The Census Bureau says that analyses have revealed an undercount of some demographic groups in the 2020 census.
The Census Bureau has a confession to make. A pair of analyses “revealed that the 2020 Census overcounted or undercounted various demographic groups”:
The results show that the 2020 Census undercounted the Black or African American population, the American Indian or Alaska Native population living on a reservation, the Hispanic or Latino population, and people who reported being of Some Other Race. On the other hand, the 2020 Census overcounted the Non-Hispanic White population and the Asian population. The Native Hawaiian or Other Pacific Islander population was neither overcounted nor undercounted according to the findings. Among age groups, the 2020 Census undercounted children 0 to 17 years old, particularly young children 0 to 4 years old. Young children are persistently undercounted in the decennial census.
Overall, the 2020 census ” incorrectly counted 18.8 million residents, double-counting some 5.2 million people, wrongly including another two million and missing others entirely, even as it came extremely close to reaching an accurate count of the overall population.” The census will never be 100 percent accurate, but 2020 presented a unique situation: the difficulty of running a nationwide count that relies on short-term employees and door-to-door contacts during a pandemic that severely restricted traditional tactics right when the 2020 census was supposed to get rolling in April 2020. There was a battery of other problems as well: litigation over a citizenship question, wildfires and hurricanes, debates over whether the Trump administration stopped the count prematurely after a late start (had it not done so, there would be even less time for states to redraw legislative districts and for courts to hear the inevitable redistricting lawsuits).
On top of that, state governments took dramatically different approaches to getting people counted. As even Elie Mystal of the Nation argues, that hurt Republicans when the announced shifts in population fell short of expected gains for red-leaning states in the House and the Electoral College:
In three states with booming Latinx populations — Arizona, Florida, and Texas — the Republican governors put few to zero dollars into publicizing the Census, even though undercounting reduced their states’ national political power. . . . Much has been made of New York’s “losing” a congressional seat by just 89 people. But people forget that New York was projected to lose two seats. Arizona, Florida, and Texas underperformed projections, and doing the Census correctly would probably place more political power in those purple states (and I’m being hopeful by calling Texas purple) and move it away from more solidly blue states like New York and Minnesota.
As Reid Epstein and Jennifer Medina of the New York Times detailed last spring:
Minnesota started its 2020 census outreach in 2015, bringing together local governments, foundations and businesses to spur participation. California allocated $187 million beginning in 2019 to get its people counted. And New York City alone spent $40 million on census advertising, texts and events. By contrast, Texas didn’t invest in a census program until last September, dedicating $15 million months after the count had begun. Florida’s governor, Ron DeSantis, said in 2019 that the state would not “have a role” in aiding the count, before ultimately agreeing to an unfunded committee in January 2020. And Arizona spent less than $1.5 million on census efforts in what remains one of the fastest-growing states.
This was, in retrospect, a big unforced error not only in how the Trump administration ran the 2020 census but also by typically savvy Republican governors such as DeSantis, Greg Abbott, and Doug Ducey, and it could haunt DeSantis or Abbott if one of them is running for president in 2024 or 2028 and finds Texas and Florida undercounted in the Electoral College. They simply did not play hardball for power in the way that Andrew Cuomo, Bill de Blasio, and Tim Walz did. That is on top of the bullet the Democrats dodged by the fact that the count predated big, further population shifts in late 2020 and into 2021 away from deep-blue, heavily locked-down states such as California and Illinois — shifts so dramatic that U-Haul reported that it ran out of trucks leaving California.
Mystal contends that Congress could either ignore the 2020 census (as it did, improperly, in 1920, and which Republicans would oppose bitterly) or call for a new one. Others have agitated for a skeptical look at the official numbers. From a purely partisan vantage point, that would almost certainly help Republicans nationally, and in many more states than it would hurt them.
But there are problems. Congress has statutorily authorized “a mid-decade census” each decade, which will be conducted in 2025, but has not authorized its use for redistricting or the Electoral College. At a minimum, a new statutory authorization would be required, and an outlay of something on the order of the $16 billion cost of the 2020 census would be a tough sell if Congress wanted a complete redo. Democrats are unlikely to be enthused about a move that would likely reduce their own power, and Republicans by nature are suspicious of messing with tradition in ways that could set precedents for future efforts to keep counting until Democrats get numbers they like. Ordinary Americans would also likely resent being asked to fill out the forms again so soon, and that would create its own problems in doing an accurate count.
There may also be constitutional hurdles. Article I, Section 2 states that “the actual Enumeration shall be made within three Years after the first Meeting of the Congress of the United States, and within every subsequent Term of ten Years, in such Manner as they shall by Law direct.” The first census was conducted in 1790, and before 2020, the most recent one was conducted in 2010. A completely new “actual Enumeration” in 2022 or later would not be “within [a] subsequent Term of ten Years” from the 2010 census.
There is a more arguable basis for Congress to direct the Census Bureau to adjust the 2020 numbers after the fact, using analytical techniques such as those used in the reviews announced today. The Supreme Court held in Department of Commerce v. United States House of Representatives (1999) that, for redistricting purposes, “actual Enumeration” means actual enumeration, not statistical sampling, but that was based on its reading of the current federal statutes, not the Constitution. In Utah v. Evans (2002), the Court upheld the use of some statistical techniques in the redistricting process as constitutional:
The Constitution’s text . . . uses a general word, “enumeration,” that refers to a counting process without describing the count’s methodological details. . . . Contemporaneous legal documents do not use the term “enumeration” in any specialized way. . . . However unaware the Framers might have been of specific future census needs . . . they fully understood that those future needs might differ dramatically from those of their own times. . . . Consequently, they did not write detailed census methodology into the Constitution.
Moreover, in Wisconsin v. City of New York (1996), rejecting a constitutional requirement of statistical sampling, the Court held that “the Constitution vests Congress with wide discretion over apportionment decisions and the conduct of the census” and that Congressionally directed decisions by the secretary of commerce “need bear only a reasonable relationship to the accomplishment of an actual enumeration of the population, keeping in mind the constitutional purpose of the census.” While the Court in Wisconsin disclaimed any explicit ruling at the time on the constitutionality of using statistical adjustments, the Court reiterated that broad, discretionary standard in Department of Commerce v. New York (2019), finding the use of a citizenship question to be constitutional. Thus, while the Court would likely find that the Biden administration lacks the power unilaterally to redo the census numbers, if Congress were to require a post hoc adjustment aimed at fixing the accuracy of the numbers, that would stand on much more defensible constitutional ground. If properly done, it could shift the balance of power further away from New York, Minnesota, California, and/or Illinois to Texas, Florida, and Arizona.
Still, given the formidable political obstacles, it seems likely that Republicans will be stuck with the consequences of the 2020 census figures until the 2032 election.
Correction: An earlier version of this article cited the New York Times article to state that the 2020 Census had undercounted the population by 18.8 million. The Times was in error. This article has been updated to quote the current version of what the Times reports.