What Medicare Part D Teaches Us About Social Insurance Markets

Summary

Seeking to reduce costs and increase the quality of services, policymakers have recently transferred the provision of certain traditional government social services to private firms. The central economic problem to be solved in this arrangement is how to structure payments to these firms and reduce their ability to capture rents from subsidies intended for consumers.

This paper illustrates the complex interplay between subsidy payments, market power, and equilibrium outcomes in the setting of prescription drug coverage in Medicare Part D, an example of such a market. These findings can apply to other privately-provided social insurance markets and highlight three key features of efficient subsidy mechanisms.

It is important to preserve the marginal relationship between the prices firms set and the prices consumers pay, which result in more intense competition. This affects firms’ abilities to raise prices and limits profits, which is significant since one of the motivations for implementing private provision of publicly-subsidized goods is that competition could deliver lower prices to consumers. The complexity of the Part D market subsidy accomplishes this by effectively acting like a voucher. Consumer-facing prices should be positively related to the social cost of providing those services. Because the Part D subsidy acts like a voucher, it preserves the marginal relationship between bids and consumer prices. The relationship between subsidies and equilibrium outcomes needs to be tempered to prevent strategic pricing by imperfectly competitive firms. This limits the ability of firms to capture an ever-increasing subsidy from the government by simply increasing prices. The Part D market achieves this by limiting the size of the subsidy so it is not so large that all consumer-facing premiums are zero.

Main article

As governments move away from directly administering public programs in favor of provision by private firms, policy discussions related to how to structure payments to such firms have become increasingly important. If incentives are properly aligned, privately provided social insurance programs may have higher quality and lower costs than traditional government-provided services. This research informs this discussion by illustrating the complexity of economic forces that policymakers must balance in the context of elective prescription drug coverage under Medicare. Efficient subsidy mechanisms share three features: firms have limited power to increase subsidies through price increases; consumers are exposed to marginal price increases; and prices are linked closely to the cost of their provision.

Social insurance programs, such as health insurance and social security, have traditionally been paid for and provided by the government. However, more recently, there have been a number of high-profile initiatives to replace government-provided services with private provision via regulated competition. The motivation for these programs is if incentives are set up in the right way, private firms may provide higher-quality services at lower cost to consumers and the government alike. A broad range of services have moved to this model, ranging from health insurance in the United States to the social security system in Chile, and have been considered in the context of social security, disability, and unemployment insurance systems in many OECD countries.

The central economic problem in this arrangement is how to structure payments to these firms? Policy makers have to confront the possibility that firms would exert market power to capture rents from subsidies intended for consumers, nullifying any efficiency gains the government hoped to achieve from private provision. This paper illustrates the complex interplay between subsidy payments, market power, and equilibrium outcomes in the setting of prescription drug coverage in Medicare Part D, an elective pharmaceutical insurance program in the United States primarily for individuals older than 65. The findings from this study can be relevant to other privately-provided social insurance markets. The authors conclude that the Part D market is relatively efficient and features three key mechanisms, which are generally applicable to other instances of regulated and subsidized competition:

  1. Firms have limited ability to increase the subsidy level through price increases – while the Part D subsidy mechanism is complex, it effectively acts like a flat voucher, limiting any strategic pricing power of firms.
  2. Consumers are exposed to price increases at the margin – the size of the subsidy is not so large that all consumer-facing premiums are zero, which helps keep demand responsive to price increases.
  3. Prices are positively linked to their underlying marginal costs – because it acts like a fixed voucher, the Part D subsidy preserves the price-cost relationship.

Historically, the government would directly provided a publicly funded good. In a regulated competition setting, the primary policy consideration is how funding should flow from the government to firms and consumers. This question is complicated by the presence of market power, which can lead to firms increasing their prices in response to subsidies intended for consumers. This could lead to both higher consumer prices and government payments to firm profits – both of which are antithetical to the motivation for such programs.

This research provides insight into the underlying economic forces by studying a specific program: the Medicare Part D prescription drug insurance. Medicare is a health insurance program primarily available to adults older than 65, and Part D is an elective prescription drug insurance component available to Medicare beneficiaries. Part D has become a model for privately-provided, publicly-financed social insurance programs in the United States. Using institutional details and data from the Part D market, this paper provides evidence on two dimensions related to the subsidy design question: how do the moving parts of the Part D subsidy mechanism interact to generate equilibrium outcomes, and what might happen under other counterfactual subsidy mechanisms?

The Part D program has several features making it worthy of study. Importantly, it has clear, well-articulated rules that allow the incentives facing firms to be modelled. Additionally, there are quality data on potential consumers and the choices available to them as well as a complex mechanism linking equilibrium market outcomes to consumer-facing plan prices and public subsidies. This research focuses on enrollees that are actively choosing plans but also models the complexities created by the presence of additional means-tested Low-Income Subsidies (LIS).

To assess how the rules of the regulated competition interact with consumer and firm incentives, the authors estimated consumer demand for Part D Prescription drug plans (PDPs. Insurance markets differ from typical product markets in that costs to the seller may vary across consumers due to differences in underlying health conditions.

To allow for this, the authors estimated demand after classifying consumers into six categories based on income and health risk. There is adverse selection if consumers that value insurance highly are also costlier to insure due to worse health conditions. Conversely, there is advantageous selection if consumers that value insurance highly are the least costly to insure due to a healthier lifestyle. Past studies in Medicare Part D have documented evidence of adverse selection.

The first years of the Part D program revealed adverse selection, as high-risk individuals showed the highest demand for the most generous insurance products. Due to its high cost, this type of plan was soon eliminated by all insurers. It is critical to account for this selection, as the set of consumers that sign-up for a plan is driven by price changes, which in turn, changes the costs of plans. Failing to model the interdependency between demand and costs may provide misleading predictions about what would happen in alternative subsidy environments.

With the demand estimates in hand, the authors then impose a structural model of differentiated product competition in prices to recover marginal costs for each plan by consumer risk type. These costs are the fundamental supply-side primitives in the model and allow the authors to predict the prices that firms would want to set under counterfactual subsidy mechanisms. After estimating demand and costs, several empirical facts become clear:

Firms have limited ability to increase the subsidy level through price increases

In the Part D market, consumers have a low willingness-to-pay for insurance products, which is not surprising given there is a closely-aligned and heavily-subsidized alternative available via Medicare Advantage. Private Medicare Advantage plans provide a comprehensive substitute to the federal Medicare program and cover both medical services and prescription drugs.

Estimates of welfare are dominated by the amount of money that the government can save by shifting consumers away from even more expensive external options. This is an important element of the model, as optimal subsidy policies induce sorting of consumers along two margins: which consumers should purchase a PDP and which plan within Part D is optimal.

Consumers are exposed to price increases at the margin.

In the Part D market, firms make modest amounts of profit. One of the key features of the subsidy mechanism is that consumers are still exposed to price increases at the margin. This affects firms’ abilities to raise prices and limits profits, which is significant since one of the motivations for implementing private provision of publicly-subsidized goods is that competition could deliver lower prices to consumers.

Prices are positively linked to their underlying marginal costs.

After parsing the institutional details in the regulatory rules, the actual subsidy mechanism operates very closely to a fixed voucher. While there is language in the subsidy mechanism that suggests an auction is happening, the actual economic mechanism sets a fixed amount of money that consumers can spend on plans within the market.

Having established these facts about the market as it works today, the authors use their model to simulate what could happen with prices and enrollment in this insurance market under under alternative subsidy rules. They considered two alternative subsidy mechanisms: a straight voucher and a proportional discount on the plan premium. While the latter mechanisms underperform the existing mechanism, the flat voucher can improve it, although not by much, supporting the idea that the existing mechanism is already quite efficient.

To understand how much any practical subsidy mechanism is leaving on the table relative to the best possible (but likely umpractical) mechanism, the authors also simulate the“social planner’s problem.” The social planner finds a set plan-specific prices that maximizethe sum of consumer well-being and producer profit while accounting for the cost of raising government dollars through taxes. The authors estimate that a social planner does significantly better than the optimal flat voucher by adjusting plan-level prices so that enrollees sort into the most socially desirable plans. In this allocation, overall enrollment into the program also increases.

The authors also simulate a scenario that  replaces all of the choices in a market with the plan that has the lowest estimated marginal cost. In this scenario, which parallels the idea of a “public option,” the government forces the monopoly firm to sell at marginal cost. This arrangement is close to the optimal, but requires the government to know which firm could provide insurance at the lowest cost. It is possible, that such information could be elicited through an auction, although whether insurers would want to participate in such an auction is less clear.

While estimates are specific to the context of Part D, these results suggest key economic forces that are likely to be important in any setting with publicly-subsidized, privately-provisioned goods and services. It is important to preserve the marginal relationship between the prices that firms set and what consumers pay, which results in more intense competition and lower prices. Consumer-facing prices should be positively related to the social cost of providing those services. The relationship between subsidies and equilibrium outcomes needs to be tempered to prevent strategic pricing by imperfectly competitive firms. Public programs are often interlinked, and in order to properly assess the value of a program, one needs to understand how, absent this program, beneficiaries would modify their behavior in the other connected public programs.

For policymakers tasked with designing such programs, understanding these economic forces is paramount to implementing an efficient and successful social insurance program via a regulated market.

This article summarizes ‘Subsidy design in privately-provided social insurance: lessons from Medicare Part D’ by Francesco Decarolis, Maria Polyakova and Stephen P. Ryan published in The Journal Of Political Economy in May 2020 by University of Chicago Press.

Francesco Decarolis is at Bocconi University and EIEF, Maria Polyakova is at Stanford University and NBER, and Stephen P. Ryan is at Washington University and NBER.

Further reading

Decarolis, F. (2015), “Medicare Part D: Are Insurers Gaming the Low Income Subsidy Design?,” American Economic Review, April 2015, 105(4): pp 1547-80.

Einav, L., Finkelstein A., and Levin, J. (2010), "Beyond Testing: Empirical Models of Insurance Markets," Annual Review of Economics, Annual Reviews, vol. 2(1), pp 311-336.

Einav, L., Finkelstein A., and Tebaldi, P. (2020) “Market Design in Regulated Health Insurance Markets: Risk Adjustment vs. Subsidies,” working paper.

Polyakova, M. and Ryan, S. (2019), “Subsidy Targeting with Market Power” NBER Working Paper No. 26367. 

Roth, A. (2002), “The Economist as Engineer: Game Theory, Experimentation and Computational Tools for Design Economics,” Econometrica, Vol. 70, No. 4 (July, 2002), pp. 1341–1378

Tebaldi, P., (2020), “Estimating Equilibrium in Health Insurance Exchanges: Price Competition and Subsidy Design under the ACA,” University of Chicago working paper.