New Report on Undercount of Young Children in 2020 Census and Comparison to 2010

A new report on the undercount of young children in the 2020 Census also provides a comparison between 2010 and 2020 Census efforts in counting young children. It “provides an overview of the research and new activities related to the net undercount of young children that occurred as planning for the 2020 Census began,” including “a review of new activity within the Census Bureau, mobilization of researchers and advocates on this issue outside the Census Bureau, and how the Census Bureau and the child advocacy community worked together on this problem.”

The report focuses “on the net undercount of young children in recent U.S. Censuses,” but notes that the “issue is bigger than the 2010 U.S. Census. Over the past 70 years, the U.S. Census has seen a consistently high net undercount of young children and there is an under-reporting of young children in major Census Bureau surveys…. Young children have high net undercounts in many other countries around the world…. In other words, this problem exists in many different cultures with different census-taking traditions. Despite changes in the U.S. Census Bureau methods over the past 70 years, as well as fundamental changes in U.S. society, the undercount of young children has remained high.”

STANDARD DEVIATIONS: How Widening Gaps in Decennial Census and American Community Survey Response in Easy and Hard to Count Communities Lead to Inequitable Allocation of Social Program Funding

By Edward Kissam

Standard Deviations blog posts represent the views of the author/organization, but not necessarily those of the Census Project.

Here, I briefly review the evidence suggesting that a combination of long-term “structural” limitations in Census Bureau methodology and the distinctive pandemic-related circumstances that likely led to systematic differential undercount in hard-to-count neighborhoods and communities in 2020 may have seriously impaired data quality for the coming decade. 

My observations stem from analysis of ACS response and Census 2020 self-response in a virtual “case study” of a California area (Fresno County) with high concentrations of agricultural workers and other low-income households of Mexican immigrants. In light of this evidence, it is almost certainly the case that currently available decennial census and ACS data will not be “fit for use” for equitably allocating public sector funding in this area and others like it.

I argue, however, that if more detailed operational metrics were made available, it would be possible to assess the extent to which the characteristics associated with low self-response in hard-to-count Hispanic census tracts with concentrations of non-citizens also undermined effectiveness of nonresponse follow-up (NRFU) and led to differential undercount in this stratum of census tracts.

–              Ed Kissam is a researcher who has focused on farmworkers and immigrant settlement in the U.S. for three decades. Since the beginning of the pandemic, he has worked on strategic response to COVID-19 in farmworker communities as part of several collaborative networks of grassroots organizations, healthcare providers, and university researchers. He is co-trustee of the WKF Fund which provides support to a range of social justice initiatives.

How Administrative Data Played into the 2020 Census

As administrative data – datasets collected by government agencies and commercial entities typically as a byproduct of nonstatistical activities, such as delivering mail – rises in importance in Census Bureau survey measurement, the Georgetown Center on Poverty and Inequality has released a new report on its use in the 2020 Census.

The Bureau has long used such data to provide “information about individuals, housing units, or businesses, including to aid in data quality analysis, build and update address lists, and produce population estimates.” The 2020 Census relied more on administrative data than ever before to “supplement key components of data collection, data processing, and quality control efforts.” The Center believes that administrative data “will likely play an increasing role in the federal statistical system, in part due to the growing cost of surveys and declining survey response rates,” but that its use “must be rigorously evaluated and potential impacts on data equity carefully considered as a civil rights issue.”

Quick Summary of President’s FY 2023 Census Bureau Budget Request

On March 28, President Biden released his Fiscal Year 2023 budget proposal. It includes $1.505 billion for the Census Bureau. The President’s request is a $151 million increase over the recently-enacted FY 2022 funding level for the Bureau ($1.354 billion). 
 
According to the White House release, this request includes $408 million “to finalize and evaluate the Decennial Census and lay the groundwork for a successful 2030 Census.”
 
The Census Bureau’s FY 2023 Congressional Justification provides more detail. Below are some highlights:

  • $153.4 million ($34 million over FY 2022) to support data collection related to the Economic Census and Census of Governments.
  • $235.6 million ($10 million over FY 2022) to support the American Community Survey (ACS), including an initiative to improve the measurement of sexual orientation and gender identity (SOGI) population in the ACS. 
  • $14.7 million increase in FY 2023 to support the High Frequency Data Program—an initiative that has emerged from the Bureau’s experimental Pulse Surveys. The request enables the Census Bureau to expand the program’s capabilities to provide “data and products quickly and more frequently in response to changing conditions in the economy and society.”
  • $9.4 million increase for Enterprise Data Dissemination, which includes funding for an initiative “to build evidence, evaluations, and improve underlying racial and ethnicity data.”
  • $68 million ($2.5 million increase over FY 2022) for all Household Surveys account, which funds the Current Population Survey and Survey of Income and Program Participation. Within this account, SIPP would receive a $1 million decrease over its FY 2022 level, while CPS would see an increase of approximately $4.5 million to develop and test a self-response web-based instrument.
  • $3.7 million increase in FY 2023 to support the Community Resilience Estimates (CRE) program that began during the pandemic to provide more granular population data. The Census Bureau plans to transition the program from focusing exclusively on the pandemic to address other disasters and “add data for steady state programs.”

How can your state/community better organize around the American Community Survey (ACS)?

“In 2020, Minnesota was able to retain a Congressional seat by the narrowest of margins (by only 26 people!). This victory was built on six years of stakeholder engagement and strategic organizing,” explains Joan Naymark, leader of Minnesotans for the ACS and 2030 Census (MACS) and a member of The Census Project Advisory Committee.

With the recent release of “ACS: America’s Data at Risk,” more communities are looking for ways to get involved and organized to support the ACS – an essential way to help support their own communities.

University of California, Riverside is hosting a webinar on March 30 at 1pm Eastern to explore that Minnesota case, as well as “early and promising efforts to organize the business community in Florida, the Asian American community in Nevada, and Native American and AAPI communities across the country.”

Register for the webinar.

Latest National Advisory Committee Recommendations on Differential Privacy

The National Advisory Committee on Racial, Ethnic, and Other Populations (NAC) met on March 7, 2022 to consider issues around differential privacy, among other topics.

The NAC approved recommendations on how the Bureau messages the issue, such as highlighting the legal requirements for the Bureau to use disclosure avoidance methods, better messaging approaches to distinct population groups, the readability of messages for public consumption, how to handle the content of messages, and the media choices for messaging.

The NAC also considered messaging on differential privacy for data users and made recommendations on the consequences of data suppression. See the meeting agenda and presentations, and read the full NAC recommendations.

Fiscal Year 2022 Enacted: Census Bureau Receives Less Than Expected

Almost six months after Fiscal Year (FY) 2022 began, Congress passed the FY 2022 Consolidated Appropriations Act, H.R. 2471, a bill funding all Federal agencies through September 30, 2022. For the bill to be properly enrolled and presented to the President for his signature, Congress passed a third continuing resolution to keep the federal government open through March 15. President Biden signed the Act into law on March 15.

The Act provides the Census Bureau with $1.354 billion, $88 million below the amount ($1.442 billion) requested by President Biden and approved by the U.S. House of Representatives and $78 million below the level ($1.432 billion) recommended by the U.S. Senate Appropriations Committee. While less than the House and Senate had recommended, the FY22 funding level is $248 million above the Bureau’s FY 2021 enacted funding level ($1.106 billion).

Of the $1.354 billion that the Census Bureau received, $300 million was allocated to the Current Surveys and Programs account along with language expressing congressional “support” for the new High Frequency Data Program and directing the Bureau to spend no less than the amount it spent in FY 2021 on the Survey of Income and Program Participation. The remaining funds, $1.054 billion, were allocated to the Periodic Census and Programs account. Language in this section of the bill directed the Bureau to follow provisions enacted in the FY 2021 funding bill regarding the American Community Survey, cybersecurity, and the Bureau’s existing partnerships with libraries and community technology centers.

In conference report language accompanying the FY 2022 Consolidated Appropriations Act, Congress directs the Bureau to “follow all administrative rules and procedures with respect to adding or modifying existing survey content, and to keep the Committees apprised of these efforts.” This language was adopted in lieu of provisions in the FY 2022 House Commerce Justice Science Appropriations report that directed the Bureau “to begin research or pilot development on proxy data collection of SOGI [Sexual Orientation and Gender Identity] questions in current surveys…” and to report to Congress “on its plan for implementing updated race and ethnicity questions for the 2030 Census and the American Community Survey, including whether the Census Bureau believes that additional testing will be necessary.”

Pundits predict that President Biden will submit his proposed FY 2023 budget to Congress by the end of March. The Census Project will be issuing its official FY 2023 funding recommendation once additional details about the President’s request are available.

Undercount and Overcount In the 2020 Census

In response to the March 10, 2022 release of analyses from the Census Bureau’s Post-Enumeration Survey (PES) and Demographic Analysis Estimates (DA) which estimate how well the 2020 Census counted everyone in the nation and in certain demographic groups, Steve Jost, consultant to The Census Project, appeared on the Morning Rush show on Newsy.

Legislation on Where to Count Prisoners

Rep. Deborah Ross (D-NC-02) introduced the End Prison Gerrymandering Act (H.R. 6550), legislation that would require prisoners to “be attributed to the last usual place of residence before incarceration” for purposes of a decennial census.

Starting with the 2030 Census, H.R. 6550 would require decennial headcounts to, “with respect to an individual incarcerated in a State, Federal, county, or municipal correctional facility as of the decennial census date, attribute such individual to such individual’s last usual place of residence before incarceration.”

The bill also would change where prisoners are counted specifically for purposes of apportionment and redistricting of Congressional seats. “If the tabulation of the number of persons in a State” in the decennial “includes an individual incarcerated in a State, Federal, county, or municipal correctional facility who is treated as a resident of the State because the tabulation attributes the individual to the individual’s last usual place of residence before incarceration,” the state would need to “treat the individual’s last usual place of residence in the State before incarceration as the individual’s place of residence for purposes of congressional redistricting.”