrants from the population count.

This order is the latest move to reshape power in the House of Representatives ahead of next year’s midterm elections, alongside ongoing national turmoil in legislative redistricting.

As academic experts in elections, redistricting and the U.S. Census, we offer three conclusions on this order here.

First, President Trump’s demand for a new Census that excludes undocumented immigrants is not just immoral and regressive, but unquestionably unconstitutional in at least two ways.

The text of Article One is clear (and echoed by Title 13 of the U.S. Code): the Census is conducted every 10 years, not whenever it is politically convenient. The last decennial Census was conducted in 2020, and preparations are well underway for the 2030 Census.

Further, the 14th Amendment plainly states that apportionment in the House of Representatives is based on the “whole number of persons in each State.”

Proposing restrictions on this bedrock democratic principle is instead reminiscent of earlier restrictive eras of American voting rights, where groups like enslaved people and Native Americans were not fully enumerated for apportionment.

Second, conducting a surprise Census is infeasible and would require compromising data quality. Enumerating the U.S. population is one of the largest, most sophisticated and most difficult statistical efforts in the world.

Conducting a census involves hiring hundreds of thousands of temporary employees, who cannot be trained on short notice. Training these “enumerators” is necessary to ensure statistical accuracy because less than seven in 10 people respond to the Census form, and instead must be reached in person.

Further, many of the hardest-to-reach households are in rural areas, which often lean Republican. Sacrificing data quality through a rushed census would likely undercount these rural populations.

Third, likely changes in apportionment from a new Census do not clearly benefit Republicans in the way that supporters might hope. The apportionment change from 2010 to 2020 created additional seats in Republican-controlled states like Texas and Florida, while Democrats lost seats in states like New York and California.

How might a new round of reapportionment without undocumented American residents play out? We lack reliable public data on the number of documented and undocumented immigrants across the United States. But, even if we consider the extreme case that all non-citizens would go uncounted, a new apportionment would likely have only limited partisan effects.

We ran a simulated apportionment scenario under the same method used to apportion the House of Representatives, after estimating new population counts with non-citizens removed based on data from the American Community Survey. We find that three states would lose seats: California (minus-4), Texas (minus-2), and New York (minus-1).

In contrast, Missouri, Ohio, Pennsylvania, Tennessee and West Virginia would each gain a seat. While many of these states lean Republican, these states’ congressional maps already have small biases toward Republicans. Apportionment changes may be unlikely to push them further in Republicans’ favor.

Other states, like Idaho and Michigan, are likely to each gain a seat, but their redistricting is controlled by nonpartisan commissions. States with commissions would find it difficult to introduce a new partisan gerrymander.

Further, a new Census would likely diminish political representation for Republican areas.

For example, consider likely changes in Texas. While many (documented and undocumented) non-citizens live in urban counties like Houston’s Harris County, large non-citizen populations often live in rural areas that lean conservative. The Migration Policy Institute estimates that Hidalgo and Cameron counties — both southern border counties won by Trump in 2024 after voting for Biden in 2020 — have two of the largest undocumented populations in the country.

Undocumented workers play key roles in the local economies of these areas. One recent survey of Hidalgo farmworkers by the National Center for Farmworker Health estimated that 80 percent were undocumented.

Beyond large population changes in some counties, changes to population counting can have large relative implications too. For example, a new Census could lessen the political representation of small, rural counties.

Again, consider Texas: Smaller counties would lose relative political weight from a new Census due to proportionally large undocumented populations in the north (such as Dallam and Ochiltree counties), west (Hudspeth County) and south (Starr County), all of which were carried by Trump in 2024.

Comparatively, the relative political power of suburban areas, where fewer undocumented people tend to live compared to urban and rural areas, would increase. If this choice to count only citizens spreads to impact how funds are distributed, these areas, alongside urban areas — many of which are conservative, rural or both — are likely to suffer.

Christopher Kenny is a postdoctoral research associate at Princeton University’s Data-Driven Social Sciences. Tyler Simko is an assistant professor of Political Science at the University of Michigan.

Copyright 2025 Nexstar Media, Inc. All rights reserved. This material may not be published, broadcast, rewritten, or redistributed.

For the latest news, weather, sports, and streaming video, head to The Hill.