1 of 8
Michelle R. Smith, Associated Press
This March 23, 2018, file photo shows an envelope containing a 2018 census letter mailed to a U.S. resident as part of the nation's only test run of the 2020 census.

SALT LAKE CITY — The Supreme Court will soon hear a case on the U.S. census, considering whether a new question asking people their citizenship would lead to better governance or bad data.

It's the latest clash in a seemingly never-ending debate over what belongs on America's decennial resident count. For centuries, government officials have struggled to justify certain questions, including ones about religious affiliation.

"Any question that raises privacy concerns or is controversial will generate a problem," said Margo Anderson, author of "The American Census: A Social History." "The question of religion is so explosive."

Government tracking of individual beliefs calls to mind dangerous possibilities, like targeted policing or faith-based concentration camps. That's understandable but also unfortunate, according to religion researchers, who say a census question on religion could revolutionize the way we understand American faith.

"In a country where religious affiliation is as dynamic as ours is, it's incredibly important just to get baseline religious identity measures," said Daniel Cox, a research fellow in polling and public opinion at the American Enterprise Institute. Census data "would be an incredible asset."

Religion and the census

The Census Bureau has weighed asking about religious affiliation multiple times since 1790, when the first census was taken. It's faced general concerns about lengthening the question form and more specific fears stemming from the sensitive nature of faith, said Anderson, who is a distinguished professor emerita at the University of Wisconsin-Milwaukee.

"Given the history of discrimination of people by religion, officials thought, 'No, we're not going to do this,'" she said, noting that the First Amendment's religious freedom protections were a key deterrent.

Still, the U.S. census hasn't always been religion-free, as the Pew Research Center reported in 2010.

Andrew Harnik, Associated Press
Then-House Minority Leader Nancy Pelosi of Calif., accompanied by Rep. Jose Serrano, D-N.Y., left, Rep. Elijah Cummings, D-Md., third from left, Rep. Al Green, D-Texas, fifth from right, Leadership Conference For Civil And Human Rights President and CEO Vanita Gupta, second from right, and others, speaks at a news conference on Capitol Hill in Washington, Tuesday, May 8, 2018, on the Trump administration's decision to add a new question on citizenship to the 2020 Census.

During the second half of the 19th century, the census included questions about religious organizations, collecting information about average church attendance and income.

In the early 20th century, these questions were shifted to a new, stand-alone Census of Religious Bodies, which took place every 10 years from 1906 to 1946. It also asked denominational leaders when congregations were established, the amount of church debt, religiously affiliated schools and demographic characteristics of attendees, Pew reported.

When the Census of Religious Bodies fizzled out in the mid-1950s due to funding problems and concerns about the government interference with faith groups, the Census Bureau again debated adding "What is your religion?" to the main decennial count. Officials were aware of religious liberty and privacy-related concerns, but they felt it would be beneficial to know more about the geographic spread of faith groups.

"What they were looking for was a simple classification: the number of Protestants, Catholics and Jews at that point," Anderson said.

Many lawmakers, faith leaders and civil liberties organizations claimed even a basic question would threaten religious minority groups, citing events like the Holocaust, she added.

It's easy to imagine similar arguments being made today, said Ryan Burge, a political science instructor at Eastern Illinois University. Just as illegal immigrants fear the backlash from a citizenship question, members of less popular faith groups would worry about what would happen if their census responses ended up in the wrong hands.

" Muslims would definitely be more reluctant to share their faith (than Christians). Atheists and agnostics also might be less likely to respond. "
Ryan Burge, a political science instructor at Eastern Illinois University

"Muslims would definitely be more reluctant to share their faith" than Christians, he said. "Atheists and agnostics also might be less likely to respond."

In the 1950s, the outcry against a religion question was too much for the Census Bureau to overcome.

"Under the circumstances, it was not believed that the value of the statistics based on this question would be great enough to justify overriding (concerns)," said Census Bureau Director Robert W. Burgess in 1957, according to Pew.

However, the debate still wasn't settled.

In 1957, the Census Bureau asked about individual religious beliefs on its Current Population Survey, which seeks to boost understanding of the U.S. labor force. They were able to analyze how religion corresponded with unemployment, income, education and other factors, Pew reported.

Throughout the 1960s and 1970s, community leaders, lawmakers and census officials continued to weigh the potential costs and benefits of adding a religion question to the main decennial questionnaire. Then, in 1976, these conversations seemed to come to a definitive close.

"Asking a (religion) question in the decennial census, in which replies are mandatory, would appear to infringe upon the traditional separation of church and state," announced a statement from the Census Bureau, according to Pew. "Regardless of whether this perception is legally sound, controversy on this very sensitive issue could affect public cooperation in the census and thus jeopardize (its) success."

Later that same year, Congress affirmed this conclusion when it updated census-related policies. Among other amendments, lawmakers banned mandatory census questions on religious beliefs or affiliation.

They decided "the danger of misuse of data is too great," Anderson said.

Powerful data

All survey designers consider factors like length, wording and question order, but few do so knowing their decisions affect the whole country, Anderson said. That distinction helps explain why debates over adding a citizenship or religion question to the U.S. census are so contentious.

"You have to count 330 million people really fast and accurately, so you don't want to complicate it," she said.

" What is the compelling government interest to have it? There are no laws in America that say we give out goods and services based on religious background." "
Ryan Burge, a political science instructor at Eastern Illinois University

Officials spend years preparing for each census, studying how potential new questions could affect participation.

Each census "affects congressional apportionment and federal funding for 10 years. It creates all kinds of problems when people say, 'Let's leave off cousin Charlie,'" Anderson said.

Due to the risks associated with a bad change, the Census Bureau avoids adjustments unless they come with a clear public benefit. That's a difficult hurdle for a religion question to clear, Burge said.

"What is the compelling government interest to have it?," he said. "There are no laws in America that say we give out goods and services based on religious background."

Cox disagreed, arguing that officials would benefit from deeper knowledge of residents' religious affiliations and faith groups' geographic spread. Government leaders would be better prepared to serve people of faith in their region. Data could also drive expanded education opportunities.

"Even just a basic question which captures some of the larger traditions and denominations, like Protestants, Muslims, Jews, Hindus, Catholics and 'nones' would be a huge benefit," Cox said.

Lost potential

Despite the 1976 law, religion researchers still think about what would be possible if the U.S. census asked about religion. Both Cox and Burge shared a variety projects they might pursue if the law changed.

"I would be able to unlock all the things we don't know about religion in America," Burge joked, adding that, at the very least, the question would help him "immeasurably."

Relying on nonprofits and universities to fund and lead religion survey work limits the scope of faith-related findings, Cox said. Few organizations pay for surveys that are large enough to allow for state-level analysis of faith groups. Those that do may not open up the resulting data to researchers who aren't affiliated with their institution.

Ross D. Franklin, Associated Press
This March 15, 2010, file photo, shows copies of the 2010 census forms in Phoenix.

Even the most ambitious current projects — like the U.S. Religion Census, which is conducted by the Association of Statisticians of American Religious Bodies and collects data on congregations in every county — run into problems, Burge said.

For example, the U.S. Religion Census doesn't explore religious "nones" since they don't belong to a congregation.

"It can't count atheists or agnostics since they're not part of a religious body," Burge said.

32 comments on this story

The U.S. census, on the other hand, collects information on everyone in America. If it asked about personal faith, religion researchers could track subtle shifts that aren't captured by other surveys.

"We can't look at things like the relationship between unemployment and religion at the county level," Burge said. "And we can't look at the religious demography of a county and voting."

Since the Census Bureau is unlikely to reconsider a religion question, some "incredibly wealthy" institution should take on the challenge, Cox said.

"It would be fantastic. There's still a large gap in our knowledge," he said.