public-policy article

Washington Policy Council: 2020 Census

October 26, 2018

Listen to a complete recording of the most recent Washington Policy Council:

Undercounting of Young Children

Deborah Stein, Network Director for the Partnership for America’s Children, joined the Washington Policy Council on Tuesday to discuss the importance of ensuring all children are counted during the upcoming 2020 Census. Undercounting young children in the census has been a historical problem. In 2010, enumerators missed more than 10 percent of children age four or younger, 2.2 million in total. The majority of these undercounted children live in communities of color.

Why are kids the missed? The answer is complex, and research led by demographer Dr. William O’Hare is still in the early stages. What we do know is that, in most cases where young children are not counted, the household returns the form but leaves off the youngest children. The Census Bureau primarily funds outreach and research on “hard to count” populations geographically.  It has not accounted for situations where the census form is returned, but a child, or children, are left off. In these cases, the Census Bureau requires a different approach.

Preliminary indicators show that kids are often not counted because they live in complex households. For instance, custody of the child may be shared across multiple households, or the child might live in a household with a grandparent or other relative. Notably, we do know that these complex family households most often are headed by a young adult, renters, or non-English speakers. These factors all pose additional problems for outreach by census enumerators.

Census data, in particular the Federal Medical Assistance Percentages (FMAP) are responsible for the funding amount that goes to states for many federal medical and social service programs. Beyond that, significant state and local planning is dependent on the census for things such as school class size, business development, and other publicly available resources. When the population is undercounted, the need for teachers, grocery stores, and more, are underestimated. Further, public policy research relies heavily on census data because it is the only place where every person in the country is counted. If the census is significantly off the mark, data that policy analysts will be working with for the next decade will be skewed.

The Count All Kids Committee is working directly with the Census Bureau to improve plans to reach families with young children. The committee also supports state and local advocates in setting up Complete Count Committees that will conduct outreach in communities and help increase the response rate and accuracy of the census with respect to young children.

What Can You Do?

You can support the Count All Kids Campaign by:

  • Sign up for the coalition and campaign at;
  • Invite someone from the Committee to speak at your conference, webinar, or other event;
  • Highlight the undercount of young children in your organization’s newsletter;
  • Ask your local affiliates to help organization Complete Count Committees in states;
  • Distribute widely the outreach materials from the campaign as they become available.

For more information you can contact Deborah Stein directly at


“The 11th Question”

Sarah Brannon, Senior Attorney with the Voting Rights Project at the American Civil Liberties Union, provided an update about the pending litigation challenging the inclusion of a question asking for respondents’ citizenship status on the 2020 Census questionnaire. The ACLU is currently involved in six lawsuits pending across three federal courts in California, Maryland, and New York. While progressing at different rates, the cases are substantively the same, and the first case is scheduled to go to trial in New York on November 5. Under federal law, no changes to the census questionnaire are permitted after June 20, 2019 without permission from Congress, so the litigation must be resolved by that time to ensure the citizenship question can be removed if the challenge proves successful.

The Census Bureau added the citizenship question “at the request” of the Department of Justice to assist the Department in enforcing the Voting Rights Act of 1965. However, citizenship status has not been asked on the census since 1950 and the Voting Rights Act has been enforced for 43 years without the citizenship question on the census. The lawsuits challenge the citizenship question on constitutional grounds and under the Administrative Procedures Act (APA). This poses a high legal burden for the challengers because these kinds of decisions by cabinet-level officials have rarely been litigated and, under the APA, the Commerce Secretary is granted wide discretion. Nonetheless, the ACLU’s complaint asserts evidence that the Census Bureau itself has stated that the new data collected from the citizenship question to enforce the Voting Rights Act is less reliable than already-available administrative records. Further, the question would increase the cost of the census by $90 million to compensate for the expected lower participation rate due to the inclusion of the question.

The Chief Scientist at the Bureau agreed that the question will have a negative impact, and citizenship data will not even be usable. This is because people will likely leave the question blank and “citizenship status” is a particularly hard question to backfill with imputed data collected from within the community or neighborhood. Because of these findings, including the citizenship question goes against long-standing survey best practices to refrain from including a question that will depress the response rate unless the question is absolutely necessary.  In this case, the citizenship question not only unnecessary, but actually counterproductive and, at worst, intentionally harmful. The biggest hurdle for the challengers, however, is that there is no quantifiable data to show the actual impact on the inclusion of the question. Because of inadequate funding, only one of five scheduled field tests have been conducted, and while test respondents demonstrated an unwillingness to respond to the citizenship question, the widespread impact on response rate will not become clear until enumerators begin collecting actual responses on April 1, 2020.

Currently, the litigation is in the trial discovery process, and on October 22 the Supreme Court temporarily granted the government’s application for a stay of the order of the lower court to allow the challengers to depose Commerce Secretary Wilbur Ross. The parties have until October 29 to file briefs on this issue and force the Supreme Court ultimately to decide whether Ross will be compelled to testify. However, a dissent attached to the grant of application for stay from Justice Gorsuch strongly indicated that the full Court would ultimately side with the government and prevent Secretary Ross from sitting for a deposition. It is nonetheless still likely that the challengers will have the opportunity to depose John Gore, the official involved in the letter of request from the Department of Justice to add the citizenship question.

Absent unforeseen circumstances, the trial will begin on November 5, and the district court judge will still be permitted consider potential deposition testimony from Secretary Ross even if the Supreme Court approves the deposition after the trial has concluded. The ACLU expects a decision from the trial judge in February 2019. Check out the National Assembly census page for future updates.