An accurate decennial census matters because the nation’s population count directs federal funding for hundreds of programs. Where people are and who people are directly informs decisions on the distribution of federal funding. The tally also determines apportionment in the U.S. House of Representatives.
Simply put, states that gain population relative to other states are likely to gain representatives in the House. The number of House seats assigned to a state also determines its number of presidential electors, members of the Electoral College who ultimately vote for president every four years. That means the number of House seats in each state could affect the outcomes of the 2024 and 2028 presidential elections, if those races are tight. If there is a state-level population undercount, it’s not just federal funding at stake — it’s democratic representation as well.
House reapportionment and the status of people living in the U.S. without legal permission are central to Trump v. New York, which the U.S. Supreme Court will hear on Nov. 30. The case arose after President Donald Trump in July directed, via memorandum, that the Census Bureau subtract immigrants living in the U.S. without legal permission from state population counts used to determine House reapportionment.
Citizens and noncitizens have been included in House reapportionment calculations since the U.S. government conducted the first national census in 1790. In his memorandum, Trump reasoned: “States adopting policies that encourage illegal aliens to enter this country and that hobble Federal efforts to enforce the immigration laws passed by the Congress should not be rewarded with greater representation in the House of Representatives.”
It’s important to note that states cannot redraw their congressional district boundaries — through a process called redistricting — until the apportionment process is finished. Here’s what else you need to know when lawyers make their respective cases before the Supreme Court on Nov. 30 in Trump v. New York.
A brief, recent history of the citizenship question
In December 2017, the Department of Justice requested that the Census Bureau, housed within the Department of Commerce, ask every person living in the U.S. for their citizenship status on the 2020 census. Justice Department counsel argued the citizenship question was needed to aid the department’s enforcement of the Voting Rights Act.
DOJ attorney Arthur Gary, a career official who typically dealt with budget and procurement issues, wrote that “the Department needs a reliable calculation of the citizen voting-age population in localities where voting rights violations are alleged or suspected.” Emails that ProPublica obtained revealed the letter was drafted by John Gore, acting head of the DOJ’s Civil Rights Division at the time and a Trump political appointee. In March 2018, Commerce Department Secretary Wilbur Ross announced the 2020 census would indeed include a question asking respondents whether they were U.S. citizens.
But the citizenship question never made it onto census questionnaires. In June 2019, the U.S. Supreme Court rejected the DOJ’s rationale, spelling the end of the citizenship question. That the citizenship question was needed to enforce the Voting Rights Act “seems to have been contrived,” according to Chief Justice John Roberts, writing for the majority.
“In the Secretary’s telling, Commerce was simply acting on a routine data request from another agency,” Roberts writes. “Yet the materials before us indicate that Commerce went to great lengths to elicit the request from DOJ (or any other willing agency).”
In a mid-July workaround, the Trump administration directed, via executive order, that executive branch departments and agencies provide the Commerce Department with any requested data that would help the Census Bureau determine the number of citizens and non-citizens living in the U.S. Ross, in his March 2018 memo, had directed the Census Bureau to use federal and state administrative records to try to fill in the citizenship blanks.
Executive memoranda and executive orders are effectively laws for the executive branch. They’re largely similar, and there’s no official definition of either. In the weeks and months before Trump’s July executive order, Iowa, Nebraska, South Carolina and South Dakota agreed to provide the Census Bureau with driver’s license and state identification card data, according to reporting from NPR’s Hansi Lo Wang. Other states — including New Hampshire, Pennsylvania and Maine — declined to provide or offer driver’s license data.
In August, the Census Bureau shortened counting by a month, “to accelerate the completion of data collection and apportionment counts by our statutory deadline of December 31, 2020, as required by law and directed by the Secretary of Commerce,” according to Census Bureau Director Steven Dillingham. Some census workers were left concerned that there was not enough time to achieve an accurate headcount. Two census workers in Massachusetts and Indiana this week told the Associated Press that they entered false information under intense pressure from supervisors because of the tight deadline.
What’s happening with the census and the Supreme Court on Nov. 30?
Federal courts in Maryland, New York and California have already ruled it would be unlawful for the U.S. government to exclude immigrants living here without legal permission from population counts used to divvy up House seats. The Trump administration appealed the New York court’s decision, entered Sept. 10. That’s the case that will soon come before the Supreme Court.
“The Presidential Memorandum violates the statutes governing the census and apportionment in two clear respects,” the three U.S. district court judges from New York write in their ruling. First, federal law requires that the Commerce Department secretary deliver a single set of numbers — the total population for each state — to the president, who, in turn, “is required to use the same set of numbers in connection with apportionment,” the judges write.
The New York judges note that Trump directed Ross to offer up not one but two sets of numbers: The census counts, including immigrants living here without legal permission, and another count of individuals living illegally in each state, tabulated based on administrative records from other federal and state agencies.
“By directing the Secretary to provide two sets of numbers, one derived from the decennial census and one not, and announcing that it is the policy of the United States to use the latter in connection with apportionment, the Presidential Memorandum deviates from, and thus violates, the statutory scheme,” the judges write.
The judges further observe that the memorandum “violates the statute governing apportionment because, so long as they reside in the United States, illegal aliens qualify as ‘persons in’ a ‘State’ as Congress used those words.” The Supreme Court will have the final say on whether the president overstepped his legal bounds in issuing the memorandum and, by extension, whether people living in the U.S. without legal permission will be included in the state tallies used for House reapportionment.
“There’s two issues here,” census historian Dan Bouk explained during an Oct. 29 media briefing co-hosted by the Beeck Center for Social Impact and Innovation at Georgetown University, Journalist’s Resource and Reveal. “One is constitutional and one is statutory. The Constitution, through the Fourteenth Amendment, declares that it’s the ‘whole number of persons’ that will be counted. And so that’s the basis on which the Supreme Court will be deciding if it’s a constitutional claim. And the statutory claim is that the current apportionment statute also uses that same language.”
It’s unclear when the Supreme Court might rule, though the case has been fast-tracked — state population totals are due to the president at the end of December.
What’s at stake?
Whether a state’s population is fully represented in the House, which could affect how much sway voters in each state will have in picking the next two presidents. Trump, in his July memorandum, described the stakes for one large state:
“Current estimates suggest that one State is home to more than 2.2 million illegal aliens, constituting more than 6 percent of the State’s entire population. Including these illegal aliens in the population of the State for the purpose of apportionment could result in the allocation of two or three more congressional seats than would otherwise be allocated.”
The New York judges in their September decision surmise that Trump was referring to California.
How does reapportionment work?
House seats are assigned based on a multistep but straightforward series of equations. Since 1790, the U.S. government has altered its reapportionment formula five times. The current formula, called the Method of Equal Proportions, has been used since 1941.
The Permanent Apportionment Act of 1929 fixes the number of House seats at 435. Each state gets at least one representative, meaning the first 50 seats are automatically apportioned. The remaining 385 seats are allocated by the current formula.
The formula produces state priority rankings based in part on population totals. The top-ranked state for the 51st seat gets that seat. The top-ranked state for the 52nd seat gets that seat, and so on until all 435 seats are allocated. The bottom line: A state that has gained population since 2010 has better odds of gaining seats in the House than a state with a stagnant or shrunken population.
What are key dates tied to the House reapportionment process?
Oct. 15: The end of Census Bureau field data gathering and the last day for individuals to postmark their census questionnaires or file them online.
Nov. 30: Supreme Court oral arguments on whether people living in the U.S. without legal permission should be included in House reapportionment.
Oct. 15 to Dec. 31: Timeframe for Census Bureau data processing. This is three months shorter than the average amount of time the Census Bureau has had for processing since the 1990 census, according to a report from the National Conference on Citizenship, a congressionally chartered nonprofit focusing on civic engagement.
Dec. 15: Expected date the Census Bureau will release its “demographic analysis,” the first look at the nation’s new population estimates, based on birth and death records, migration numbers and Medicare enrollment data.
Dec. 31: The statutory deadline for the secretary of commerce to deliver to the president the state population counts used for House reapportionment. During the first week the newly elected Congress meets in the New Year, the president “shall transmit to the Congress a statement showing the whole number of persons in each State.” Within 15 calendar days after that, the clerk of the House must send the allocation of House seats to state officials, according to a 2019 report from the Congressional Research Service.
April 1: The statutory deadline for the Census Bureau to deliver detailed census information to state officials in charge of redistricting.
These dates are subject to change, even though many are just weeks away.
“On the timing front, Congress could still weigh in, which may make news,” Denice Ross, a fellow at the Beeck Center, said during the Oct. 29 media briefing. “A majority of both houses [of Congress] supports moving the deadline back to allow for sufficient time for processing the data.”
Even if these dates come and go that doesn’t mean the assignment of House seats is set in stone for the next decade. If the Supreme Court allows Trump to exclude some number of people living in the U.S. illegally from the apportionment calculation, Congress could still pass and the president could still sign legislation specifying that those individuals be included in reapportionment, according to Bouk, who works at the nonprofit research organization Data & Society.
If Congress changed the apportionment process after the decennial census, it wouldn’t be the first time. Congress codified the Method of Equal Proportions in legislation passed in November 1941. That apportionment formula was in force for the 78th Congress, seated in January 1943 — roughly three years after the 1940 census was enumerated.