I hope people read Lori Montgomery’s article taking a look behind the curtain at the Congressional Budget Office and realize that if you think there’s really no way to know for sure what the long-run budgetary impact of a complicated series of health policy reforms would be, you’re right. The CBO employs a lot of smart people, and they employ a lot of smart computer models, but forecasting the future is hard to do.
And I think the CBO’s role as a budget scorekeeper prevents it from doing what I would like to see from a general forecaster and issuing predictions in terms of a range of probabilities rather than a number. I think this probably becomes a bigger problem in the realm of climate change than in the realm of health care. But our climate science models, though pretty good, involve a fair amount of uncertainty. It’s possible, as a result, that things will turn out to be much worse than the modal prediction of the models. Importantly, the downside risk is basically unbounded while the upside risk is pretty sharply bounded. Conversely, when it comes to adapting ourselves to a low carbon economy the situation is reversed. There’s a sharp bound on how economically difficult such a transition could possibly be, while the sky’s potentially the limit in terms of possible future technologies for energy efficiency and clean energy. The result is an account that’s probably too conservative about how bad business as usual might get and too pessimistic about the costs of changing things.
As a matter of budgetary accounting, I can’t think of any better way to do it. But climate change isn’t primarily an issue about budget forecasts and it’s important for Congress to make policy based on its best possible overall understanding of the issues.
A related, but much more specific, issue is the subject of this great piece by Dave Roberts. It turns out that the CBO has a fairly arbitrary scoring rule that makes spending cap-and-trade revenue on energy efficiency programs look much worse than spending it on just handing the money to whomever.