The voices of Urban Institute's researchers and staff
March 28, 2018

Five reasons the addition of a citizenship question to the 2020 Census is detrimental to its goal

March 28, 2018

On Monday, Commerce Secretary Wilbur Ross announced that the 2020 Census would include a citizenship question to better enforce the Voting Rights Act per a request from the Justice Department.

The Voting Rights Act was created to ensure that all Americans had equal access to the polls—a protection of the civil rights to which we have all been guaranteed. But by introducing this question into the Census, the civil rights of many groups are at risk of being further disenfranchised. This decision also risks politicizing the Census Bureau, the key government data collection agency that depends on public trust to complete its mission.

Here are five reasons the introduction of a citizenship question is based on false premises that put the 2020 Census, government data, and our democracy at risk.

1. The claim that the citizenship question is needed to enforce the Voting Rights Act is not based on data.

The Justice and Commerce Departments’ claims that this question is necessary to enforce voter fraud are not based on evidence, as incidents of voter fraud are exceptionally rare. Additionally, the Voting Rights Act of 1965, specifically section 2, was introduced in large part to protect black voters from structural racism—in the forms of poll taxes, literacy tests, and blatant denial of voters who tried to register in some locations.

The Voting Rights Act was not introduced to weed out a few bad actors. It was an attempt to right the wrongs that were widespread in the system. Arguing that the citizenship question is needed for the Voting Rights Act does not align with facts or with the act’s purpose.

2. A new question to enforce the Voting Rights Act at the block level is not necessary.

The Justice and Commerce Departments also argued that having data at the smallest geographic level—the block level—is crucial for the Voting Rights Act and that only the decennial Census can provide this granularity. But this claim is false.

The American Community Survey (ACS) captures citizenship data every year with a robust sample of 3.5 million housing arrangements. The five-year ACS data files provide data about the voting-age population by citizenship and race down to the block group level. The addition of a citizenship question on the decennial census (asked only once every 10 years) is not necessary when the data are updated and available every year through the ACS.

3. Changes to the census require extensive testing, and testing was not completed for the citizenship question.

The field of survey research exists because it is a science—writing and asking questions in an unbiased way is challenging—and is built upon experimentation, testing, and research. This is especially true with a survey as monumental as the decennial census. Changing the relationship question to more accurately count coupled US households was a decade-long process because researchers sought to ensure accuracy and to prevent related issues like nonresponse. The improved relationship question will be included on the 2020 Census because of extensive preparation and vetting.

By introducing the citizenship question on the 2020 Census without adequate testing, the Commerce Department is disregarding the advice of experts and may be putting the entire census at risk. Legislators recently introduced an amendment to existing law, mandating that the commerce secretary provide advance notice before changing questions on the decennial census. But that amendment may be too late for the citizenship question.

4. Many are already suspicious of the census, and this question will heighten their reluctance to participate.

The Constitution mandates that everyone is to be counted. All people, regardless of citizenship status, are to be represented in the census. Enumerators had difficulty counting immigrants during the 2010 Census, which was before the introduction of recent anti-immigration executive orders.

There has been growing concern that the current climate would suppress participation in the census, meaning that immigrants (and their children) who are already undercounted will be further suppressed and less likely to receive their fair allocation of government money and political representation. This is likely to hurt states across the political spectrum with high proportions of immigrants, such as California, Florida, New York, and Texas.

5. An accurate decennial census relies on a good public image and respect for the Census Bureau’s work. This question could threaten that image.

The Commerce Department’s recent decision at the behest of the Justice Department represents a concerning slide toward the entanglement of politics in census operations. Because of recent issues related to data breaches at Equifax and Facebook, public confidence in data has diminished. Although Title 13 guarantees that the Census Bureau keeps all data in its possession confidential, public perceptions are critical to its mission. The politicization of the last-minute addition of the citizenship question could chip away at the public’s confidence in an agency that has carried out its mission with integrity.

The decennial census is one of the safeguards of US democracy. It is mandated by the Constitution and is the guarantor of proportionate representation in Congress and fair allocation of government funding. Introducing the citizenship question on the decennial census is a notch against democracy and evidence.

This post was updated on April 19, 2018, to clarify that the five-year ACS data files provide data at the block group level.

Illustration by Neil Tony/Getty Images.

SHARE THIS PAGE

As an organization, the Urban Institute does not take positions on issues. Experts are independent and empowered to share their evidence-based views and recommendations shaped by research.