How a census undercount shortchanges housing, medical programs for the poor

The ongoing legal battle over adding a citizenship question to the 2020 U.S. census is often framed in terms of political representation—population counts in the census determine how many Congressional seats a state gets. Efforts by the Trump Administration to add a question to the decennial count about a person’s immigration status—the subject of a recent Supreme Court decision—have raised worries among census experts and demographers that the query would lead to a significant undercount of the population in the United States, because immigrant self-response is predicted to decline for fear of being targeted by immigration status.

While the apportionment of political power through Congressional seat counts is foundational to democracy, it’s not the only issue at stake with the census. The country’s once-a-decade count is both a survey and a massive data collection enterprise, the key method the federal government uses to properly apportion money for social programs, not mention the basis for research by an untold number of third parties like nonprofits and universities.

How does that impact local government and the community programs that cities and municipalities depend on? According to Sadaf Knight, chief executive officer of the nonpartisan advocacy group Florida Policy Institute, there are billions of dollars in vital funding tied to the outcome of the census, most notably for Medicaid, the Supplemental Nutrition Assistance Program, Medicare Part B, and Pell Grants for college education.

“There are 132 programs and more than $690 billion in funds, as of 2015, tied to census population data,” says Knight.

NY Daily News via Getty Images

How an undercount would impact one county

Curbed decided to examine how the outcome of the census might impact a single county to see how a population miscount might affect it. Curbed chose Osceola County, Florida, part of the larger Central Florida metro region, because it’s one of the fastest growing counties in the U.S. and has seen an increase in its Hispanic community (due in large part to the Puerto Rican migration after Hurricane Maria), one of the populations predicted to be impacted the most by a potential citizenship question.

The funding formulas for federal programs can be complicated, making it hard to parse out the exact impact on one place. For instance, Medicaid is administered through the Federal Medical Assistance Percentage (FMAP). That’s basically the federal government’s share of payment for certain programs⁠—it reimburses states for the money they spend⁠—which can vary from a 50 percent split to an 83 percent split. The federal share is determined by the state’s per capita income (PCI), determined by dividing a state’s total income by its total population. The lower the PCI, the higher the FMAP. Hence, undercounting means a higher PCI, and lower federal contributions.

A study by the George Washington Institute for Public Policy found that for every person not included in the census count, Florida loses $946 in federal dollars that would go toward Medicaid and a handful of other health programs—Children’s Health Insurance Program, Title IV-E Foster Care, Title IV-E Adoption Assistance, and the Child Care and Development Fund. And that’s far from the worst shortfall. The same study found every missed person cost Vermont $2,309. (The study noted that it can’t draw a straight line between an undercount and all census-guided programs).

Keeping that $946 as a rough estimate, Osceola County had a population of 268,685 in 2010. Estimates predict the 2020 population will reach 352,180, with an Hispanic population of 53.7 percent. A Harvard study looking at the impact of a citizenship question on Hispanic census participation estimates 12 percent of this demographic will be missed, or roughly 22,694 people. That means an undercount in Osceola would cause Florida to lose at least $21 million in federal funds.

According to Knight, the threat of an undercount due to the inclusion of the citizenship question is especially dire because it reduces the counts of specific populations, namely the most vulnerable. Often, Census Bureau staff struggle to count low-income families and people of color, due in part to disproportionately being renters and moving more often between addresses. Undercounts of these populations, as well as immigrants, means the money and programs targeted to these groups won’t be enough to cover everyone because many federal grants allocate funds on the basis of census population characteristics, such as age and income.

For instance, the National School Lunch Program uses national census-derived data, so while an Osceola-specific undercount wouldn’t cause a direct shortfall for the county, it could change the program’s overall funding level if the same populations are missed across the country. It’s the same with Section 8 Housing Choice Vouchers because census-derived data determines local housing needs, median income, and market-appropriate contract rents, so undercounts can skew those formulas.

George Washington compiled a long list of other programs impacted by census figures. Sixteen of the largest such programs provide approximately $29.2 million to Florida each year.

Knight says that in 2019 alone, Osceola County received $1.6 million in Community Development Block Grants, which are funds distributed through HUD that support a variety of local programs centered on housing and community needs, based in part on population figures. The funding formula is based on both population and the economic make-up of a community. Losing funding for these kinds of programs would create “ripple effects throughout the economy, with working families, especially children, hardest hit,” Knight says.

Any undercount puts Osceola County—and virtually anywhere in the country—at risk of losing funds. According to Knight, Florida received roughly $44 billion spread across 55 programs from the federal government in fiscal year 2016 to fund nutrition programs, education, transportation, and health care. The last census also awarded Florida, due to the state’s rapid population growth, two additional Congressional seats.

“The needs of our newest residents should be met, and they should be counted and represented,” says Knight. “The services they need, and their representation in Congress, will be inadequate.”

Has the census count already been compromised?

As part of the Supreme Court’s recent decision in Department of Commerce vs. New York, which considers the citizenship question, the majority opinion, authored by Chief Justice John Roberts, said that the rationale that was being used to add the citizenship question was faulty, not that the Commerce Department wasn’t allowed to add the question.

The Trump Administration continues to argue that it should add the question, and even considered delaying the printing of census forms in a bid to get the question added at a later date. Delays in printing and distributing forms may delay the date of the final count, as well as determinations of funding and the redistribution of Congressional seats. It could add to the cost of what’s expected to be the most expensive census in history—an estimated $15.6 billion.

While local governments await the Trump administration’s next move, and the court’s response officials are breathing a temporary sigh of relief.

“The Supreme Court’s ruling to effectively block a citizenship question on the U.S. census survey is a victory for New York, where an undercount of state residents would have serious consequences in terms of loss of federal funding and Congressional seats, among other things,” said Kathryn Wylde, President and CEO of the Partnership for New York City, in a statement.