The Department of Health and Human Services' budget neutrality practices for state Medicaid waivers have occasionally been inconsistent and opaque, according to the Government Accountability Office.
In four out of 10 Medicaid waivers analyzed by the GAO, HHS approved spending limits based on cost growth assumptions higher than state baseline rates, and in some cases, "included costs states never incurred in their base year spending."
In six demonstrations -- Washington DC, Idaho, Michigan, Missouri, New Mexico and Wisconsin -- the GAO found that the approved spending limits aligned with the states' actual historical costs, a sign that the waivers are not increasing federal spending.
But in four states, including Arizona and Texas, the GAO estimates, HHS' five-year spending limits could have been trimmed back by about $21 billion in federal spending and $11 billion in state spending.
HHS has been following a budget neutrality policy for section 1115 demonstrations since the 1980's, with the most recent version published in 2001, but updated at least internally since then, as HHS noted in comments on the GAO's report -- the latest of several going back to the early 2000s.
"We have had long-standing concerns with HHS's policy, process, and criteria for reviewing and approving section 1115 demonstrations, including the absence of a federal process for obtaining public input and a lack of transparency in the basis for approved spending limits," the GAO wrote.
In February 2012, HHS opened section 1115 Medicaid waivers to public comment, as required by the Affordable Care Act, and waiver applications are now posted online via Medicaid.gov.
Between 2007 and May 2012, CMS accepted 62 waiver applications from 37 states, plus Washington, DC, with 13 pending and three withdrawn.
According to the GAO, in Arizona, Indiana and Rhode Island, the demonstrations' spending models used limits "that exceeded benchmark rates," and in Texas included hypothetical costs in the base year of spending.
In setting up baselines for spending in waivers, CMS generally uses either the state's historical spending trend rate or the president's budget trend rate -- whichever is lower. But, the GAO wrote, "We found that the criteria and methods for making the adjustments in these states were not always clear or well supported."
Both Arizona and Texas's waivers, the GAO wrote, "established funding pools to make new supplemental payments beyond what they could have made under traditional Medicaid requirements and receive federal matching funds for the payments."
Arizona received a waiver from 2011 to 2016 letting the state scale back adult Medicaid coverage, freeze childless adult enrollment and create a mostly federally-funded supplemental fund to pay providers for uncompensated care. The waiver's budget calculations for per person spending -- the largest portion of the spending limit -- used "an outdated baseline," with estimates from national growth rates higher than the state's historical data, the GAO wrote.
By the GAO's sampling of recent waivers, HHS' "budget neutrality policy and process did not provide assurances that all recently approved demonstrations will be budget neutral."
Responding to the GAO, HHS' assistant secretary for legislation, Jim Esquera, wrote that "the agency's policy has been consistently applied over the years," even if "budget neutrality has not always been conveyed as broadly and as clearly as it could have been."
HHS disagreed with the GAO on a few points, Esquera wrote, including the GAO's argument that some of the recently approved demonstrations may not end up being budget neutral.
"GAO used only a subset of the data that CMS uses to assess budget neutrality and to determine the appropriate trend rate and estimates used in assessing budget neutrality."
Also, Esquera wrote, in October 2012, CMS released a model template for states to use, to standardize the process, with an accompanying budget worksheet and guidance on commonly used data elements in budgeting.