For years, the Count All Kids Campaign has been lifting up the need to test new strategies for counting young children. The Census Bureau’s own data shows that the count of young children has been getting steadily worse since 1980 even as the count of adults has improved. Dr. William O’Hare’s research, published on the Count All Kids website, has shown that the reasons children are missed are typically different than the reasons adults are missed. Our own research shows that even when adults respond to the Census, one in five families either won’t include their young child or do not know whether they should. Census Bureau’s efforts to improve the overall census count is unlikely to improve the count of young children. Improving the count of young children needs to be a special focus.
Members of the Count All Kids Campaign have been pressing the Census Bureau for years to test new strategies for counting young children because what the Bureau have been doing over the past 40 years is not working. We believe that if they keep doing what they have been doing, the problem will just get worse.
We were delighted when the Bureau undertook preliminary testing of some new strategies and when we learned last year that they planned to test them in the 2026 Census Test.
Unfortunately in February, the Bureau released a request for comments that showed that their plan for the 2026 Census test has been dramatically changed. It now will not include any of the tests of strategies to count young children. In addition, the test will include a citizenship question, which is likely to depress the count of young children because one in four young children live with at least one immigrant parent. It is also profoundly flawed in many other ways. The results will be meaningless and potentially misleading. This is also a wasted opportunity to test approaches that might improve the count of young children—and approaches that could improve the count of other historically undercounted populations.
A number of children’s advocates submitted these comments, spelling out why the Bureau must test new strategies for counting young children, and asking the Bureau to complete its original 2026 test plan, if necessary by deferring the test until 2027. Count All Kids will continue to press for tests of new strategies for counting young children wherever possible. We may also urge the Bureau to use those new strategies without testing, where we think they are promising and believe it will not affect the accuracy of the Census. But this is deeply discouraging and frustrating.
A wide range of organizations submitted comments on other aspects of the test, including why they should not include a citizenship question, why using postal workers makes no sense, why using the 45 minute ACS test rather than the actual Decennial questionnaire means the results will be invalid, why it’s a problem that the test was reduced to two sites from six, and many additional important points.
These comments are important as reference tools, because if the Bureau releases its results (the supporting materials both say they will and they won’t), we will want to be able to assess what, if anything, they have learned and we want to be able to articulate why the results are misleading, if they are.
