A new article in a Wyoming blog about Medicaid expansion offers a helpful view of the decision facing states. Here is an excerpt from that article:
A report for the Wyoming Department of Health prepared by Milliman, Inc., an actuarial consulting firm, forecasts the added costs of the program at $116 million to $148 million between 2014 and 2020, based on their best estimate of 28,200 new enrollees. The report said the enrollment could be as low as 17,000 and might exceed 44,000. Under the best estimate, Milliman expects about 3,700 “woodwork” cases that the federal government would reimburse at only 57 percent.
In the same report, Milliman notes potential for significant savings. Some current state health programs would be at least partially subsumed under Medicaid, enabling the state to discontinue their funding and save money. But forecasting these savings — frequently called “cost offsets” — is much more difficult than predicting the cost of enrolling new patients in Medicaid.
“Detailed data was available for the Medicaid cost analysis,” said Jill Van Den Bos, a senior consultant at Milliman and the lead author of the study. The researchers used U.S. Census data and claims-data, among other sources, to predict costs.
“But it was harder on the cost-offset side,” Van Den Bos said. Eligibility for some free services offered by the state — such as colorectal cancer screening, and breast and cervical cancer treatment — is set at 250 percent of FPL. But it is difficult to know how many participants in those programs would fall under the eligibility limit of 133 percent of FPL.
“There is no hard data,” Van Den Bos said. “Assuming uniform distribution, it’s about half.” The state might also be able to reduce its bill for funding the Wyoming State Hospital, which took $60 million from the general fund and “generated only $1.4 million in revenue from all third party payers,” including Medicaid, the report said. But once again, the savings are hard to pin down, since it is unclear how much Medicaid will pay for the indigent — for example, how many days of care per year — until more information arrives from Washington.
“The uber-person who had access to all of the data on Earth would still have a better data for the cost side than the cost-offset side,” Van Den Bos said.
This excerpt gets at two of the complicating factors surrounding this kind of analysis:
- First, a range is important. States want to know their full budget exposure, and thus need 100% enrollment scenario estimates, but also have to account for the behavioral vagaries of other enrollment scenarios.
- Second, the data supporting cost estimates is clearer than the data supporting cost offsets. This is compounded with each state having a unique Medicaid situation and its own local set of programs that may be subsumed by expansion, such as the payments to the Wyoming State Hospital mentioned here.
States face a complex decision as they wrestle with whether or not to expand their Medicaid program.