In a reply to David Leonhardt of the New York Times, my Economics 21 colleague Christopher Papagianis raises the following interesting point regarding Social Security:

[T]he source of our entitlement problems is not aging so much as the automatic increases in real (inflation adjusted) benefits. For example, the effective Social Security tax rate equals the wage replacement rate divided by worker-to-beneficiary ratio. Today, an average 42% wage replacement is financed by a 3.4 worker-to-beneficiary ratio, which works out to about a 12% payroll tax. When the ratio drops to 2 workers per beneficiary, we can’t maintain the replacement without increasing taxes to 21%.
Much the same is true with Medicare, except that program also suffers from the very poor design of its payment system – an open-ended fee-for-service design that incentivizes more, not less, health care spending.

Later in the piece, Papagianis addresses criticisms of House Budget Chairman Paul Ryan’s Medicare reform proposal:

Most critics of Mr. Ryan’s plan have focused on this last point – the indexation of the credits to CPI – arguing that this is unfair because health care costs are projected to increase at a faster clip. As Jim Capretta has argued, “this is sheer hypocrisy on their part because the indexing of government-financed premium credits below cost growth is in the president’s plan too, and yet not a complaint has been heard about that from its advocates.”
If the government’s costs for premiums and cost-sharing subsidies exceed roughly .5% of GDP after 2018, the new health care law requires the government’s contributions toward coverage to rise with GDP growth. The bottom-line is that critics of Mr. Ryan’s plan contend that the premium support credits will not keep pace with expectations of rising health costs – but this is the same problem faced under Obama’s 2010 health law.

Loren Adler and Shai Akabas have summarized the growth rate in Medicare across three proposals, Domenici-Rivlin, the Ryan proposal, and PPACA. Domenici-Rivlin grows premium support at GDP +1, or roughly 4.3% per year; Ryan grows premium support at CPI-U, or roughly 2.3% per year; and PPACA directs IPAB to take action if Medicare cost growth exceeds 3.8% per year.

James Capretta, of course, was taking about premium support credits on the exchanges for people not yet eligible for Medicare:

Ryan’s critics have focused particular attention on his plan’s indexation of the Medicare “premium support credits” to the CPI in the years after 2022, suggesting that this idea is somehow beyond the pale. But this is sheer hypocrisy on their part because the indexing of government-financed premium credits below cost growth is in the president’’ plan too, and yet not a complaint has been heard about that from its advocates. That’s right. After 2018, if the aggregate governmental cost of premium credits and cost-sharing subsidies provided in the state-run exchanges exceeds about 0.5 percent of GDP (a condition that the Congressional Budget Office says will be met), the recently-enacted health law requires the government’s per capita contribution to health plan premiums in the exchanges to rise more slowly than premiums. The administration actuaries interpret the law to mean that the government’s contributions toward coverage will rise with GDP growth after 2018. CBO appears to have a different interpretation. Still, under all interpretations and projections, it’s clear that the exchange credits in the new law will not keep pace with expectations of rising health costs. And that’s exactly what the president is now saying is so wrong with Ryan’s Medicare plan.

One rejoinder from advocates of the president’s approach is that Medicare should be preserved as a defined benefit while the new health entitlement should be a defined contribution. In the case of Medicare, we’re talking about maintaining the status quo, albeit at crippling expense. In creating a new program from scratch, perhaps it makes sense to be somewhat less generous, hence one can maintain a meaningful, non-hypocritical distinction.

My own view is that Capretta is right to favor a defined contribution approach to Medicare. Yet I also think that the Domenici-Rivlin rate of premium support growth might be the better way to go.