Superwonk James Capretta has a must-read post at NRO on CBO's recent analysis of health care reform's probable consequences for insurance premiums. Capretta makes two key points. First, even though the CBO report contains data helpful to Senate Democrats, it also undercuts a recent study by liberal economist Jonathan Gruber at MIT, who argues that Obamacare would lower the cost of premiums on the individual market. CBO says otherwise. The administration could back the Gruber study, or it could embrace the CBO. But it would be hard for the White House to do both. Second, Capretta continues, "it appears that none of most-cited models used to estimate the impact of health-reform plans, including CBO's, has an explicit capacity to calibrate insurance take-up rates based on the penalties imposed on those who go without coverage. Apparently, the premium estimates are based as much on judgment as analytics, and CBO's judgment is clearly on the optimistic side. But what if they are wrong?" Translation: When it conducted its analysis, the CBO thought that the glass was half-full. But it always could turn out to be half-empty. We really don't know. And for that reason, it's worth exercising some epistemological modesty when discussing the potential benefits -- and costs -- of health care reform. Megan McArdle has more here.