A Question on the 2020 U.S. Census Raises Issues for Marketers

By Michael J. McDermott

Once every 10 years, the U.S. Census Bureau undertakes a gargantuan task, one that the founding fathers of the United States considered so important they mandated it as part of the Constitution. The decennial census exists to compile an accurate count of every person living in the U.S. and to record basic demographic data such as age, sex, and race. Its primary purpose is to serve as an underpinning for the country’s representative democracy, making sure each community gets the right number of representatives in Congress and that public funds are equitably distributed.

Over time, the decennial census has also come to play a secondary role that is critical to businesses, especially those that rely on sales to individual consumers. Census data tracks how many people live in a given area. It describes their households, ages, income levels, occupations, and other characteristics that are important to businesses intent on targeting their marketing accurately.

Why a question about citizenship on the 2020 census could be bad news for marketers.

That is why the marketing community is concerned about plans to include a citizenship question in the 2020 census. When the ANA recently surveyed a group of key committee members about the inclusion of a citizenship question in the next census, more than 80 percent of those with knowledge of the issue said they opposed it.

“Opponents of the citizenship question believe it will create an environment of fear and distrust in immigrant communities and depress response to the census from noncitizens and even legal immigrants,” says Bill Duggan, group EVP at the ANA. He added that such an outcome raises the risk of nonrespondent bias by significantly undercounting immigrant, minority, and low-income populations.

“Census costs could rise, as increased nonresponse would lead to increased in-person follow-up,” Duggan adds. “Most importantly, if immigrants and others — for example, their families, even if some of them are citizens — avoid the national headcount, the census results will be flawed.”
 
Precedent Exists, but Concerns Remain

The 2020 census would not mark the first time a citizenship question was asked of all households included in the census. It was first asked in 1890, during an era of high immigration, according to a Pew Research Center report, and it was included in subsequent decennial censuses through 1950. Since then, it has been asked only of a sampling of households on the census long form or, since 2010, on the American Community Survey, a Census Bureau sample survey that covers 2.6 percent of the population every year.

Late last year, the Department of Justice (DOJ) submitted a request to the Department of Commerce, which oversees the Census Bureau, asking it to reinstate the citizenship question on the 2020 census questionnaire. The request stated that the DOJ “is committed to robust and evenhanded enforcement of the Nation’s civil rights laws and to free and fair elections for all Americans.” It said that data provided by the inclusion of a citizenship question would be “critical” to the department’s enforcement of Section 2 of the Voting Rights Act, which provides protections for minority voters.

On March 26 of this year, Wilbur Ross, secretary of the Department of Commerce, sent a letter to the Census Bureau stating he had “determined that reinstatement of a citizenship question on the 2020 decennial census is necessary to provide complete and accurate data in response to the DOJ request. To minimize any impact on decennial census response rates, I am directing the Census Bureau to place the citizenship question last on the decennial census form.”

The question to be included on the decennial census form is worded and formatted in the same manner as the citizenship question on the ACS. It asks, “Is this person a citizen of the United States?” Five response options are given:

•     Yes, born in the United States
•     Yes, born in Puerto Rico, Guam, the U.S. Virgin Islands, or Northern Marianas
•     Yes, born abroad of U.S. citizen parent or parents
•     Yes, U.S. citizen by naturalization (followed by instructions to print the year of naturalization in a box below the response)
•     No, not a U.S. citizen

Seventeen states and several major cities have filed suit to prevent the citizenship question from being included in the 2020 census, arguing that it violates both the Constitution’s requirement that a census count everyone residing in the United States and laws governing data quality and administrative procedure.

Another lawsuit, filed by a coalition of organizations and citizen plaintiffs, contends that the Trump administration added the question to deter Asian-Americans, Latinos, and some immigrants from completing census forms for fear that the information would be used against them or members of their household, and thus undercount them in the final census tally.

What It Means for Marketers

Other issues aside, therein lies the big concern for marketers. Flawed results would distort the representation of U.S. population estimates and the research benchmarked to it. “Certain populations will likely be undercounted,” Duggan notes. “It could have a particularly negative impact on media that cater to those communities, the companies that research them, and the agencies that help advertise to them. The value marketers see in those consumer segments will be understated, and investments will be reduced.”

Just how dramatic the undercount might be is hard to predict, but Lisa Torres, president, multicultural, at Publicis Media, says she’s seen estimates as high as 60 million people uncounted, a figure equal to almost 18 percent of the Census Bureau’s projected 2020 population.

“In past censuses, the undercount has been less than 4 percent, often significantly less,” says Torres, who is a member of the Alliance for Inclusive and Multicultural Marketing (AIMM). She is also on AIMM’s Data & Measurement working group. “The fact is, this data is essential not only to marketers, but to any organization’s ability to make good business decisions. Think about retailers that have to figure out where to open new stores and what merchandise to put in them. They rely on census data to make those decisions. If that data is flawed, there are huge ramifications to the business’s ability to forecast accurately. That’s a very big concern.”

The effects of such a dramatic undercount would be far-reaching, as well. Torres points out that most datasets used by marketers and research firms are based on census data, and even those that aren’t still use census data as the “true set” against which they are benchmarked. “If the census data is wrong, all data is wrong,” she says. “So suddenly you have goods, products, and services that are not going to the right places or are not addressing the right people, because all the decisions made are based on census data.”

Compounding the challenge for marketers is the fact that for the past decade the lion’s share of U.S. population growth has been coming from foreign-born residents, a trend that is expected to continue for at least 40 years. The Census Bureau estimates that the foreign-born population will have increased 19.9 percent for the period 2010 to 2020 and projects it will grow 18.7 percent from 2020 to 2030. Native-born population growth projections for those two periods are 6.4 percent and 5.6 percent, respectively. Growth rates will be somewhat lower for both groups from 2030 to 2060, but foreign-born will continue to outpace native-born by a factor of three or four to one.

“There is a significant shift taking place in this country. We are moving toward a more multicultural society,” Torres says. “It’s most evident among the millennial generation. Marketers rely on census data to validate where growth is taking place. If segments of the population are not being accurately counted in the census, it could cause marketers to dial back their efforts in the very segments where they should really be increasing them.”

The ANA survey that found more than 80 percent opposition to the inclusion of a citizenship question polled members of the association’s Data & Measurement, Legal Affairs, and Multicultural Marketing & Diversity committees. Many of their comments reflected the same concerns Duggan and Torres raise.

“Minorities would most likely be miscounted and therefore misrepresented. The data would be bad. That would make locating and targeting multicultural groups harder,” wrote one respondent. “It would vastly affect our view of the size of the multicultural market and ultimately is likely to pull dollars away from multicultural marketing and marketing overall,” wrote another.

In a joint letter to the Department of Commerce, ANA CEO Bob Liodice and leaders of other major trade associations representing the advertising industry voiced their members’ opposition to the inclusion of a citizenship question on the 2020 census. They relayed their members’ belief that it would depress response to the census from both noncitizens and even legal residents. The resulting risk of nonrespondent bias “raises troubling issues in the world of marketing, as undercounting would distort the representation of U.S. population estimates and the research benchmarked to it,” the letter states.

Many other organizations, including the Advertising Research Foundation, the American Association of Advertising Agencies, and the Media Rating Council, have voiced support for the ANA’s position on this issue, Duggan says. The Cultural Marketing Council has also sent a letter to Secretary Ross stating its opposition to the inclusion of a citizenship question on the 2020 census.