POLICY BRIEF | MARCH, 2019 Who Benefits from Increasing Health Insurance Subsidies: Patients or Providers? By Marika Cabral KEY TAKEAWAYS Medicare is one of the pillars of the U.S. social insurance system and the primary source of health care coverage for n Medicare Advantage (MA) plans cover nearly 20 million those over 65. More than 58 million people were enrolled people — about 34 percent in Medicare in 2017. In that year, Medicare spending grew of the Medicare beneficiary to $705.9 billion, up 4.2 percent from the year before, population. representing 20 percent of total national health expenditures n Roughly half of MA subsidy (U.S. Centers for Medicare & Medicaid Services 2019). increases are passed on to Since originating in 1965 as a federal government insurance program, beneficiaries in the form of Medicare has increasingly become privatized. So have other publicly funded lower premiums or improved health insurance programs, such as Medicaid, which provides health care benefits. coverage to low-income individuals. Medicare’s private plan option, known as Medicare Advantage (MA), is based on a system of subsidies to private n Pass-through of subsidy insurance providers designed to compensate them for bearing the risks and payments to MA beneficiaries costs of managing health care for seniors. In addition to collecting these varies greatly, ranging government payments, insurers often charge enrollees a premium. from 13 percent in the least competitive markets In most of the United States, Medicare beneficiaries can choose between the to 74 percent in the most traditional fee-for-service program, in which the federal government pays competitive. health care providers directly, or a private MA plan. Enrollment in MA plans has roughly doubled over the past decade. By 2018, MA covered nearly 20 million n As many MA markets are people, or about 34 percent of the Medicare beneficiary population. highly concentrated, limited pass-through of subsidies is a pressing concern for seniors Do Medicare Advantage Subsidies Go to Plan Members who depend on Medicare for or Providers? health care coverage. MA subsidies to private insurers have been adjusted up or down a number of times in the history of the program. Most recently, the Affordable Care Act included an estimated $156 billion reduction in MA subsidies (Kaiser Family John A. and Cynthia Fry Gunn Building siepr.stanford.edu 366 Galvez Street, Stanford, CA 94305-6015 @siepr facebook.com/SIEPR/ 1 POLICY BRIEF | MARCH, 2019 Foundation 2010). These adjustments have spurred We found that MA plans passed along a little over half a sharp debate. If subsidies are raised, how much of their capitation payment increases to beneficiaries in the these higher payments are passed through to Medicare form of lower premiums or improved benefits. enrollees in the form of lower premiums and more For each dollar in higher payments, we estimate MA generous benefits and how much goes to health care premiums fell about 45 cents in the three years following providers, including insurers, doctors, and hospitals? the change. In addition, an estimated 9 cents went to MA Similarly, when subsidies are trimmed, how much enrollees in the form of more-generous benefits, such as do providers take a hit and how much are premiums lower medical co-payments and added coverage. raised or benefits cut to keep profits intact? Simply put, do changes to MA subsidy levels impact patients or That combination of lower premiums and improved providers more? benefits indicates a total pass-through rate of 54 percent. Statistically, we have 95 percent confidence that the In a recent paper, my University of Texas colleague combined subsidy pass-through rate fell between 37 Michael Geruso, Neale Mahoney of the University of percent and 71 percent. Moreover, we found that affected Chicago Booth School of Business, and I took advantage counties and unaffected counties were following the same of a legislative overhaul that raised MA subsidies to trend before the new payment system was implemented. examine how gains were distributed (Cabral, Geruso, and This suggests that the patterns observed when BIPA was Mahoney 2018). implemented were a result of the subsidy increases. Congress enacted the Benefits Improvement and We investigated the possibility that MA plans were Protection Act (BIPA) in 2000. The law instituted far- improving coverage in ways that couldn’t easily be reaching changes in how capitation, or per-beneficiary, measured, such as offering better customer service. payments to MA providers were calculated. These However, a review of beneficiary evaluations and survey payments previously had been based largely on data yielded no evidence of changes in plan quality other historical expenditures of the traditional Medicare than the premiums and benefits we studied. program at the county level. BIPA created a system of rural and urban payment floors that boosted subsidies in 72 percent of U.S. counties. Previously, capitation Explaining Incomplete Pass-Through of payments in those counties had been on a similar track Medicare Advantage Subsidies as payments in counties unaffected by the new floors. But when the new subsidy calculation method was Under certain economic assumptions, such as a fully implemented, per-beneficiary payments rose $600 per competitive insurance market, the entire MA subsidy year on average, or 12 percent, in the counties where the increase should be passed through to beneficiaries. In floors were binding. our theoretical model, two explanations for partial pass- through are possible: These sharp payment increases represent a natural experiment allowing us to analyze how higher • When MA insurers lowered premiums, they may have subsidies were apportioned and the extent to which attracted a riskier, less-healthy pool of beneficiaries. these increases reduced MA premiums and improved This advantageous selection of beneficiaries could coverage. To estimate how much of the MA subsidy lead to higher health care costs, which would prompt increases went to beneficiaries, we used an empirical insurers to adjust premiums upward to recover these strategy known as difference-in-differences to compare additional expenses. Thus, even in a highly competitive changes in counties affected by the new floors with market, insurers would not be able to fully pass changes in unaffected counties. through MA subsidies. John A. and Cynthia Fry Gunn Building siepr.stanford.edu 366 Galvez Street, Stanford, CA 94305-6015 @siepr facebook.com/SIEPR/ 2 POLICY BRIEF | MARCH, 2019 • A county MA market may not be fully competitive. A Our results do not address how the share of subsidies not high degree of concentration, that is, with one or more passed through to MA beneficiaries was divided among firms holding dominant market share, would give insurers and other health care providers (e.g., doctors, providers market power to set quasi-monopolistic hospitals). Nevertheless, the high correlation between prices and keep a large share of the subsidy increases. pass-through and insurer market concentration is a powerful indication that insurers are likely capturing the To investigate the possible impact of advantageous lion’s share of the portion of payments going to providers. selection, we analyzed how the BIPA-generated Other research supports this interpretation. Duggan, Starc, payment shock impacted enrollment in MA and insurer and Vabson (2016) found that a large MA payment increase costs. For this analysis, we use data on traditional was followed by a sharp, substantial rise in returns on the Medicare beneficiaries to estimate how the selection stocks of large MA insurers. At the same time, the share of beneficiaries into MA (and out of traditional price of the largest publicly traded hospital company was Medicare) impacted MA insurer costs. We found limited unchanged. This indicates that investors made their own advantageous selection into MA plans. In our model, judgments about where the subsidies were going. under conditions of perfect competition, increased risk in the beneficiary pool would reduce pass-through to an estimated 85 percent. In other words, of the 46 cents Can Policy Increase Pass-Through? of every dollar of increased subsidy not passed through to beneficiaries, advantageous selection accounts for Medicare provides health care to tens of millions of 15 cents, or about a third of the gap between the share Americans and there is a national interest in ensuring the passed through to consumers and full pass-through. program provides affordable, high-quality health care. Thus, when payments to private MA insurers are raised, That leaves market power as a possible explanation of one goal of policy may be to pass along to beneficiaries incomplete pass-through. We used two measures of as much of these subsidy increases as possible in the pre-BIPA competition to determine the degree of insurer form of lower premiums and improved benefits. market power in a county: (1) the Herfindahl-Hirschman Index (HHI), a standard statistical metric of competition Our research points to the important role of market in a market; (2) the number of firms offering MA plans. power in determining the extent of MA subsidy pass- Both metrics showed a strong relationship between through. Beneficiaries in more competitive markets market concentration and incomplete pass-through. benefit more from subsidy increases—and suffer more from subsidy cuts—than their counterparts in less On the first point, we divided counties into three groups competitive markets. ranging from most to least concentrated according to the insurer HHI. In the most concentrated counties, where This suggests that steps to increase market competition market power was greatest, pass-through was a mere 13 are key to boosting pass-through. We didn’t examine percent. By contrast, in the least concentrated counties, what kinds of government initiatives would make we estimated pass-through at 63 percent. Similarly, in MA markets more competitive. Put plainly, we don’t counties with just one insurer, our pass-through estimate know what works. We need to understand better was also 13 percent. But in counties with three or more what policies are effective in promoting competition. insurers, pass-through rose to 74 percent. In sum, pass- Nevertheless, even though examining specific policy through of subsidy payments to MA beneficiaries varies proposals is outside the scope of our analysis, there is greatly, ranging from 13 percent in the least competitive some indication that the number of competitors is an markets to 74 percent in the most competitive. important factor. Consequently, measures to encourage new entrants to the MA market may be worth exploring. John A. and Cynthia Fry Gunn Building siepr.stanford.edu 366 Galvez Street, Stanford, CA 94305-6015 @siepr facebook.com/SIEPR/ 3 POLICY BRIEF | MARCH, 2019 Our research is particularly relevant to one of the most contentious questions on the health policy agenda: what Marika Cabral is a SIEPR Visiting to do with the Affordable Care Act (ACA). As mentioned Fellow during the 2018-19 earlier in this policy brief, the ACA provided for $156 academic year. She is an Assistant billion in MA subsidy reductions. Our results indicate Professor at the University of Texas that these reductions have been borne only partially at Austin. Much of her research by Medicare beneficiaries, with a significant fraction of focuses on understanding the role of asymmetric the cuts falling on insurers and other providers. If the information and the impact of government ACA were repealed, higher payments to MA plans would intervention in health (and health-related) be restored. Our results indicate market power would insurance markets. be a key factor determining how these gains would be apportioned. Sam Zuckerman contributed editorial assistance to this policy brief. Our study looked at the MA program in the early 2000s, but insurance markets are still dominated by a small number of insurers. In 2014, 88 percent of MA markets had a Herfindahl-Hirschman Index measure of insurance The Stanford Institute for Economic Policy competition that met the regulatory definition of highly Research (SIEPR) catalyzes and promotes evidence- concentrated. In most parts of the United States, MA plan based knowledge about pressing economic competition is in short supply—suggesting that limited issues, leading to better-informed policy solutions pass-through of subsidies remains a pressing concern for generations to come. We are a nonpartisan for seniors who depend on Medicare for health care research institute, and SIEPR Policy Briefs reflect coverage. the views and ideas of the author only. References Cabral, Marika, Michael Geruso, and Neale Mahoney, 2018. “Do Larger Health Insurance Subsidies Benefit Patients or Producers? Evidence from Medicare Advantage.” American Economic Review, 108(8): 2048-87. Duggan, Mark, Amanda Starc, and Boris Vabson, 2016. “Who Benefits When the Government Pays More? Pass-Through in the Medicare Advantage Program.” Journal of Public Economics, 141: 50-67. The Kaiser Family Foundation, 2010. “Explaining Health Reform: Key Changes in the Medicare Advantage Program.” https://kaiserfamilyfoundation.files. wordpress.com/2013/01/8071.pdf U.S. Centers for Medicare & Medicaid Services, 2019. “Historical NHE, 2017.” https://www.cms.gov/research-statistics-data-and-systems/statistics- trends-and-reports/nationalhealthexpenddata/nhe-fact-sheet.html John A. and Cynthia Fry Gunn Building siepr.stanford.edu 366 Galvez Street, Stanford, CA 94305-6015 @siepr facebook.com/SIEPR/ 4