Background and Context
The decennial census is mandated in article 1, section 2 of the U.S. Constitution to ensure that “Representatives and direct Taxes shall be apportioned among the several States which may be included within this Union, according to their respective Numbers.” This country’s founding fathers understood that an accurate population count was central to establishing a fair and representative democracy.
Since 1790, the census has been conducted every 10 years. It is one of our country’s longest-standing legacies. The extensive thoroughness of its data has been the source for equal political representation under the Constitution through redistricting and reapportionment of legislative seats across the nation. Lawmakers use census data to identify community needs and to distribute a substantial proportion of federal financial assistance program funding to the states and localities based on population numbers or other community characteristics that the census measures. Business and private industry leaders decide where to locate new plants and services based on census data, creating new jobs and promoting economic growth. It is imperative that all necessary steps are taken to ensure accuracy in population numbers. A simple undercount of one percent could lead a state to forfeit a substantial amount of Census-related federal funds over the next 10 years.
It is this commitment to accuracy that raises questions over the move by the U.S. Department of Commerce last week. Based on a request from the U.S. Justice Department, U.S. Commerce Secretary Wilbur Ross announced on March 26th that the 2020 Census will inquire about citizenship—making this the first time the question will be asked of all U.S. households since 1950.
Historical Undercount in Census
Unfortunately, the Census Bureau has a bad track record of properly counting certain population groups. These groups—referred to as “hard-to-count” by the Census Bureau—are at higher risk of not being fully counted in the decennial census. They include communities of color, people that live in rural areas, and undocumented immigrants. According to the Leadership Conference Education Fund, “These groups have been historically underrepresented in the decennial census for decades; some may experience new or increased vulnerability due to major changes in methodology, such as relying on the internet as the primary way for households to respond to the 2020 Census; and some may be reluctant to respond due to current concerns about how their data may be used.”
In its extensive assessment following the 2010 census, the Census Bureau showed the 2010 census over-counted the total U.S. population by 36,000 people, or 0.01 percent, due mostly to duplicate counts of affluent Whites owning multiple homes.
While this was an improvement from the 2000 census, the census missed about 2.1 percent of Black Americans and 1.5 percent of Hispanics, together accounting for some 1.5 million people. This was in addition to undercounting about 5 percent of American Indians living on reservations and nearly 2 percent of minorities who marked themselves as “some other race.”
Consider for example Harris County—data from the Census 2020 Hard-to-Count Maps estimates that approximately 24% of Harris County’s current population (1,080,750 people) lives in hard-to-count neighborhoods. For comparison, neighboring Waller County has approximately 82% of its current population (38,752 people) that live in such neighborhoods. In addition, in 2016, 19.2% of Harris County’s households had either no Internet subscriptions or dial-up only access, according to the latest American Community Survey estimates. Being left out of the census systematically deprives these groups and their communities of vital public and private resources.
There is plenty of evidence to suggest that one would be inevitably suppressing the number of people counted in the census by including a question inquiring about citizenship. In speaking with Steve Murdock, a Rice University sociologist and demographer and Census Bureau Director from 2007 to 2009, under President George W. Bush, he shared that exact opinion by saying he thinks those that are here with undocumented status will simply not respond out of fear. “I think it will have the effect of suppressing the count and it will lead people to try to stay out of the Census rather than get in it,” said Murdock. Ultimately, local governments are hurt the most by Census undercounts when they try to make the case that there is need for more funding due to population increases. Murdock added, “We are going to find ourselves underfunding some cities, including large cities, in terms of providing the services they need. It will make it difficult for local government to budget for services their community needs when there are inadequate data on the number of persons that need those services.”
Fiscal Implications for Next Decade
As mentioned previously, numerous federal financial assistance programs rely on data derived from the Decennial Census to determine the distribution of federal funds. To better understand the fiscal impacts of the upcoming 2020 Census on states and local areas, George Washington University professor Andrew Reamer conducted an analysis titled Counting for Dollars: The Role of the Decennial Census in the Geographic Distribution of Federal Funds. The project digs into the geographic distribution of funds from the 16 largest census-guided programs and is designed to help stakeholders and policymakers understand the extent to which federal financial assistance is distributed on the basis of census-derived data.
In Fiscal Year (FY) 2015, the 50 states plus the District of Columbia received $589.7 billion from 16 large Census-guided programs, such as Medicaid, Highway Planning and Construction, the National School Lunch Program, and Head Start. Out of that amount, there was $286.1 billion in reimbursements to the states through five grant programs (Medicaid, the Children’s Health Insurance Program (CHIP), Foster Care, Adoption Assistance, and the Child Care and Development Fund) administered by the U.S. Department of Health and Human Services (HHS). Those federal reimbursements are determined by use of the Federal Medical Assistance Percentage (FMAP) formula.
|Table 1: Texas: 16 Large Census-guided Financial Assistance Programs (FY2015)|
|Medical Assistance Program (Medicaid)||HHS||Grants||States||$21,731,712,107|
|Supplemental Nutrition Assistance Program (SNAP)||USDA||Direct Pay||Households||$5,265,698,536|
|Medicare Part B (Supplemental Medical Insurance) – Physicians Fee Schedule Services||HHS||Direct Pay||Providers||$4,548,698,536|
|Highway Planning and Construction||DOT||Grants||States||$3,633,437,152|
|National School Lunch Program||USDA||Grants||States||$1,394,918,358|
|Title I Grants to Local Education Agencies||ED||Grants||LEAs||$1,321,089,793|
|State Children’s Health Insurance Program (S-CHIP)||HHS||Grants||States||$1,068,727,000|
|Section 8 Housing Choice Vouchers||HUD||Direct Pay||Owners||$1,045,825,000|
|Special Education Grants (IDEA)||ED||Grants||States||$984,945,560|
|Head Start/Early Head Start||HHS||Grants||Providers||$678,806,882|
|Supplemental Nutrition Program for Women, Infants, and Children (WIC)||USDA||Grants||States||$534,860,939|
|Section 8 Housing Assistance Payments Program (Project-based)||HUD||Direct Pay||Owners||$330,585,412|
|Health Center Programs (Community, Migrant, Homeless, Public Housing)||HHS||Grants||Providers||$245,890,848|
|Child Care and Development Fund-Entitlement||HHS||Grants||States||$221,754,000|
|Foster Care (Title IV-E)||HHS||Grants||States||$210,415,000|
|Low Income Home Energy Assistance (LIHEAP)||HHS||Grants||States||$117,570,163|
|Source: Counting for Dollars: The Role of the Decennial Census in the Geographic Distribution of Federal Funds|
|Table 2: Texas FMAP-Based Funds Received by Programs, FY2015|
|Traditional||Medicare Part D Clawback||Foster Care||Adoption Assistance||Matching Funds|
|Sources: Traditional Medicaid, Kaiser Family Foundation (sourced from HHS); Medicare Part D Clawback, Federal Funds Information for State (sourced from HHS, for calendar 2015); remaining programs, HHS.|
As Table 1 shows, Texas received a total of $43.3 billion in federal funding for the 16 large census-guided programs in the analysis. $22.3 billion of that came directly from reimbursements through FMAP formula funding (Table 2). That funding ranked third among all states, just behind California and New York.
Each state’s FMAP moves in the opposite direction from its per capital income (PCI), meaning that the lower a state’s PCI, the higher its FMAP and federal dollars it will receive. This is important because the more people that are undercounted in the census in that state, the higher that state’s PCI. Since PCI is calculated by dividing a state’s total income by the total number of population, if there is a lower population count in the decennial census, then that state’s PCI will go up and lead to a lower FMAP percentage and less federal dollars.
The George Washington University analysis projected that if Texas were to have experienced an additional 1% undercount in the 2010 Census, it would have been estimated to lose roughly $291.9 million in FMAP funding in 2015 alone. This is funding for Medicaid, CHIP, foster care, and adoption assistance around the state that would have been needed to be taken from other parts of the budget in order to help fill the gap. Among all states, Texas ranked number one in the potential funding loss in FMAP funding had there been an additional 1% undercount in 2010.
While this analysis only accounts for FMAP funding, there is no denying the countless other financial assistance Texas received from the federal government is reliant on precise census data.
The Census Bureau estimated that the net 2010 undercount for Texas was 0.97%. That means nearly a quarter of a million people in Texas were not counted—and that is just an estimate. Compounding the fact that minorities and rural residents are already undercounted in Texas, with an added layer of fear by undocumented populations knowing the 2020 Census will ask if they are citizens, the potential impact of a Texas-sized undercount could cause economic ripples that last the whole decade.
It’s evident that states with a disproportionate large amount of undocumented populations stand to lose the most from an undercount. We’re talking about states like Nevada, Texas, California, New Jersey, Arizona and Florida. There are 12 states that have undocumented populations higher than the national average of 3.5 percent. Some of those states are Republican seats, but despite the potential of losing Congressional seats in Republican states, the Trump administration seems determined to move forward.
Prior to this decision, Texas was projected to gain as many as three congressional seats. It’s unknown how the inclusion of this question will affect that projection. Dr. Murdock was asked about this potential risk of losing seats. He did not appear at all surprised and suggested Republicans may have bigger plans in mind by having an semi-accurate number of undocumented immigrants in the country. “If I had to say, this hurts Democrats more than Republicans,” Murdock said, “it makes Democrats more vulnerable in that it could diminish minority seats in government.”
Just to be clear, current federal law prohibits the census bureau from releasing data that would identify individuals. However, it does not prohibit the release of individual characteristics of a certain population group. It is not yet clear what the Secretary of Commerce or Justice Department hope to do with this information. The federal government has tied the collection of 2020 Census information to federal law enforcement already by suggesting this will help the Department of Justice better enforce the Voting Rights Act by more accurately measuring how many people are eligible to vote.
Ultimately, the impact on any state, not only Texas, due to a 2020 Census undercount would extend beyond the fiscal concerns and the five HHS programs mentioned in this publication. This is why the Texas Legislative Study Group stands firmly against the inclusion of any question regarding citizenship, or any other question that may have the potential effect of reducing the accuracy of the 2020 Census. It is in each state’s fiscal self-interest to ensure every person is counted.