Over the past half century, per-beneficiary real spending growth in Medicare has exceeded real per-capita gross domestic product (GDP) growth by more than 2 percentage points. It is widely agreed that this gap should not continue. In fact, the Patient Protection and Affordable Care Act (PPACA) is projected to close the gap completely, largely by reducing the rate of increase in fees paid to health plans and providers. Yet the ability to sustain the proposed fee cuts is highly uncertain. Moreover, even if the PPACA spending trajectory is realized, Medicare spending as a share of GDP will grow due to the aging of the population.

Policy makers from both parties have struggled to identify ways to reduce the fiscal burden of Medicare on the federal budget. The strategy most widely identified with the Republicans involves premium support. Details of such proposals vary (see Aaron and Frakt¹ for a critique of existing proposals), but the basic notion is that individuals would be given a fixed subsidy (voucher) that could be used to buy coverage. By setting the voucher rate, the government can define, and therefore control, the level of spending.

An alternative strategy that was incorporated into the PPACA (thus identified with the administration and Democratic efforts to reform Medicare) involves moving away from fee-for-service (FFS) payments to fixed payments per episode (bundled payment) or global payments per beneficiary. The global-payment model involves setting a fixed-payment rate to cover all of a beneficiary’s care. Many variations of global payment are being explored. Perhaps the most prominent involves the global payment made to an Accountable Care Organization (ACO), which is a provider system capable of accepting and managing financial risk.

Although the parties may be far apart on a range of issues, and these 2 strategies at first blush appear divergent, they share a number of features. Both define a federal budget contribution to Medicare. Both preserve beneficiary choice of provider, and both preserve physicians’ autonomy (and responsibility) in making clinical decisions with their patients.

Yet the 2 approaches also have important differences. Identifying those differences and tracing their consequences can help clarify the relevant issues and contribute to defining common ground. There are 3 main differences between the 2 strategies for controlling spending growth. At the heart of the differences are preferences about how to divide the responsibility and risk associated with controlling Medicare spending.

1. How Should the Fixed Payment Be Set and Updated Over Time?

The current FFS Medicare system places responsibility for controlling spending on the government, and the taxpayers bear the cost if spending rises rapidly. However, the existing FFS system hampers the ability of the government to reduce spending growth because it does not encourage accountability for total spending, and our existing system has not been able to reduce fee updates enough to slow spending growth, despite the Sustainable Growth Rate (SGR) system. Under a voucher or global-payment approach, federal spending growth depends entirely on the method for updating the voucher level or the global budget. In Paul Ryan’s original proposal, the voucher was essentially increased each year by the Consumer Price Index.² This increase is well below the historical rate of healthcare spending growth. In the more recent Ryan-Wyden plan, vouchers are updated based on competition, with a cap at 1 percentage point above GDP growth (which is greater than the forecasted spending growth under the PPACA). In existing global-payment models (such as the Pioneer ACO program), payment rates would be updated by the rate of spending growth in the FFS system. The use of FFS spending as a benchmark is problematic because (1) a policy aim is to shrink the FFS system and (2) the FFS system is viewed as a source of the spending problem in the first place.

In choosing a method of updating the value of a voucher or the level for global payments, one must balance fiscal concerns with the desire to promote development and adoption of new technology. This choice will be governed by values, the perceived potential for important medical breakthroughs, and demographic trends. Some stakeholders would like to limit total Medicare spending growth to the rate of GDP growth, thereby stabilizing the share of GDP consumed by Medicare. During a period of rapidly rising Medicare enrollment due to the aging of baby boomers, such a target would require per-beneficiary spending to grow at a rate below GDP growth. For example, holding Medicare growth to GDP growth rates would require an update in the per-beneficiary payment rate of GDP minus about 1.25 percentage points (the subtraction of 1.25 is required to offset demographic trends). Enacting such real cuts for older and disabled adults is viewed by many as undesirable. This issue highlights the fact that stabilizing the share of GDP devoted to Medicare is very challenging.

Because spending growth has historically been driven by new medical innovation, the reduction in payment increases would likely alter the nature or slow the rate of introduction of medical technology.^3-5 This could imply a slowdown in the rate of improved medical treatment (as opposed to an absolute reduction in quality of care; providers would still be paid more, but considerably below historic trends). Yet because payment is bundled—at least in the global-payment model, and maybe in the premium-support model—introduction of new technology may not be as adversely affected as some fear. If there is considerable waste in the system, the clinical and administrative flexibility in the bundledpayment arrangements gives providers incentive to capture savings from increased practice efficiencies and elimination of low-value care. A new approach to payment may shift medical innovation toward cost-reducing technologies and processes. Such savings could help finance high-value medical innovation.

Ultimately, the choice of how rapidly to allow the fixed payment to rise depends on our willingness to support the system with taxes (and bear the economic costs associated with mildly higher tax rates). Views on this issue will vary, but it is important that the system be flexible enough to allow more rapid spending growth if we are blessed with medical discoveries (eg, a cure for Alzheimer disease or diabetes) that we are willing to finance. Strategies such as competitive bidding could allow markets to set update rates. However, because competition is imperfect, a bidding system might not sufficiently slow spending growth, so a cap on public spending seems reasonable under either the bundled-payment system or the premium support system.