Joan Alker, the Executive Director of the Georgetown Center for Children and Families, has theorized that falling Medicaid and CHIP enrollment among children is a troubling sign that the One Big Beautiful Bill (OBBB) is harming children. More recently, Maya Goldman, citing Alker, repeated a similar argument in Axios, suggesting that declining child enrollment reflects “spillover effects” from the law even before its major provisions have been broadly implemented.
One of us (Brian) was contacted by Axios and quoted in the piece. Both pieces rely on assumptions that are not demonstrated by the available evidence and overlook several alternative explanations that are at least as plausible—and in many cases more obvious—than the explanations they advance.
Both articles begin with a decline in child Medicaid enrollment and then work backward toward their preferred explanation. Yet they do not actually know whether uninsured rates among children have increased, which the Axios article does explicitly acknowledge.
The entire narrative rests on the assumption that declining enrollment must mean children are losing coverage. And then they further claim that the loss of coverage is because of the OBBB. And because the OBBB contains no direct child eligibility changes, they rely heavily on vague and ill-defined “chilling effects” (Alker) or “spillover effects” (Goldman).
The Pandemic Enrollment Surge Still Shapes Medicaid Enrollment
Serious analysis of Medicaid enrollment must include the COVID-era continuous coverage requirement. Both Alker’s piece and Goldman’s write-up miss this vital element.
For more than three years, states were largely prohibited from removing people from Medicaid, even if they were no longer eligible. By the spring of 2023, roughly 18 million people enrolled in Medicaid were no longer eligible for the program. They remained on the rolls because states were prohibited from conducting normal eligibility reviews.
Everyone understood that enrollment would decline once states resumed redeterminations in 2023. Yet enrollment proved remarkably sticky. Even after well over a year of unwinding activity, Medicaid and CHIP combined enrollment in January 2025 (79.4 million) remained roughly 8.0 million beneficiaries above February 2020 levels (71.4 million). Medicaid enrollment alone remained more than 7.5 million above pre-pandemic enrollment. The lingering effects of the continuous coverage requirement continue to distort enrollment statistics today.
Alker compares current enrollment trends against a pandemic-inflated baseline and treats enrollment declines as inherently alarming. A much more plausible interpretation is that enrollment remains elevated because improper enrollment accumulated for years and is taking a long time to reverse. Medicaid and CHIP enrollment peaked in April 2023 at 94.7 million people and has steadily declined since but remains 3.5 million people above pre-pandemic levels.
Duplicate Enrollment Is Not Coverage Loss
Another major omission from Alker’s piece and Goldman’s write-up is duplicate enrollment. During and after the continuous coverage period, many people were enrolled in multiple state Medicaid programs or both Medicaid and an Obamacare plan. These situations inflate enrollment numbers without increasing the number of people actually receiving health coverage.
This distinction is critical because many commentators implicitly assume that every enrollment reduction represents a newly uninsured child. That assumption is overly simplistic and false. A child enrolled in Medicaid in two states counts as two enrollments even though there is only one child. Removing the duplicate record lowers enrollment but does not reduce coverage.
The Congressional Budget Office itself projects that approximately 600,000 children will be removed from duplicate Medicaid enrollment during the budget window. Those children are not becoming uninsured. They are not losing eligibility. They are being counted correctly. Yet many critics of the OBBB characterize these enrollment reductions as evidence that children are losing coverage. This conflates administrative cleanup with actual coverage loss. Indeed, CBO’s own estimates show that projected reductions in child Medicaid enrollment are not synonymous with projected reductions in child coverage.
Demographics Matter
The Georgetown and Axios pieces also ignore demographic trends. The United States continues to experience a sustained decline in fertility. The United States recorded roughly 4.3 million births in 2007 compared with about 3.6 million births in 2024, a decline of more than 15 percent. Fewer children are being born, which means fewer young children are entering Medicaid and CHIP eligibility categories. At the same time, older teenagers continue aging out of children’s coverage categories. These demographic forces alone place downward pressure on child enrollment.
These demographic changes matter. In 2024, the Census Bureau reported 3.6 million babies under age 1, 3.7 million 1 year-olds, 4.5 million 17-year-olds and 4.4 million 18-year-olds. As fewer children are entering the population while more 17- and 18-year-olds are aging out of child eligibility categories, child enrollment will naturally decline over time. Yet fertility trends are largely overlooked in discussions that instead rush toward policy explanations without first accounting for demographic change. Despite about one million fewer children in the U.S. since 2020, enrollment in Medicaid/CHIP is at pre-pandemic levels—42.2 million children in both February 2020 and February 2026. It reached a high of 49.6 million children in April 2023 (when states started doing Medicaid redeterminations).
The Data Do Not Support the “Chilling Effects” Theory
Unable to identify a direct child-eligibility provision that would explain the enrollment decline, Alker referred to “chilling effects” and Goldman to “spillover effects.” Since the law’s major eligibility provisions do not apply to children and largely have not yet taken effect, the theory is that parents may become confused, discouraged, or fearful and therefore fail to enroll eligible children.
This argument is largely speculative and ignores several more obvious explanations.
The Axios article quotes Alker suggesting that confusion surrounding work requirements and immigration enforcement may be discouraging enrollment. But the article provides no evidence that these purported spillover effects are actually responsible for the enrollment declines. Instead, readers are asked to accept a chain of assumptions. Enrollment is falling. Therefore, confusion must be increasing. Therefore, children must be losing coverage.
A rudimentary empirical investigation would have cast doubt on the “chilling effect” theory whereby, as Alker puts it, “parents of citizen children living in mixed status families are too fearful of deportation and other negative consequences if they share contact information.” If that explanation were accurate, one would expect states with the largest undocumented populations to report the steepest declines in child Medicaid enrollment. But that’s not what the data shows. The correlation between Alker’s estimates of the percent change in child Medicaid enrollment from January 2025 to January 2026 and estimates from the Pew Research Center of the share of undocumented immigrants by state in 2023 is -0.02. Thus, there is no meaningful relationship between these variables.
Once again, the available evidence does not support Alker’s preferred explanation.
Meanwhile, several far more obvious explanations are readily available: the prolonged unwinding of continuous coverage, removal of duplicate enrollment, correction of improper enrollment, declining fertility, aging out of child eligibility categories, and efforts to ensure that public benefits are reserved for people who are legally eligible to receive them. The American people overwhelmingly support ensuring that taxpayer-funded benefits are not provided to unauthorized immigrants. Yet that possibility is often treated as politically inconvenient and therefore excluded from the analysis.
The OBBB Did Not Target Children
The timing problem is equally damaging to the narrative. The eligibility provisions in the OBBB were directed at able-bodied, working-age adults enrolled through the Affordable Care Act’s Medicaid expansion population. The work requirements, which have not yet taken effect, apply to expansion adults, not children. More frequent eligibility reviews apply to expansion adults, not children. The law’s most consequential Medicaid reforms were aimed at provider tax schemes, state directed payment abuses, and financing gimmicks that have dramatically inflated federal spending while diverting resources away from Medicaid’s traditional vulnerable populations.
In other words, the legislation was designed to strengthen program integrity and refocus Medicaid on children, pregnant women, seniors, and people with disabilities—not reduce coverage for these vulnerable groups.
Enrollment Is Not the Same Thing as Coverage
The discussion becomes even more misleading because enrollment is often presented as synonymous with access to care. Medicaid does not work that way. Children retain important protections through retroactive eligibility. In general, Medicaid can cover qualifying medical expenses incurred during the prior two months before enrollment for eligible children. That means eligible children who need care can receive services and subsequently enroll to obtain coverage for those expenses. This important feature of the program is almost never mentioned in stories about enrollment declines—and was not mentioned in Georgetown’s misleading study or Axios’s coverage—even though it fundamentally changes how enrollment statistics should be interpreted.
The Goal Should Be Healthy Children, Not Maximum Medicaid Enrollment
Most importantly, the goal should not be maximizing Medicaid enrollment. The goal should be healthy children. Enrollment is not the same thing as coverage. Coverage is not the same thing as access to care. And access to care is not the same thing as child health.
Health coverage matters, but it is only one factor affecting child well-being. Family stability, nutrition, exercise, educational attainment, social engagement, mental health, and increasingly the effects of social media and excessive screen time often have a greater impact on long-term outcomes than whether a child is enrolled in a particular public program at a given moment.
Even if the discussion is limited to insurance coverage, Medicaid enrollment should not be treated as the ideal outcome. When families move from Medicaid to employer-sponsored insurance, they generally gain access to broader provider networks, greater physician participation, and better access to specialists. A family that moves from Medicaid to private coverage has not suffered a loss. In most cases, it has moved to a better form of coverage.
Summary of Alker’s Mistakes
Neither Georgetown nor Axios identify whether enrollment declines reflect:
- continued clean-up of rolls from the pandemic-era continuous coverage measures,
- movement to employer-sponsored insurance,
- movement to Obamacare exchange coverage,
- duplicate enrollment removal,
- unauthorized immigrant enrollment cleanup,
- demographic changes, or
- actual increases in uninsurance.
The pieces don’t reflect much intellectual curiosity about finding a true explanation. Alker’s analysis omits key information and relies on a series of assumptions that have not been demonstrated—and when some of these assumptions are tested, they don’t hold up to scrutiny. Until the issues we raise above are explored empirically, claims that the OBBB is causing children to lose coverage remain speculative and likely wrong. The available evidence is far more consistent with the delayed unwinding of pandemic-era enrollment distortions, demographic change, duplicate enrollment removal, and improved program integrity than with the narrative advanced by opponents of the Medicaid reforms in the OBBB.
