Indiana is one of six states that have expanded Medicaid using waivers that give them additional flexibility in designing their expansions, which these states have used (depending on the state) to charge premiums to poor beneficiaries, enroll beneficiaries in marketplace coverage, and/or delay the effective date of coverage in ways that aren’t usually allowed, among other things. Several other states may seek to adopt the expansion through a waiver, and some states that have expanded without a waiver — such as Arizona, Kentucky, and Ohio — have filed or are preparing waiver proposals to charge premiums and make other changes.
Some of these states are looking to the “Healthy Indiana Plan (HIP) 2.0” as a model. The most notable features of HIP 2.0 are “POWER Accounts” modeled on health savings accounts, a delay in the effective date of coverage until income-based premiums are paid, and a rule that locks some beneficiaries out of coverage if they don’t pay their premiums. HIP 2.0 is far more complicated than traditional Medicaid, and mounting evidence suggests that its complexity has made it difficult to implement in a way that is consistent with the terms agreed upon by the state and federal government.
Like all waivers, HIP 2.0 is a demonstration project intended to promote Medicaid’s objective of delivering health care services to vulnerable populations who can’t otherwise afford them. Indiana said it would use its waiver to test several hypothesis regarding the impact of premiums on participation in health coverage and the efficient use of health services. State and federal evaluations are in progress to test these hypotheses, and the results aren’t yet available. However, extensive prior research shows that premiums significantly reduce low-income people’s participation in health programs, and Indiana’s early experience under its waiver raises serious questions regarding HIP 2.0 in general and how the state is assessing premiums in particular. To avoid depriving low-income residents in other states of needed health care, these questions should be answered before the Centers for Medicare and Medicaid Services (CMS) allows other states to replicate its design.
HIP 2.0 has two parts: HIP Plus and HIP Basic.[1] Adults with incomes below the poverty line can enroll in either program, while adults with incomes between 100 percent and 138 percent of the poverty line can only enroll in HIP Plus. HIP Plus enrollees must pay a monthly premium of 2 percent of household income (or $1 a month for those with incomes below $50 a month) into “Power Accounts” modeled on health savings accounts (HSAs). HIP Basic has no premiums, but beneficiaries must pay co-pays at the maximum level Medicaid allows, and they receive a more limited benefit package than under HIP Plus.
Enrollment in HIP Plus isn’t effective until a beneficiary makes a premium payment; beneficiaries with incomes below the poverty line who don’t make a payment within 60 days after their eligibility is approved are moved to HIP Basic. After enrollment, HIP Plus enrollees with incomes below the poverty line are moved to HIP Basic if they don’t pay their premiums, while those with incomes above the poverty line lose their coverage and can’t return to the program for six months.
Indiana’s reports to CMS and press accounts suggest that the state has had difficulty implementing some of its core provisions, as explained below.
A central objective of Indiana’s waiver is to test how the program’s premium structure affects enrollment and access to care, and one of its hypotheses is that “POWER account contributions for individuals in the HIP Plus plan are affordable and do not create a barrier to health care access.”[2] But this hypothesis can’t be fairly tested if the state isn’t charging premiums based on beneficiaries’ actual incomes, as Indiana agreed to do. Instead, Indiana appears to be counting an improbably large number of enrollees as having little or no income, which makes their premium amount easier to calculate and collect.
Indiana reports that over 175,000 HIP 2.0 enrollees have incomes below 5 percent of the poverty line.[3] This is extremely implausible, given that Census data show only about 94,000 non-elderly adults in Indiana have incomes this low. Moreover, while Indiana reports that these very low- income enrollees make up over half of HIP 2.0 enrollees, Census data show that only 12 percent of non-elderly adults with incomes below the HIP 2.0 limit of 138 percent have incomes below 5 percent of poverty.[4]
One possible explanation of this highly unlikely enrollment pattern is that it is easier to collect a fixed premium of $1 per month (the amount charged for people with incomes under $50 a month) than a premium set as a percentage of income, which changes as income rises or falls.
The premiums charged in HIP 2.0 — 2 percent of income for people with incomes below 138 percent of the poverty line — are modeled on the amounts that enrollees in marketplace coverage with similar incomes are expected to contribute. However, contributions from marketplace enrollees are based on their annual income, while premiums in HIP 2.0 are based on a beneficiary’s current income, which can change from month to month.[5] Indiana must ensure that HIP 2.0 enrollees don’t pay more than 5 percent of their income for premiums and cost-sharing on a monthly or quarterly basis and must allow beneficiaries to adjust their contributions when their income drops. The managed care organizations that provide health care services to HIP 2.0 enrollees collect the premiums, so they must adjust the amounts they bill beneficiaries each time their income changes — a far more complex task than simply charging a flat $1 a month.
Governor Mike Pence recently wrote that “nearly 70 percent of enrollees have made contributions to their health savings accounts” and argued that programs like HIP 2.0 should be the model for states going forward.[6] But if large numbers of beneficiaries are being charged $1 a month instead of a higher amount based on their actual income, participation in HIP Plus and in HIP 2.0 as a whole is likely higher than if the state were implementing the program in accordance with its agreement with CMS. Thus, Indiana’s experience may not provide a true test of how premiums affect participation in health coverage.
Third parties are allowed to contribute to POWER accounts. Indiana reports on the number of individuals whose premiums are paid by employers and non-profit organizations, but not by health care providers or other third parties. The most recent quarterly report, covering November 2015-January 2016, said that only 32 enrollees had their premiums paid by their employers and 1,054 enrollees had their premiums paid by non-profit organizations.[7]
However, the state’s most recent quarterly report to CMS acknowledges that “some informal arrangements or payments on behalf of members” — from hospitals, friends, and family, for example — “may not be included in these numbers.” And in a recent news story, the director of an Indiana advocacy group familiar with how HIP 2.0 works said that “payment help is often off the grid.”[8] If third parties are paying premiums for a large share of beneficiaries and this is not being reported to CMS, the demonstration is not a true test of the impact of premiums on enrollment and utilization of services.
Indiana’s waiver also seeks to test whether “presumptive eligibility and fast-track prepayments will provide the necessary coverage so as not to have gaps in health care coverage.” Presumptive eligibility allows health care providers and other groups authorized as “qualified entities” to screen patients or clients for Medicaid eligibility and immediately enroll individuals who appear eligible for a temporary period while they complete the eligibility process.
CMS required Indiana to implement a robust presumptive eligibility process to mitigate the potential harm from delaying Medicaid coverage until beneficiaries paid their premiums (or were shifted to HIP Basic after 60 days if they had incomes below the poverty line and didn’t make a premium payment). Under presumptive eligibility, people whom providers screen as eligible for HIP 2.0 can obtain needed health care before paying their premiums or waiting 60 days.
From February through September of 2015, however, only 11.5 percent of beneficiaries approved for HIP 2.0 had a presumptive eligibility period prior to enrollment. Going forward, the state wants the standard to be 8 percent, which means the vast majority of beneficiaries would have their coverage delayed until they paid a premium or waited 60 days.[9]
Indiana claims to have reached out to all providers that could participate in presumptive eligibility as “qualified entities.” Yet only 41 percent participated in the quarter ending in January 2016, a decline from the previous quarter, which raises questions about how Indiana is implementing presumptive eligibility.[10]
Fast-track prepayments are $10 payments made upon application and before eligibility is determined that allow individuals found eligible for coverage to have it start on the first day of the month they are found eligible. But Indiana’s monthly and quarterly reports lack data on the number of individuals making prepayments, so it isn’t clear whether the ability to make these payments mitigates harm from delays in coverage.
Under the Medicaid statute, states can’t charge premiums to beneficiaries with incomes below 150 percent of the poverty line and can require only nominal cost-sharing for beneficiaries with incomes below poverty. There’s a good reason for this: Medicaid beneficiaries have very low incomes and often can’t afford to pay for care.
A robust body of research, including evidence from Medicaid waivers in effect before health reform, shows that premiums have harmful impacts on individuals with low incomes.[11] Enrollment in Oregon’s waiver program dropped by almost half after the state gained CMS approval in 2003 to test raising premiums. Similarly, after CMS allowed Utah to impose a $50 annual enrollment fee on adult Medicaid beneficiaries, one-third of those who lost coverage cited financial barriers as the reason for not re-enrolling.[12] People with incomes below half of the poverty line are especially sensitive to even modest changes in out-of-pocket costs, a recent report from the Department of Health and Human Services shows.[13]
CMS has allowed five of the six states with a Medicaid expansion waiver (all except New Hampshire) to charge premiums to a subset of the expansion population. Now Arizona, Ohio, and Kentucky — which have had very successful Medicaid expansions in terms of health coverage gains and state budget savings — are looking to switch from a straight Medicaid expansion to a waiver. Their proposals borrow certain elements from Indiana’s waiver:
- Arizona’s December 2015 proposal to CMS would allow the state to give each beneficiary an HSA-like account into which premiums and co-payment obligations would be deposited. It also would allow the state to impose a work requirement on beneficiaries and a five-year lifetime limit on coverage for non-disabled adults, neither of which Medicaid has ever allowed.
- Ohio will release a proposal on April 15 that borrows elements from Indiana, such as HSA-like accounts and premiums for people living in poverty.
- Kentucky’s new Governor Matt Bevin said at his inauguration, “I intend to copy the best parts of what they [Indiana] are doing” in Medicaid.[14] Bevin has also spoken of requiring more “personal responsibility,” likely through premiums and increased cost-sharing. Kentucky is likely to release its waiver proposal later this year.
Evaluations by Indiana and other states charging premiums are testing hypotheses regarding the impact of premiums on participation and use of health care services, and CMS has commissioned a cross-state evaluation to examine the impact of premiums on take-up of coverage.[15] CMS has also commissioned its own evaluation of HIP 2.0, including a survey of current and former beneficiaries about their experiences paying premiums and obtaining care; a report on the survey is scheduled to be completed by the end of 2016.[16]
Indiana’s experience in implementing its premiums shows that conclusions can’t be drawn just from enrollment data showing participation rates in HIP Plus and HIP Basic. Other information also is essential. Results from the federal and state evaluations should shed light on how the state is determining premium amounts, the extent of third-party payments, and whether coverage delays are resulting in unmet needs for health care. CMS should wait for these results before approving proposals to replicate features of the Indiana waiver, particularly from states that have already expanded Medicaid.
In a state that hasn’t expanded, a demonstration project allowing premiums could arguably increase coverage for previously uninsured low-income adults, although likely by less than a traditional Medicaid expansion would. But allowing states that have already expanded to start charging premiums and eliminating coverage for people who don’t pay them — before results are in from Indiana (and other waiver states) — poses unacceptable risks of weakening coverage for low-income people.[17] Such proposals would fail to satisfy the criteria CMS developed to guide the Secretary’s approval of demonstration projects.[18]
