Medicaid cost projections vary by millions

Only time will tell which Medicaid expansion cost and enrollment projection is correct, and the difference could be hundreds of millions of dollars.

The State of Alaska is using lower estimates generated by the Portland, Ore.-based research firm Evergreen Economics, which projects the number of individuals newly-eligible for Medicaid services under the state’s expansion plan at 41,900 in the 2016 state fiscal year that began July 1. Evergreen expects little growth in that number, with 42,260 expansion-eligible Alaskans in 2021.

The newly-eligible group is generally healthy adults, without dependents, who make less than 138 percent of the federal poverty level; for single individuals that is $20,314 per year and for married couples it’s $27,490 annually. 

Evergreen made its projections to the state last winter. A Feb. 6 memo from Evergreen Economics Vice President Dr. Ted Helvoigt to Department of Health and Social Services Commissioner Valerie Davidson presents the numbers as preliminary results of Evergreen’s Medicaid expansion analysis.

States are eligible to expand Medicaid under the Affordable Care Act. The cost for new enrollees in Alaska will be covered completely by federal money in fiscal year 2016, and for the first half of fiscal year 2017.

Alaska’s share of that cost will gradually increase to a 10 percent match by 2021, which is similar to much of the federal money the state accepts, namely Transportation Department funding for roads and airports.

Gov. Bill Walker announced last month that he plans to accept federal Medicaid expansion funding beginning Sept. 1 without legislative approval.

The newly-eligible population is the foundation for what a larger Medicaid program will cost Alaska, and an April 2013 study by the health care research firm The Lewin Group projected the new Medicaid market to be 54 percent larger than Evergreen.

The Lewin projection was for a newly-eligible population of 64,700 persons in the 2015 calendar year, with 1.4 percent annual growth to 69,700 by 2020.

The Alaska Department of Labor and Workforce Development’s population projection for all adults ages 19-64 over the same period is 0.03 percent growth.

Helvoigt said in an interview that Evergreen used Labor Department population data combined with Behavioral Risk Factor Surveillance System, or BRFSS, data from the state Division of Public Health to form its projection.

Helvoigt presented the Evergreen report to the Legislature in March. The BRFSS data provided demographic information that was weighted and added to the population figures, he said.

“When we reweighted the data using the Department of Labor data, there was essentially no difference from what Public Health came up with in the BRFSS with their weighting scheme,” Helvoigt said.

He noted that the similar results added confidence to the final numbers.

Representatives from The Lewin Group could not be reached for comment on their study completed in early 2013, but the methodology behind the results is detailed in the report.

The Lewin projection of 64,700 newly eligible Alaskans was generated from a simulation based on the federal Census Bureau’s Current Population Survey data from 2008-10, which included demographic characteristics, according to the study.

“The (Current Population Survey) reports up to 40 percent fewer Medicaid enrollees than program data show actually participate in the program,” the Lewin study states. “To correct for this problem, we identified people who appear eligible for Medicaid in these data and assigned a portion of them to Medicaid covered status.”

It is important to note that Alaska’s population was growing at a faster rate from 2008 to 2010 — 1.2 percent annual growth — than between 2012 and 2014, when the state’s population grew just 0.2 percent per year.

If the Lewin projection for the expansion-eligible population is ultimately correct, that would mean nearly 1-in-11 Alaskans fall into that demographic.

Helvoigt surmised that many rural Alaska Natives could fall into the income demographic and technically be eligible for expanded Medicaid coverage. However, that group would already be covered 100 percent by the federal government through Indian Health Services programs, which could skew some data sets.

While Helvoigt said he is “extremely confident” in Evergreen’s eligible population projection, forecasting the number of actual enrollees is much more difficult.Evergreen used Lewin’s 63 percent enrollment rate in its projections.

“What’s that take-up rate going to be?” he said in an interview. “Is it going to be that 63 percent? I just don’t have a great feel for that and I don’t think anyone really does. There’s just a lot of different things that drive that.”

The Lewin study suggests an increase in enrollment among those currently eligible for Medicaid but not enrolled. That could come from increased media coverage of the Medicaid program.

Helvoigt said that is exactly why it is so difficult to predict how many people will sign up. Looking at what has happened in other states that have accepted new Medicaid money — some have spent tens of millions more in match dollars when initial projections turned out to be low — isn’t reliable because public outreach and sentiment is different in each state, he said.

Of the 31 states that have accepted Medicaid expansion, half are likely to be greater than projected enrollment and half are likely to be fewer, Helvoigt predicted.

“No one’s going to hit it right on,” he said.

The demographics of the newly-eligible population could also lend to less enrollment, according to Helvoigt.

“There’s reason to believe that the expansion population —working age, not disabled, no children — that their take-up rate is going to be lower than other current Medicaid eligible (populations) simply because this population is going to be more mobile, in general younger and more male,” he said.

Male or female, young people just out of college could fall into the income demographic but be looking for work that will soon make them ineligible or plan on moving out of Alaska and thus not sign up, he suggested.

In the end, The Lewin Group projects a final tally of about 43,000 new Medicaid enrollees, while Evergreen Economics pegged 26,600, which is the number the State of Alaska is using, at the end of the forecast periods.

The disparity between the enrollment projections and a difference in per-individual cost sums to a significant overall cost gap.

In fiscal year 2021 (state fiscal years runs from July 1 to June 30), Evergreen and the state project Medicaid expansion to cost $224.5 million, of which the state will pick up almost 10 percent.

The Lewin Group projects a cost of $517.2 million in calendar year 2020. That would mean the State of Alaska is on the hook for about $50 million that year, more than the total amount it would pay from fiscal year 2019-21 under the Evergreen projection.

Evergreen estimated a cost of $7,200 per individual in fiscal year 2016. In the 2015 calendar year, Lewin had a cost estimate of $9,700 per person.

The origin of this disparity is less clear, as both firms report using historical Medicaid cost data from groups as demographically similar as the newly-eligible population as possible.

Helvoigt said he ran cost estimates for former Gov. Sean Parnell’s administration, which ultimately turned down expansion, and investigated if there would be an initial spike in Medicaid costs because of health services use by people previously without insurance.

The assumption that new enrollees will seek health care at an abnormally high rate doesn’t hold water based on historical data, according to Helvoigt.

“Regardless of who they were, and even after accounting for inflation, you find that people enroll and then they start using a little more (health care) each year,” he said. “It’s possible that there’s something about this population that’s going to make them very costly, but I don’t know what that would be.”                  

Elwood Brehmer can be reached at [email protected].

Updated: 
11/20/2016 - 12:00pm