Harvard Study Highlights Wide Range Of Medicaid Expansion Estimates

If one thing is certain about the 2010 health law, it’s that almost nothing about it is certain. And according to a newly released report, one of the major unknowns involves the law’s expansion of Medicaid, the federal-state health insurance program for the poor.

In a Health Affairs article released Wednesday, researchers at the Harvard School of Public Health say the expansion of Medicaid under the Affordable Care Act  (ACA) could add between 8.5 million and 22.4 million people to the rolls by 2019. It could also drive up annual federal spending on the program by anywhere from $34 billion to $98 billion.

Medicaid expansion begins in 2014, when the threshold for eligibility will be raised nationwide to 133 percent of the federal poverty level. Currently, eligibility varies widely among the states once certain federal minimums have been met.

In 2010, 69 million Americans were enrolled in Medicaid, which provides health insurance to low-income children and adults, including 9 million people who also qualify for Medicare.

In the Harvard study, the researchers’ best estimate is that 13.4 million more people will enroll in Medicaid by 2019 as a result of the law’s expanded eligibility, with annual federal spending going up $58 billion as a result. In March, the Congressional Budget Office estimated that 16 million will enroll by 2019 if the health law is implemented. The CBO did not offer a range for its enrollment estimate, nor did it estimate the increase in annual federal spending in its most recent analysis.

The CBO has not published a full explanation of how it reached its estimates. The new study authors used data from two Census Bureau surveys of income – one that was used by the CBO, and another that was not. The Harvard researchers used the latter survey for their best estimate because it had a larger sample size.

“Our intention was to show that an estimate is an estimate; all estimates have ranges,” said Katherine Swartz, one of the authors and a professor of health economics and policy at Harvard. “Policy-making would be better served by acknowledging that there’s an ‘as high as’ and an ‘as low as.’”

Using one Census Bureau survey over the other had the largest effect on the estimates. The second most important factor in the researchers’ simulations was how they determined eligibility. While the surveys aggregated income among extended family, Medicaid eligibility is determined by household income and other factors, which the authors adjusted for.

Third most important was the expected enrollment rate among currently uninsured adults. The 13.4 million enrollment estimate assumes a 62 percent uptake by adults who will be newly eligible for Medicaid. The CBO has not released its estimated uptake rate, but the researchers believe the CBO assumed 55 percent in its analysis.

Swartz emphasized that the study’s enrollment and spending ranges should not be used to cast doubt on the new law’s effectiveness. Rather, she said, the analysis should be used to inform further policy development that supports the health law goal of decreasing the number of uninsured.

“If fewer people than expected enroll in Medicaid, we need to figure how to get them enrolled so they don’t forgo” health care, Swartz said.

christiant@kff.org