I think there’s no denying that the Obama administration is in some ways disappointed with what the CBO had to say yesterday about their proposal to create a beefed-up MedPAC called IMAC. As Peter Orszag explained at the end of his blog post on it, they wound up taking on a speculative long-term issue and attaching a very precise cost estimate to it without any clear reasoning:
A final note is worth underscoring. As a former CBO director, I can attest that CBO is sometimes accused of a bias toward exaggerating costs and underestimating savings. Unfortunately, parts of today’s analysis from CBO could feed that perception. For example, and without specifying precisely how the various modifications would work, CBO somehow concluded that the council could “eventually achieve annual savings equal to several percent of Medicare spending…[which] would amount to tens of billions of dollars per year after 2019.” Such savings are welcome (and rare!), but it is also the case that (for good reason) CBO has restricted itself to qualitative, not quantitative, analyses of long-term effects from legislative proposals. In providing a quantitative estimate of long-term effects without any analytical basis for doing so, CBO seems to have overstepped.
That said, it’s worth being clear that we’re dealing with two separate issues here. Barack Obama has laid out two different cost constraints on health care reform. One is that reform must be CBO-certified as deficit neutral within the 10-year scoring window. That means what however much new coverage costs will be balanced out, within the 10-year window, by either offsetting spending reductions or else new revenue. The other is that the bill must “bend the cost curve” over the long run, meaning it must slow the rate of growth in long-run health care expenditures.
These are totally different things. And the point of IMAC is to meet the second goal, not the first. And the CBO score speaks to the first goal, not the second. Both goals are important, but to learn that a particular proposal aimed at issue two doesn’t solve issue one doesn’t actually tell us much of anything.