In March, the U.S. Department of Commerce announced that the 2020 census would include a question on respondents’ citizenship. The inclusion of a citizenship question prompted alarm among immigration, faith and civil rights groups, and led elected officials to express concern that undocumented persons would be discouraged from filling out the census, reducing federal funding and congressional representation in the communities in which they live. The census has not contained a citizenship question since 1950, although surveys of sample populations in the Current Population Survey and the annual American Community Survey have included that question as recently as 2000.
Analysts have expressed concerns that the inclusion of the citizenship question could “turbo-charge . . . fears among Hispanics and other immigrant groups” that government agencies will access citizenship information and use it to carry out deportations. And if the undocumented response rate falls, “[t]he result would be systemic undercounting and underfunding of states, cities and towns with substantial populations of Hispanics and other immigrants.”
Initially, many noted that a reduced response rate from undocumented persons would primarily hurt Democratic-leaning “blue states,” like California and New York. However, a closer examination of the undocumented population reveals that the impact of the citizenship question could also have a negative impact on Republican-leaning “red states” given the rapid growth of immigrant populations in many red state rural communities and metro areas over the past two decades.
Impact on States
The decennial census is essential in determining congressional apportionment and federal funding. States with higher populations receive more representatives in the U.S. House of Representatives and a greater share of federal dollars. Because the U.S. Constitution requires that the census be a count of all “persons” residing in the states, not merely a count of citizens or legal residents, undercounting the undocumented population will carry tremendous costs. Alabama has recently filed suit challenging the counting of undocumented people by the census, but that suit faces an uphill battle in the courts, including a 2016 Supreme Court decision that reaffirmed the ability of states to rely on total population – including nonvoters – in redistricting.
Because Congress is reapportioned in accordance with overall population, states with large undocumented populations that would go uncounted stand to lose representation. Due to the growth of the immigrant population in the southeast in recent years, in both rural towns and large southern cities like Atlanta and Charlotte, the impact of a census undercount will be felt in blue and red areas alike. As one expert has noted, the states “most disadvantaged, however, are not those with simply the most undocumented people,” like New York or Illinois. Rather, the states with the highest proportion of undocumented people compared to overall population would be the most impacted. These states include solid blue states like California, Maryland and New Jersey, but also a number of red states and swing states – Arizona, Florida, Nevada, and Texas. To the extent the citizenship question drives down the response rate, these states are most likely to lose congressional representation.
Figure 1: Change in foreign-born adults as a share of total adult county population from 1990 to 2012 (Source: Pew Charitable Trusts)
The impact of the citizenship question goes beyond the undocumented population. The question is likely to have an impact on the census response rates of the 8.8 million citizens and legal residents who live in mixed immigration status households, who may fear responding to the census out of concern that the survey could adversely impact undocumented family members and friends.
According to a recent study, red states stand to lose representation if the response rate of the undocumented population falls. The first state in line to lose a seat in Congress is Florida. If the response rate for undocumented residents falls by 10 percent, Florida loses a seat in Congress. And if 40 percent of the undocumented population elects not to respond to the census, blue states will lose one seat in Congress (in solid blue California), while three states that supported President Trump will each lose a seat (one each in Arizona, Florida and Texas).
A census undercount’s impact will also adversely impact red states’ federal funding for many key programs. A number of red states with poverty rates above the national average could see federal funding for healthcare, housing, and nutrition programs reduced as a result of an undercount.
A March 2018 report by the GW Institute of Public Policy at George Washington University explains that red states face significant losses in financial resources over the next decade if a census undercount impacts the more than $800 billion allocated each year on the basis of census-derived statistics. Of note, five federal grant programs administered by the U.S. Department of Health and Human Services (HHS) utilize the Federal Medical Assistance Percentage (FMAP) to determine funding levels to states. The FMAP, which is based on state population levels determined by the decennial census, determines state reimbursements to and payments from states relating to Medicaid, the Children’s Health Insurance Program, Title IV-E Foster Care, Title IV-E Adoption Assistance, and the Child Care and Development Fund. States with lower per capita incomes tend to be more reliant on these FMAP programs and more likely to be affected by a decline in the census response rate due to the addition of a question about citizenship.
According to the GW study, in FY 2015, 37 states – 27 of which voted for President Trump in 2016 – lost out on funding for these five FMAP programs due to undercounting in the 2010 census. The median FY 2015 loss in those states was $1,091 in federal funding per person living in the state. The other 13 states tend to have higher per capita incomes and generally received lower FMAP funding, making them less susceptible to funding cuts due to undercounting. Ten of these 13 states voted for Hillary Clinton in 2016.
With a citizenship question expected to increase undercounting in the 2020 census, the effect is likely to be even more pronounced after the 2020 census. The GW study calculated the projected loss in FMAP-based program funding for each state if there was an additional 1 percent undercount in the 2010 census. Nine of the ten states with the highest projected losses were 2016 Trump states.
Figure 2: States with largest projected losses in FMAP program funds after additional 1% undercount (Source: GW Institute of Public Policy)
|State||Projected Loss in FMAP Funds|
While much of the conversation surrounding the impact on a 2020 census citizenship question has centered on the impact on “blue states,” many “red states” stand to lose just as much – if not more – were the census response rate were to fall.
Florida and Texas – both 2016 Trump states – would stand to lose representation in Congress after the census. Similarly, because many red states have lower per capita incomes than blue states, and are more reliant on federal funding for key state budget programs, many red states stand to lose millions of dollars in federal funds – an average of more than a thousand dollars for each uncounted resident. Immigrant undercounts pose a significant risk to red states’ congressional representation and budgets.