He spoke at the Wilson School. The audio sounds bad on my computer, but he makes an important point. If health care costs are 16 percent of GDP, and we could reduce spending by 30 percent with no worse outcomes, that saves 5 percent of GDP. There is no economic policy change in any other area that has the potential to even come close to that.
Orszag argues, as others do, that because health spending is highly concentrated, catastrophic health insurance will not reduce spending very much. The point is that for very sick people, catastrophic coverage is like comprehensive coverage. This in turn suggests that there is not much to be gained by trying to tinker with insurance incentives.
I tend to disagree. First, I think that having a multi-year term for catastrophic policies, as sketched out in my book (soon to be out in paperback, by the way), would mean that a much larger percentage of health care spending would be subject to incentives than the typical analysis would suggest. Second, it is possible to restructure insurance to use co-payments rather than deductibles as the main mechanism for cost sharing. Third, if we were to phase out Medicare and replace it with actuarially fair catastrophic insurance, the remaining lifetime deductible (again, this is discussed in my book) would be so high that even high-spending consumers would still be responsible for much of their spending.
On balance, however, I am in the same camp as Orszag. Our camp believes that the United States could reduce health care spending substantially without hurting health care outcomes. However, the path to getting there involves lots of research into the efficacy of various procedures as well as changes in behavior (Orszag refers to physician norms as an example).
The other camp, which includes Jacob Hacker writing in today's Washington Post, says that all you have to do is socialize medicine and magically costs will come down. That is unlikely.