This week the U.S. Census Bureau released the Current Population Survey (CPS) Annual Social and Economic Supplement (ASEC) data, which examine poverty, income, health insurance and other trends for calendar year 2025. Normally, the Census Bureau would also release the American Community Survey (ACS) this week, which provides considerably larger sample sizes and hence much more detail on state-by-state comparisons and smaller population groups. But these are not normal times, and we do not yet know when the ACS data will be released. The data have been delayed indefinitely after an unprecedented departmental administrative order a few months ago from the Commerce Department that restricts the methods the Census Bureau can use to protect data confidentiality.
Although both are large federal surveys run by the Census Bureau, the CPS has a number of differences from the ACS data that readers may be more accustomed to seeing in places like CCF’s annual report on children’s health insurance coverage and our State Data Hub. For example, while the ACS is fielded each month and asks respondents about their coverage at the time they answer the survey, the CPS ASEC is fielded in February through April each year and asks respondents about their coverage throughout the previous full calendar year as well as at the time they’re taking the survey, giving us different ways to look at coverage trends.
CPS data indicate that in general, there were no statistically significant changes in the rate of children under age 19 going without health insurance coverage from 2024 to 2025. The rate of children uninsured for the entire previous year dipped slightly, but not by a statistically significant margin, from 6.1% in 2024 to 5.8% in 2025. The rate of children who were uninsured for part of the year ticked up from 4.5% in 2024 to 4.8% in 2025 although this wasn’t a statistically significant change either. Finally, the rate of children who were uninsured in the spring when their families filled out the survey remained at 6.5%.
The only statistically significant changes in children’s coverage were regional: uninsured rates for children in the Northeast fell considerably to just 2.6%, while child uninsured rates in the West increased by 1.1 points to 5.5%. Child uninsured rates remain highest in the South at 8.0%.
However, there are some important reasons to think these numbers may not be telling a complete story. First, the CPS child uninsured rate only counts children as uninsured if they lacked coverage for the entire calendar year, so children who were disenrolled from Medicaid during the year, for example, are not included in that figure. The CPS sample size is also much smaller than that of the ACS, about 100,000 addresses in the CPS compared to 3 million in the ACS. While both surveys are designed to be representative, the smaller sample in the CPS can mean that the estimates are less robust, and they don’t offer ready analysis of smaller geographies like states and counties. On top of this, response rates, which have been falling in both surveys for decades, still haven’t recovered since the pandemic, leaving the CPS sample skewed towards higher-income households and therefore potentially less likely to fully capture individuals who are uninsured or who are enrolled in Medicaid and the Children’s Health Insurance Program (CHIP).
Furthermore, according to administrative data, the number of children enrolled in Medicaid and CHIP declined sharply in 2025 and has continued dropping even more swiftly through the first half of 2026. As of July 2026, CCF’s enrollment tracker is showing almost 2.5 million fewer children enrolled in Medicaid/CHIP as compared to January 2025 when President Trump took office. We have blogged before on why we think the number of children is declining so quickly, including anecdotal reports of children living in mixed-status families whose parents are too fearful to enroll or renew their citizen children in public coverage due to the Trump Administration’s aggressive deportation agenda and unprecedented decision to share Medicaid data with the Department of Homeland Security. Cuts to Medicaid at the state level have also started, though they will really get going in 2027 as many of the nearly $1 trillion in federal cuts to Medicaid start taking effect.
Another question is whether the CPS data ,is experiencing a drop in the number of non-citizens answering the survey. One in four children in the U.S. lives in a family that includes a non-citizen or foreign-born member. If a parent is fearful of enrolling their child in Medicaid or CHIP it seems plausible that they may not be willing to answer a survey fielded by the U.S. Census Bureau. This could be why the CPS data show that the uninsured rate for noncitizen children unexpectedly declined by 2.8 points from 2024 to 2025, although the change was again not statistically significant. (Children who are not citizens still have an uninsured rate more than 3 times higher than children overall at 18.3%.)
If the ACS data had been released today, we would be hard at work dissecting the top-line numbers to see which states were seeing improvements or setbacks in child uninsured rates, whether children of different races and ethnicities, income groups, or age were experiencing different coverage trends. Until those data are released, our analyses will have to wait.
When any number of children lack health insurance, they and their families are decidedly worse off. Research has shown that having Medicaid during childhood (versus being uninsured) leads to better health outcomes in the short and long term, and improved educational and economic outcomes in the long term. And of course, comprehensive, affordable health insurance, such as the coverage provided by Medicaid and CHIP, offers fundamental protection for families’ economic security, particularly at a time when inflation has been stubborn and fuel costs have increased markedly. If there is a window of opportunity to prevent the country from going backwards on such an important indicator of child well-being and family economic security, policymakers must act quickly.