After a setback at the Supreme Court, President Donald J. Trump has backed off efforts to ask every household in the United States about the citizenship of its members. But his plans to seek the same information by other means could still have profound implications for the distribution of legislative seats and vast federal resources.
The Supreme Court ruled in June against the Department of Commerce, which sought to ask on census forms about every household member, “Is this person a citizen of the United States?” The administration claimed that the question would help it enforce the Voting Rights Act.
States and cities, as well as numerous advocacy groups, argued that the question’s inclusion would compromise the accuracy of the population count, violating the constitutional requirements of an “actual Enumeration” and equal protection. They said immigrant and minority households would be deterred from completing the census, especially given the hostile political environment.
The assertion is backed by the Census Bureau’s own research, which suggests that some nine million people—out of a total population of roughly 330 million—would not return a census form that asked about citizenship status. Many of those who’d go uncounted are citizens or legal residents, and urban areas would be disproportionately affected.
Chief Justice John Roberts did not rule on whether the citizenship question would be constitutionally permissible; instead, he ruled on procedural grounds, calling the administration’s stated rationale “contrived.”
This month, President Trump said he would press ahead on acquiring citizenship data. He signed an executive order directing the Social Security Administration, the Department of Homeland Security, and other federal agencies to share such data with the Census Bureau.
What’s at stake?
The 2020 census—a $15.6 billion undertaking that the Government Accountability Office has deemed “high risk”—will determine the distribution of legislative seats and vast federal resources for a decade to come.
The House of Representatives’ 435 seats will be reapportioned among the states based on population changes registered between the 2010 census and the 2020 one. If there is a significant undercount, states with large immigrant and minority populations will be underrepresented in Congress, as well as in the Electoral College. California would be poised to lose at least one congressional seat; Texas, Arizona, and Florida might as well.
Since states redraw legislative districts to contain roughly equal numbers of people after every census, a flawed count will leave immigrant and minority neighborhoods underrepresented as well.
Also at stake is the distribution of nearly $900 billion per year in federal spending on education, health care, housing, and other social services allocated according to census data.
What happens next?
If the Trump administration successfully pairs citizenship data with census returns, it could set in motion far more transformative changes in the way legislative power is distributed.
States currently draw legislative districts on the basis of the total number of residents, with each state legislator or member of Congress in a state representing roughly the same number of people. But with information about citizenship, states would then have the data they would need to draw their districts based on the number of eligible voters—citizens aged eighteen or above—instead. Trump said as much as he signed the executive order.
In states they control, Republicans could then seek electoral advantage by drawing “wider, whiter districts beyond urban boundaries,” writes Brookings Institution demographer William Frey. Youth and immigrants would find themselves without representation.
If states opt for this method, they will no doubt come up against legal challenges. On a technical level, courts will have to decide whether the federal citizenship data is reliable and thorough enough to be the basis of districting. The Supreme Court will likely find itself grappling with a more profound issue as well: the meaning of “one person, one vote.”