Last March, Secretary of Commerce Wilbur Ross forwarded to Congress a list of proposed questions to be included in the United States 2020 census, including an item asking respondents about their citizenship status. The item was inserted into the census questionnaire in response to a letter from the Department of Justice asserting that citizenship information was needed to facilitate enforcement of the Voting Rights Act’s districting provisions. The Census Bureau also says on its website that information about the distribution of citizens and noncitizens is of interest to policymakers, researchers, and advocacy groups. In addition to the citizenship question, the census questionnaire will also include items regarding age, race, gender, ethnicity, marital status, and homeownership.
A question about citizenship status was included in one form or another in every decennial census from 1890 through 1950. In subsequent censuses from 1960 through 2000, that item appeared in the so-called long-form questionnaire, mailed to a sample of 15 percent or 20 percent of the population; it was not included in the short-form questionnaire administered to the other 80 percent to 85 percent of households. The Census Bureau continues to include the citizenship question in its more detailed American Community Survey (ACS), administered to samples of the population every month, with cumulative results summarized annually. With the inauguration of the ACS in 2005, the Census Bureau decided that it could dispense with the long-form questionnaire in the 2010 census while administering the streamlined short form to the entire population. When the long form was discarded, the citizenship question disappeared along with it, and thus was not included in the 2010 census.
Secretary Ross’s announcement provoked lawsuits from 17 states (including New York and California), several cities, and a coalition of civil liberties organizations that want the question removed, claiming that it will discourage noncitizens from participating in the census. To the extent that this happens, the plaintiffs say, the census will produce an undercount of the population in states and cities with large concentrations of immigrants, leading to losses in congressional and state legislative seats in those jurisdictions and parallel losses in federal dollars allocated on the basis of the population count in the census.
New York attorney general Barbara Underwood summarized the plaintiffs’ case: “The federal government has a solemn obligation to ensure a fair and accurate count of all people in this country. By demanding the citizenship status of each resident, the Trump administration is breaking with decades of policy and potentially causing a major undercount that would threaten billions in federal funds and New York’s fair representation in Congress and the Electoral College.” The plaintiffs go on to allege that the administration deliberately inserted the citizenship question into the census precisely for this reason.
In July, a federal judge in New York City allowed the lawsuits to go forward, even as he expressed doubts about their likelihood of success. Congress, he pointed out, delegated broad responsibility to the secretary of commerce to administer the census and to devise appropriate questions for it. The judge narrowed the case to the claim that the Trump administration inserted the citizenship question on the basis of biased motives, a catchall argument according to which measures that are otherwise perfectly legal and constitutional are ruled out of bounds if adopted for politically questionable purposes. No matter what result the district court reaches, the issue will be appealed through the federal system to the U.S. Supreme Court, with a resolution due in time for the census to proceed on April 1, 2020. In all likelihood, the citizenship question will appear in the census.
The case against the citizenship question rests on two trivial and unsupported claims but also upon two profoundly important and controversial arguments.
One trivial claim is that the citizenship question must be taken out because it’s unnecessary for an accurate enumeration of the U.S. population. As the federal judge in New York City pointed out, the census from the beginning has included various items unrelated to a simple enumeration of the population as required in Article I, Section 2 of the Constitution—for example, items about race, gender, marital status, age, and other demographic factors. In the past, the census has included questions about running water, indoor plumbing, monthly rent, the number of radios in a household, type of heating equipment and fuel, occupation and education, and the like. The Founders viewed the census as an opportunity not only to perform a count of the population but also to gather economic and demographic information useful in monitoring national progress. The Constitution neither requires nor prevents these additional items from being included, according to the judgment of political authorities. The citizenship question is of a like character.
A second doubtful argument is that the citizenship question will lead to an undercount of the population because unauthorized immigrants will be unlikely to respond. This assumes that they would otherwise respond to the census if not for the citizenship question—a dubious proposition. Those with reason to remain in the shadows are unlikely to be lured into the open by assurances that the census will not ask about citizenship status. Even in 2010 (when the question was not asked), the Census Bureau estimated that the U.S. Latino population was undercounted by about 1.5 percent. Some private groups maintain that the undercount was several times greater than this. The California Department of Finance claims that the 2010 census undercounted that state’s population by 1.5 million residents. Illegal immigrants already have reasons to be suspicious of government information collectors, leaving aside the citizenship question. In the meantime, the paranoia about the citizenship question in establishment circles will only reinforce those suspicions and discourage cooperation with the census, which is required by law.
The deeper questions have to do with whether illegal immigrants must be counted for purposes of political representation and the allocation of legislative seats within the states and congressional seats among the states. Most federal benefit programs, including Medicaid, TANF (welfare), and SNAP (food stamps), already restrict eligibility to citizens or to legal residents. Those attacking the citizenship question maintain that representation should be allocated according to total population counts rather than on the basis of other factors that would exclude illegal immigrants, such as legal residency, citizenship, or voting eligibility. A basis for this claim can be found in the census clause in the Constitution stating that representation in Congress must be based on a count of “persons” rather than “citizens.” On the other hand, federal courts have never made an official determination as to whether unauthorized immigrants or people in the country illegally must be included in the category of “persons” for the purpose of representation.
Critics of the citizenship question in the census fear that this information will provide a basis for lawsuits challenging the allocation of legislative and congressional seats according to total population. Though the ACS gathers information on citizenship, the Supreme Court has ruled that such sampling results can’t be used by states for purposes of legislative and congressional districting. The states must rely on actual counts from the decennial census.
In 2016, the Supreme Court decided a case (Evenwel v. Abbott) in which two voters challenged the apportionment of state legislative seats in Texas on the basis of total population according to the 2010 census. The plaintiffs claimed that the apportionment based on total population counts violated the Equal Protection Clause of the Fourteenth Amendment because voters in districts with greater numbers of unauthorized immigrants had greater weight than voters in districts with smaller numbers. In their brief, the plaintiffs showed that some state senate districts had more than 500,000 registered voters while others had slightly fewer than 300,000; voters in the smaller districts had 1.6 times the voting power of those in the district with larger numbers of registered voters.
The plaintiffs relied on Supreme Court precedents in the reapportionment cases of the 1960s in which the justices laid down the principle of “one person, one vote.” Those cases addressed the problem of “malapportionment” caused when state governments created legislative districts of vastly different population sizes, usually to the detriment of urban areas and in favor of rural districts with inherited legislative power. As people within various states moved from rural to urban and suburban areas, legislatures were understandably reluctant to adjust district lines to accommodate that evolution—because such an adjustment would force many members out of office. By the early 1960s, when this process had been going on for more than half a century, voters in rural areas in many states had much greater influence in the legislatures than those in more populous areas of those states. Some compared the situation with the old “rotten boroughs” in Great Britain, in which ancient seats in Parliament were controlled by a few prominent families or absentee landlords. Since the legislatures in the American states drew the district lines, no means of remedy existed, absent intervention by the federal courts.
In Reynolds v. Sims (1964), the Court ruled: “To the extent that a citizen’s right to vote is debased, he is that much less a citizen . . . But the basic principle of representative government remains, and must remain, unchanged—the weight of a citizen’s vote cannot be made to depend on where he lives.” The Court was particularly concerned with situations in which voters in some districts are allocated disproportionate influence compared with voters in other districts—precisely the issue of “vote dilution” raised in Evenwel. In Reynolds, the Court went further: “The Equal Protection Clause requires substantially equal legislative representation for all citizens in a State regardless of where they reside.” Those words suggest that citizenship ought to be the basis of representation. On the other hand, the Court also said in that case that legislative districts must be based on population and that “[l]egislators represent people, not trees or acres.” The Court thus conflated the two concepts of citizenship and population, no doubt because those cases were decided at a time when immigration (legal and illegal) was at near its lowest point in the history of the nation.
But in a subsequent apportionment case, Burns v. Richardson (1966), the Court clarified this point in permitting the state of Hawaii to use a districting formula based on registered voters, rather than total population. In giving Hawaii leeway to experiment with different districting formulas, the Court wrote: “Neither in Reynolds v. Sims nor in any other decision has this Court suggested that the States are required to include aliens, transients, short-term or temporary residents, or persons denied the vote for conviction of crime, in the apportionment base by which their legislators are distributed and against which compliance with the Equal Protection Clause is to be measured.” The opinion went on to add in a footnote that states may exclude for the purposes of apportionment persons who fail to meet its legal residency requirements (a category that might be read to include unauthorized immigrants). There is thus no precedent in these cases for the claim that states must apportion legislative districts by population.
In Evenwel, the Court ruled that the state of Texas could use population as the basis for legislative districts, in part because, as the state maintained, detailed information on citizenship and registered voters was unavailable or too unreliable to be used for legislative districting and the proposed alternatives to population were unworkable. But the justices, following the precedents established in Reynolds and Burns, stopped short of saying that states must use population as the standard, thus leaving the door open for states in the future to use some other metric (e.g., the number of citizens or registered voters) as the basis for representation. As Justice Samuel Alito wrote in a concurring opinion, “Whether a state is permitted to use some measure other than total population is an important and sensitive question that we can consider if and when we have before us a state districting plan that, unlike the current Texas plan, uses something other than total population as the basis for equalizing the size of districts.” Justice Clarence Thomas elaborated on this theme in writing in a separate concurrence that states should be left free to experiment with other metrics, such as “total population, eligible voters, or any other nondiscriminatory voter base.” The Supreme Court has thus left up in the air the issue of whether representatives at the state and local levels must be allocated on the basis of total population, citizenship, eligible voters, or some other metric.
In response to Evenwel, several states have taken steps to apportion legislative districts on the basis of citizenship. Missouri recently approved a constitutional amendment that would accomplish this goal. A similar measure is under consideration in Nebraska, though that state (like Missouri) has a small number of legal and illegal immigrants, in comparison with such states as California, Florida, and Texas. In any event, the issue is beginning to move in states across the country. In the likely event that some states move further in this direction, they will certainly face lawsuits challenging those moves, with the Supreme Court being asked eventually to resolve the unanswered questions from the reapportionment cases.
This is one reason that the citizenship question on the U.S. Census now carries so much weight: it will provide a factual basis to test claims about the appropriate foundations for legislative and congressional representation. If the citizenship question goes forward, state governments will have the information they need to apportion legislative and congressional districts on some basis other than total population. Inclusion of the citizenship question will also grant legitimacy to the distinction among citizens, legal immigrants, and illegal immigrants—distinctions that some wish to abolish. The furious opposition to the citizenship question is thus no mystery.
It’s no secret, either, that this issue has grown in importance in response to the surge of immigration that the United States has experienced since the 1970s, largely in response to the Immigration and Naturalization Act of 1965 and to the rapidly globalizing economy. About 44 million immigrants currently reside in the United States (defined as those living in the U.S. but born in other countries), or about 14 percent of the population, compared with less than 5 percent in 1970 and 6 percent in 1980. There are, in addition, another 42 million children of immigrants, nearly all now U.S. citizens. Among the immigrant population, roughly half (21 million) are naturalized citizens, and the remainder are legal residents on temporary visas, lawful permanent residents, and illegal immigrants. There are about 1.5 million legal arrivals per year, with 60 percent of new (legal) immigrants coming from China, India, Mexico, Cuba, and Central and South America. Evolving changes in the demographic composition of the United States are thus mostly due to the effects of legal (not illegal) immigration, though the consequences of unauthorized immigration should not be discounted.
The Department of Homeland Security estimated that 12.1 million unauthorized immigrants resided in the U.S. as of January 2014 (the date of its most recent estimate). DHS further estimated that this number increased, on average, by 125,000 per year between 2011 and 2014. If one uses those estimates and projects them into mid-2018, the size of the unauthorized population would currently stand at 12.5 or 12.6 million. The Federation for American Immigration Reform (F.A.I.R.), using a somewhat different methodology, arrives at a current number of about 12.5 million, plus 4.2 million children of illegal immigrants, who, according to current interpretations of the law, are classified as American citizens. The estimate of 12.5 million represents about 4 percent of the total population of about 320 million Americans.
The unauthorized population is not distributed evenly around the country—instead being concentrated in a handful of states and, within those states, in a small number of cities and counties. F.A.I.R. estimates that more than 2.6 million reside in California, 1.9 million in Texas, nearly 1 million in Florida, and close to 900,000 in New York. Thus, more than half of the unauthorized population resides in those four states; and within those states, they are mostly concentrated in Los Angeles and Orange Counties, South Florida, the border areas of Texas and the Houston and Dallas–Fort Worth metropolitan areas, and the New York City metropolitan area. It appears that legal immigrants tend to cluster in many of those areas as well, adding to the situation addressed in Evenwel, where legislative (and congressional) districts contain widely divergent numbers of citizens and voters.
According to the current allocation of congressional seats, the figures regarding unauthorized immigrants add up to about four additional seats for California, three for Texas, and one or two each for Florida and New York (plus an equivalent number of electoral votes). One could argue that those seats have been distributed at the expense of several states with far smaller numbers of unauthorized immigrants, such as Michigan, Minnesota, Ohio, Alabama, Pennsylvania, Rhode Island, and West Virginia—all states slated to lose at least one congressional seat following the 2020 census.
Could those states file lawsuits against the House of Representatives or the Bureau of the Census, claiming that they have been harmed by the population formula for allocating congressional seats to the several states? That question could easily arise in response to the 2020 census, particularly if the citizenship question goes forward as planned. Given prevailing judicial precedents, the plaintiffs would have a reasonable chance of success.
On a deeper level, the controversy over the census raises questions about what the United States is or should be as a nation-state in the modern world. The New York Times, in an editorial attacking the citizenship question, claims that the Trump administration is trying to “weaponize the census to redefine American democracy for a narrow set of people. They’re trying to fundamentally change what this country is, and aspires to be, by creating different classes of people.” The editorial implies that in the United States, no distinctions can be made between citizens and noncitizens—a manifestly false claim, since the right to vote has always been limited to citizens, and most government benefits are similarly restricted to citizens or legal residents.
This argument and others like it suggests that the United States is evolving into a new and different kind of state—a universal state, as it were, in which there are no distinctions among citizens and noncitizens, where the government represents everyone without regard to citizenship, where all are welcome and have equal rights, and with those residing elsewhere having rights to enter without restriction (because restriction would violate universal rights). It’s easy to see how such an idea could take hold in a country like ours, based, as it is, upon universal principles of equal rights and individual liberty and with a founding document stating that “all men are created equal.” Philosophers and historians in the past have speculated about the possibilities of a universal state of this kind, though they have disagreed as to whether it would be a good or a bad thing, and whether or not it would work in practice. Yet until recently, it was regarded as mostly a speculative idea.
A nation-state, as traditionally understood, protects, represents, and secures the rights of its citizens first and foremost, while, of course, respecting the rights of visitors and noncitizens. For more than two centuries, the United States has worked out successful compromises between its commitment to universal rights and its real-world existence as a nation-state representing the American people—save for a singular exception in the middle decades of the nineteenth century. Those compromises are beginning to come apart once again, under pressures of immigration and globalization, with the census controversy now throwing into clear relief polarized interpretations of the nature of the American regime. The controversy may prove difficult to resolve under present conditions, and could turn out to raise the most explosive national questions since 1861.
Photo by Drew Angerer/Getty Images