Inaccuracies in the 2020 Census Enumeration Could Create a Misalignment Between States’ Needs

Abstract

As the most accurate reflection of the United States population, the US decennial census is vital to health policymakers and others at all levels of government. Competing priorities related to cost containment and the introduction of new reforms raise concerns about the resources available to the US Census Bureau to conduct an accurate population enumeration in 2020.

We examined the state of the Census Bureau’s preparations for the 2020 Census and how inaccuracies in the coming census enumeration could influence public health and health equity in the coming decade. The results of the 2020 Census will be used to allocate trillions of dollars in federal funding to states, including support for programs vital to public health such as Medicaid and the Special Supplemental Nutrition Program for Women, Infants, and Children.

Inaccuracies in the census enumeration could create a misalignment between states’ needs and allocation of federal resources. Also, a census miscount of the population could create challenges for public health surveillance and research activities that inform public health policies and interventions.

The US decennial census provides the most comprehensive portrait possible of the United States population. The accuracy of its results is vital to policymakers at all levels of government and affects the health and well-being of people in the United States. The function of the census enumeration is to count each person in the United States only once and in the correct place. The breadth of this mission supports its wide-reaching impact on public policy, including public health policies and interventions that persons in the United States navigate every day.

The 2020 Census’s success is dependent on the groundwork laid in the years prior, including development of technological and personnel infrastructure. An estimated funding shortfall of $3.3 billion and leadership gaps currently leave the US Census Bureau without the resources it needs to prepare for the 2020 Census. These concerns have led to placement of the 2020 Census on the Government Accountability Office’s high-risk list, which calls attention to federal agencies and programs at high risk as a result of mismanagement.¹ Although other elements of federal health policy have recently attracted far more attention, much is at stake for the public health community as the 2020 Census draws closer. Here we examine two important topics: the state of preparations for the 2020 Census with respect to the potential for inaccuracies in the decennial enumeration and how census enumeration inaccuracies could affect public and population health.

CHALLENGES FACING THE 2020 CENSUS

The challenges currently facing the Census Bureau are substantial. Specific challenges confronting the 2020 Census include new costs and insufficient funding appropriations, field testing issues, and leadership changes.

Decennial Census Cost and Funding Shortfall

The per-household cost of conducting a US census has risen in recent decades: in 2020-adjusted dollars, the 1970 per-household cost was $16, whereas in 2010 the cost had risen to $92 per household.² These cost increases have been driven by declines in response that require follow-up with households in person or over the telephone. In 2010, nearly 50 million housing units required follow-up at a cost of more than $1.5 billion.³ Magnifying this issue, the 2020 Census will enumerate the largest US population in history. To confront these costs, the Census Bureau is seeking to implement numerous information technology reforms, including allowing respondents to complete census forms online for the first time and allowing enumerators in the field to complete data collection on mobile devices.²

At the same time, the Census Bureau has revised its 2020 Census life-cycle estimate from $12.3 to $15.6 billion, driven largely by the costs of new technologies.⁴ This comes after several years during which White House requests and congressional appropriations for the Census Bureau were lower than needed. The Census Bureau has been instructed by Congress over the past decade that the 2020 Census should cost less than the 2010 Census.^5,6 This funding shortfall affects testing of new cost-saving technologies, establishment of technological and personnel infrastructure, and creation of the community partnerships necessary for success. Although Census Bureau funding increased in early 2018, a much greater increase will be needed in 2019 as the Bureau rapidly builds its capacity to carry out the 2020 Census.

Field Test Cancellations

To ensure that these technological reforms are carried out securely and successfully, testing of new technologies is essential. Beginning in 2016, lack of funding spurred the cancellation of census field test operations at several sites.¹ The Census Bureau planned to conduct three “end-to-end” tests of the census in 2018 (effectively a dress rehearsal for the full 2020 Census carried out in selected locations). In July 2017, however, it was announced that the tests in two of the three locations had been cancelled because of budgetary constraints.⁷ This is of particular concern because these tests, including new survey methodologies and outreach strategies, will no longer be conducted in rural areas that have historically been undercounted in past census enumerations.⁸ Without adequate testing, the Census Bureau cannot ensure that new methods will lead to collection of high-quality data, that they can be successfully implemented at a national scale, or that they will ensure equitable representation of all communities in the 2020 Census.

Leadership Changes

In May 2017, Census Bureau director John Thompson abruptly announced his resignation. As of spring 2018, a replacement has not been nominated for either Census Bureau director or deputy director. Although interim leadership is in place, uncertainty created by Thompson’s departure came during a crucial period in the Bureau’s preparations for the 2020 Census. This is a critical time in which the Bureau requires consistency in vision and priorities to advocate for additional funding to put in place the technological and personnel infrastructure needed to conduct the census.

THE CENSUS AND PUBLIC HEALTH

An accurate census enumeration is among the most basic yet essential tools used to guide and inform public health policymakers. The Census Bureau’s funding shortfall and inability to adequately prepare for the 2020 Census should concern the public health community because of the field’s reliance on accurate data. Census data are vital to three domains of public health relevance: health equity, allocation of funding for health-related federal programs, and the ability of public health practitioners and researchers to quantify disease burdens in the communities they serve.

Undercounting and Health Equity

Although differential undercounting of minorities, well documented since the 1940s, has improved in recent census enumerations, the uncertainties facing the 2020 Census could undermine the Bureau’s efforts to ensure continued representation of historically undercounted populations in the census.^8,9 Undercounting can occur for a variety of reasons, including omission of household addresses from Census Bureau mailing lists, residences in informal housing, or nonresponse to census follow-up efforts.^10,11 A 2010 Census report showed that White populations were overcounted by 0.8%, whereas Black and Hispanic populations were undercounted by 2.1% and 1.5%, respectively.¹²

Minority children in rural areas fare worse: Hispanic children in rural communities were undercounted by 17.4% in the 1990 Census, as compared with an undercount of 6.9% among Hispanic children in urban areas.⁸ Renters, who are disproportionately likely to have low incomes and to be members of minority racial/ethnic groups, were undercounted by 1.1%, whereas homeowners were overcounted by 0.6%.¹² Although it is possible to statistically adjust census counts to reflect undercounting via estimation techniques, it would likely be highly controversial and without historical precedent.

In addition to challenges of undercounting faced in the past, there are concerns about the willingness of Hispanic and Latino immigrant communities and Muslim communities to participate or self-identify in the 2020 Census because of increases in hostility directed at these communities, as well as the addition of a citizenship question in this census. Although the Census Bureau is by law prevented from disclosing personally identifiable information, public perceptions of risk and lack of trust in government may decrease participation in the 2020 Census among members of these communities.¹³

The same factors that contribute to undercounting in the census, such as poverty and unstable housing, are also associated with countless adverse health outcomes that are of specific interest in public health equity efforts.^14,15 As discussed subsequently, however, differential undercounting has the potential to undermine health equity because representation in the census plays an essential role in identifying public health concerns and how resources will be used to address them.

Funding Population Health Programs

Public policy and spending priorities are among the most important societal determinants of health and health outcomes.¹⁶ Census-derived estimates are used widely by the federal government to inform how billions of dollars are allocated each year.¹⁷ A miscount in the decennial census could misalign the distribution of funds with the reality of the population’s need for federal resources. In particular, programs that support direct health care access, such as Medicaid, and programs that address social determinants of health, such as the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC), will be affected.

Medicaid, the third largest domestic program in the federal budget, is financed jointly by states and the federal government. The share of program costs paid for by the federal government is determined through the Federal Medical Assistance Percentage (FMAP). The formula used to determine each state’s FMAP is inversely proportional to the state’s per capita income. States with lower per capita income levels have a greater number of Medicaid enrollees and therefore a higher FMAP, requiring the federal government to pay a greater share of their Medicaid program costs.¹⁸ Were the 2020 Census to undercount a state’s population, it would artificially inflate the state’s per capita income, leading the federal government to pay a smaller share of the state’s Medicaid costs.¹⁹ An undercount, however, would not be uniformly distributed among the US population; states with larger Medicaid-eligible populations could experience a greater degree of undercounting, meaning that states with the greatest need for federal Medicaid support would also be those most likely to be fiscally affected by a census undercount.

Census data are also used, indirectly, in determining the distribution of federal funds to states in support of programs such as WIC. The amount of funding each state’s WIC program receives is proportional to the state’s share of the national population that is eligible for WIC according to age and income. The federal government uses the American Community Survey to determine the size of this population in each state. The survey’s estimates, produced each year by the Census Bureau, are weighted to adhere to decennial census enumerations. An undercount of a state’s WIC-eligible population in the 2020 Census would be propagated in American Community Survey estimates for the next decade and could result in an insufficient allocation of WIC funds to that state.¹⁷

Similar concerns apply to Section 8 housing choice vouchers, Head Start, Early Head Start, the Low Income Home Energy Assistance Program, Title I grants to local education agencies, and many other programs that promote health and the social and economic well-being of US residents.¹⁷ These federal programs support the foundations of public health for the most vulnerable populations in the United States and counter the socioeconomic drivers of health disparities. The long-term consequences of inadequate or misaligned funding for public health and social services are difficult to quantify but have the potential to undermine improvements made in population health.^20–23

Public Health Practice and Research

Public health agencies’ ability to address population-level health is closely tied to the decennial census. Routine and ongoing public health surveillance efforts, including those related to disease-specific incidence, prevalence, and mortality, play a central role in informing the strategy and priorities of public health agencies. The accuracy of these metrics depends on the use of population denominators that reflect the true population.^21,24 Such population estimates are calculated each year after the decennial census through a combination of census results and vital records.²⁵

Problems with the 2020 Census enumeration, such as inconsistent undercounting or overcounting, could therefore bias the population estimates used to determine population-level disease burdens. Without reliable population estimates, public health surveillance could misidentify community disease burdens, making it difficult to target public health interventions to the right population at the right time or to evaluate the effects of interventions already in place. An undercount would have a disproportionate impact on the very populations (e.g., low-income, housing-insecure, or minority groups) most likely to be the target of public health activities as a result of their overall increased disease burden.^14,15

Finally, numerous data sets widely used in public health and health services research rely on accurate census counts, including the American Community Survey, the National Health Interview Survey, and the Medical Expenditure Panel Survey. Subsamples of the US population, derived in part from decennial census enumeration results, are contacted to take part in surveys used to monitor public health and identify emerging trends. These data sets provide invaluable information to researchers and policymakers about topics such as community-level demographic characteristics, access to health care services, health insurance coverage, health-related behaviors, and immunizations.

In each of these data sets, population estimates derived from the decennial census are used to develop survey weights that make results generalizable to the national population. A decennial census undercount of marginalized populations could therefore mean that members of marginalized communities are inadequately or incorrectly represented in research findings used to steer health policy decisions, identify new public health priorities, and evaluate progress made toward public health goals.

CONCLUSION

The mission of the US census to give representation to every person within the country’s borders requires a complex combination of administrative, financial, logistical, and community-engaged coordination. Although the census is historically underfunded, the 2020 Census has particularly inadequate financial resources relative to previous decades, leading to the cancellation of important tests of new census methodologies. Of vital importance to the public health community, accurate enumeration is essential to the health and well-being of people in the United States through support of social programs such as Medicaid. Also, basic public health activities, research, and surveillance will be disrupted without reliable census data.

Representation in the census is a basic right of residents of the United States and is outlined in the US Constitution. The ability of our country and our communities to provide responsive public health support to those they serve is dependent on that right being fulfilled in 2020.