November 28, 2016 stgblase

By Brian Blase

A new study by Jonathan Gruber, one of the Affordable Care Act’s (ACA) chief economic architects, suggests that roughly two-thirds of new Medicaid enrollees in 2014 were eligible for the program under previous state eligibility criteria—meaning that they were not made eligible by the ACA. If accurate, then a much smaller share of new Medicaid enrollees were made eligible for the program by the ACA than Washington experts commonly believe. For example, the Congressional Budget Office’s (CBO) most recent projection is that only one of six new Medicaid enrollees were eligible for the program before the ACA. Gruber’s results, combined with much higher than expected Medicaid enrollment and spending over the past three years, has profound implications for the distribution of program costs and the effect of a repeal of the ACA.

First, Gruber’s results suggest that people who were already eligible for Medicaid before the ACA have been inappropriately categorized as though they are newly eligible because of the ACA. This is important because the ACA requires the federal government to pay a lot more of the cost of covering newly eligible enrollees than it does for those eligible for Medicaid before the ACA. Thus, it appears the federal government is paying a lot more than it should be and that states are paying far less. Second, Gruber’s results suggest that if the ACA were repealed, a lot fewer people would likely lose coverage than previously thought.

Fewer new Medicaid enrollees were made eligible by the ACA than thought

In March 2016, CBO estimated that 13 million additional people would be enrolled in Medicaid on average in 2016 as a result of the ACA. The agency projected that 11 million, or roughly 85%, would be eligible because of the ACA. These are non-disabled, working age, adult enrollees with income between the state’s previous eligibility threshold and 138 percent of the federal poverty line—the income level to which the ACA expanded Medicaid. The other 2 million consist of people who were eligible for Medicaid under states’ previous eligibility criteria but who came out of the “woodwork” as a result of the ACA.

Gruber and his coauthors, using data from the Census Bureau, estimate that Medicaid “produced 63% of the gains [in coverage] that we identified” for 2014. They also found that much of this gain was attributable to the enrollment in Medicaid of people who were eligible for the program under criteria that preceded the ACA’s Medicaid expansion:

Perhaps less obviously, we also found a substantial increase in Medicaid coverage among children and adults who were already eligible for the program before 2014. This population accounted for 44% of the coverage increase.

Dividing Gruber’s estimate of the percentage gain in coverage of Medicaid enrollees who were eligible before the ACA by the percentage gain in coverage attributable to Medicaid overall means that 70% of new Medicaid enrollees in 2014 were eligible for the program under pre-ACA rules. However, Gruber and his co-authors included people who gained enrollment in six states that initiated the Medicaid expansion prior to 2014 as enrollment of previously eligible enrollees. According to CMS, the total number of people who gained Medicaid coverage in these expansion states prior to 2014 was 950,000. If you assume that all of these people were made eligible for Medicaid by the ACA, then about 40% of new Medicaid enrollees added by 2014 were made eligible by the ACA and about 60% were woodwork enrollees.

Medicaid enrollment and spending explosion

According to Brookings Institution expert Stuart Butler, the ACA has largely become the “Medicaid Expansion Act.” This is because less than half as many people have enrolled in the ACA exchanges as expected and Medicaid enrollment has far exceeded expectations in states that adopted the expansion.

According to a report from the Foundation for Government Accountability, enrollment substantially exceeds states’ expectations of maximum Medicaid expansion enrollment. In some states, the enrollment surge is virtually unbelievable. For example, enrollment of the purported newly eligible Medicaid population in California topped 3.8 million people, more than 4 times the number projected; enrollment in Colorado, Kentucky, Nevada, New York, and Washington was more than double the projected number.

Spending per Medicaid expansion enrollee is also significantly above projections. In its 2014 Medicaid Actuarial Report, the Centers for Medicare and Medicaid Services (CMS) estimated that newly eligible enrollees would cost $4,281, on average, in 2015. In the 2015 report, released in July 2016, CMS upped that cost to $6,366—a 49% increase.

Since enrollment and spending per enrollee are both far above projections, so is total spending on the expansion. In April 2014, the CBO estimated that the federal share of Medicaid expansion spending would be $42 billion in 2015. The actual federal cost was about $68 billion—62% higher than predicted.

A large share of federal spending on the Medicaid expansion may be unlawful

The ratio of newly eligible enrollees to woodwork enrollees is crucial because of very different federal reimbursement rates of state Medicaid spending on the two groups. The ACA created a much higher federal reimbursement rate for state spending on newly eligible enrollees than for previously eligible enrollees—mainly lower-income children, pregnant women, seniors, and people with disabilities.

If Gruber’s estimates are correct, then a far lower percentage of people added to Medicaid over the past three years were made eligible by the ACA than states are reporting to Washington. This means that the federal government has likely paid billions more each year than the law allows for the expansion population while states have spent billions less.

Repealing the Medicaid expansion would likely not result in nearly as much coverage loss as expected

Last year, Congress passed a bill that ended the ACA Medicaid expansion after 2017. Presumably, if a state wanted to continue to cover the Medicaid expansion population they could submit a waiver to CMS, and, if approved, could cover that population at the same reimbursement rate the state receives for the traditional Medicaid populations.

Many experts who are projecting the number of people who would be uninsured if a similar bill is enacted by Congress next year assume that states would not seek to cover the Medicaid expansion population without the elevated federal reimbursement rate. While it is difficult to predict how states might respond to this change of incentives, it is reasonable to assume that some states would continue to cover this population through waivers. Six states and the District of Columbia provided Medicaid for non-disabled, working-age, childless adults before the ACA.

Crucially, CBO must look closely at the Gruber analysis, since it may substantially affect its estimate both of the ACA’s impact and the impact of its repeal. If Gruber is right and a much greater percentage of new Medicaid enrollees are woodwork enrollees than previously thought, then significantly fewer people would lose eligibility for Medicaid if the expansion is repealed. This means that the repeal bill being considered by Congress—even in the absence of any replace bill—will likely increase the number of uninsured by several million people less than widely believed.


Expanding Medicaid was an unwise policy decision made by the drafters of the ACA. The program needed reform, not expansion.

The experience of the Medicaid expansion adds further evidence to why reform is desperately needed, as program enrollment and spending have exploded. States, which had already devised numerous schemes to collect large amounts of federal money under the traditional reimbursement rate, appear to be inappropriately categorizing people as newly eligible enrollees and setting extremely high payment rates to insurers for the Medicaid expansion population. This suggests that fundamental Medicaid reform—a fixed amount of federal tax dollars to states combined with real state flexibility to set the rules for their programs—is needed more than ever.

Update: Gruber’s coauthors on the study were Molly Frean and Benjamin Sommers.

Leave a Reply

Your email address will not be published. Required fields are marked *


Contact Dr. Brian Blase for more information or if you believe he can help your organization achieve its goals and desired policy outcome.


Reston, VA 20191


Blase Policy.