Upcoming Presentations by Staff of CBO’s Health Analysis Division

Presenters
Noelia Duchovny (CBO), Sean Lyons (CBO), Jared Maeda (CBO), J. Michael McWilliams (Harvard Medical School), Mark Shepard (Harvard University), and Bryan Tysinger (USC Schaeffer Center)

Moderator
Tamara Hayford (CBO)

Session
Spotlight Session: Putting Research to Work: How CBO Uses Research to Inform Policymakers

Description
CBO uses published research and discussions with experts to inform the agency's analytical work, including estimates of the budgetary effects of proposed legislation. In addition, CBO actively engages with the research community to develop the agency's evidence base and analytical methods. For example, CBO has published four blog posts in the past two years calling for new research on topics in the health space ranging from how health care providers respond to cost shocks to take-up rates and costs of treatment for complications from hepatitis C. Because CBO is often tasked with projecting the impact of new programs or policies on the federal budget, descriptive data and research on the impact of policy, health, and pricing shocks on health care service use and spending are particularly valuable to the agency. In this session, CBO's Health Analysis Division highlights three examples of research that have been particularly helpful. The session will include short presentations by study authors, discussions of how the work was used and what made it helpful by CBO analysts, and a Q&A at the end for session attendees to learn about CBO's work and what makes research particularly useful and informative.

Presenter
Eric Schulman

Session
Poster Reception

Authors
Jared Maeda (CBO) and Eric Schulman (CBO)

Abstract
This paper investigates how nominal payments to Medicare Advantage (MA) plans have changed over time and how those changes have shaped economic outcomes for plans and beneficiaries. The question is policy-relevant because of differences in the relationships between benchmarks, enrollment, and benefit generosity under the Benefits Improvement and Protection Act (BIPA) and the Affordable Care Act (ACA), which potentially contributed to inaccurate forecasts of the ACA's impact. A potential reason for the difference between projected and actual outcomes is that regulatory rulemaking shaped actual payment amounts in unexpected ways. For example, although the ACA lowered MA benchmarks, total payments increased because of simultaneous growth in risk adjustment payments, complicating efforts to measure the economic incidence of the ACA (that is, which parties ultimately benefited from or bore the burden of payment changes). Our project examines the extent to which local-level, rule-based changes in payments affected who benefited or bore the burden of the ACA payment changes, compared with changes to plans and beneficiary behavior.

Initial descriptive statistics suggest that BIPA's payment floors functioned as a progressive subsidy that primarily benefited rural, low-spending counties, whereas the ACA payment cuts were proportional to spending and concentrated in high-spending urban areas. That geographic pattern suggests that local market conditions—such as insurer competition, beneficiary health profiles, and physician practices—may significantly shape the incidence of payment reforms. Our descriptive findings so far highlight the need to further investigate how local market characteristics interact with the Centers for Medicare & Medicaid Services's (CMS's) rulemaking around plan finances to determine which parties benefit from or bear the burden of payment changes. As a next step, we intend to decompose the variation in plan payments, premiums, and benefits into components explained by observable factors (such as benchmarks, risk scores, quality bonuses, and rule-based elements in the CMS bidding tool) and unobservable factors (such as behavioral responses) to better understand the economic incidence of the ACA.

Presenter
Daria Pelech

Session
Enrollment Decision-Making in Medicare Advantage

Authors
Daria Pelech (CBO) and Margaret Kallus (UC Berkeley Economics)

Abstract
Though Medicare fee-for-service (FFS) covers many benefits, the program's cost-sharing structure can expose beneficiaries to substantial financial risk. To protect against that risk, most beneficiaries obtain additional coverage. Historically, many beneficiaries have purchased supplemental Medigap policies—private insurance plans that "wrap around" Medicare FFS and cover Medicare FFS deductibles, copayments, and coinsurances. Increasingly, however, Medicare beneficiaries have enrolled in Medicare Advantage (MA) plans, in which a private insurer covers enrollees' Medicare benefits and often restructures their cost-sharing to reduce enrollees' financial risk. As of 2024, just over half of all Medicare beneficiaries were enrolled in MA, up from less than a third a decade earlier. Because Medigap and MA fill some of the same financial needs for many beneficiaries—defraying the cost of the traditional Medicare benefit—the premiums and availability of Medigap policies likely influence beneficiaries' choices between Medicare FFS and MA. In the past decade, Medigap premiums have grown with inflation while MA premiums have largely stayed flat. The diverging affordability between those options might explain some of the increase in MA enrollment over the past 20 years.

This paper combines detailed data on Medigap premiums and Medicare enrollment to estimate the effect of Medigap premiums on MA enrollment. Specifically, we estimate how variation in Medigap premiums affects the probability that beneficiaries who are newly enrolling in Medicare choose MA. We also estimate the probability that MA enrollees leave their plans following a health shock or an increase in MA premiums. Using brokerage data on Medigap premiums, we estimate the average and minimum premiums each individual might face on the basis of their age, gender, zip code, and eligibility for an open-enrollment period. We combine that with a 10-year panel of individual Medicare enrollment data identifying beneficiaries who are new to Medicare or who disenroll from their MA plans. For enrollees who are new to Medicare, we estimate their probability of enrolling in MA on the basis of Medigap premiums and characteristics of the MA market in their area. For individuals who are enrolled in MA, we match Medicare enrollment data to external data from the Centers for Medicare & Medicaid Services using MA plan identifiers to identify which plans have experienced changes in premiums or benefits (which might induce a beneficiary to shop for a new plan) and which plans have been canceled (which requires a beneficiary to select a new plan). We also use individual-level beneficiary risk scores to identify beneficiaries who experience large health shocks, to compare their probability of leaving MA with that of other beneficiaries.

The results of this study will help us better understand Medigap's role in the increase in MA enrollment, which in turn will help improve future projections of MA enrollment growth. The results will also inform estimates of policies that might change Medigap or MA availability—such as guaranteed open-enrollment periods for those who leave their MA plans or changes in payments to MA plans.