Last week, the Department of Commerce, which oversees the Census Bureau, announced that the 2020 census will ask respondents whether they are citizens of the United States. In response to what should be an unremarkable item of news, some on the left have — of course — forecast a parade of horribles. They warned that Donald Trump was “sabotaging” the survey in a sinister ploy to consolidate power and strike fear in the hearts of immigrants.
But one purpose of any national census is to obtain basic and reliable data about the nation, and how many citizens reside in a country is as basic as it gets. That is why several countries, from Australia to the United Kingdom, include the citizenship question in their surveys. It is also why the United States includes the question in its own. Before 1960, the decennial census asked respondents if they were citizens. Afterwards, the question was included on the long-form questionnaire (issued as a supplement to the decennial survey to one in six households). In 2010, the American Community Survey replaced the long-form questionnaire, and it, too, asks the citizenship question.
Some have attacked the decision on quasi-constitutional grounds. The census is principally for determining the apportionment of House representatives, and federal law requires the census to count citizens and non-citizens alike for that purpose. But it does not follow that the census must preserve public ignorance about the number of citizens. The two are not mutually exclusive, and this decision does not change the way representatives are apportioned.
Others have voiced the concern that including the question may depress the survey’s response rate, yielding unreliable data and potentially diminishing the apportionment of House seats and electoral votes in states with immigrant-heavy populations. Immigrants fearful of the Trump administration, the worry goes, will decline to fill out the survey, thus making the data less accurate and leading to an undercount. In addition to determining apportionment, the census affects the distribution of federal dollars and is used to enforce voting-rights laws, so it is reasonable to want its data to be as accurate as possible. But no evidence supports the claim that including the citizenship question on the census will have a significant effect on either its response rate or the reliability of its data.
The more deranged corners of the Left, meanwhile, have declared that the president wants to root out unauthorized immigrants and deliberately engineer an undercount to ensure that Republicans remain in control of the government. Of course, a citizenship question is just that: Respondents will not be asked about their legal status. And in the event that the census actually undercounts the population in immigrant-heavy regions, red states such as Arizona and Texas would lose House seats and electoral votes.
In reality, the Commerce Department is making a mundane change to grant a December request from the Department of Justice. The Justice Department said it needed more data on the location of voters to buttress its enforcement of Section 2 of the Voting Rights Act, which prohibits states from passing racially discriminatory election laws. The scope of that provision has been interpreted very broadly to block assorted redistricting efforts and to require majority-minority districts, and the DOJ says it needs more data to properly enforce the law.
Regardless, there is nothing untoward about asking people whether they are citizens of a country or not. If Democrats really feared the possibility of an undercount, they would be vigorously urging people to disregard the fearmongering and respond to the census. Instead, they are throwing a conspiratorial fit.