Sunday, March 05, 2006


WSJ Mumbo Jumbo

I previously commented on a CBO letter to Senator Charles E. Grassley that explained why the increase in capital gain revenue could not be attributed to the cuts in the tax rates on capital gains. At the conclusion of the posting, I said, "I'm waiting for the WSJ to spin [the CBO letter]." On March 2, the spin began with an editorial entitled "Non-Dynamic Duo." (Subscription required.)

Not surprisingly, the WSJ had to resort to name-calling, dubbing the detailed analysis offered by the CBO as "economic mumbo jumbo." Apparently, "economic mumbo jumbo" is a WSJ term for "technically difficult analyses that contradict the WSJ's ideological biases." At the core of the dispute is whether the CBO's and the JCT's methods of estimation are reasonably accurate. They are.

Late last week, the CBO issued another study, The Uncertainty of Budget Projections: A Discussion of Data and Methods. That study showed that, over the long haul, CBO estimates have generally been on the mark. Thus we have this graph which shows that, generally, the CBO estimates of budgetary surpluses or deficits have been within their 90% confidence range:

(Click for larger image.)

I could respond in kind to the WSJ's editorial by claiming that it's written by a bunch of knaves appealing to the baser instincts of a bunch of fools. But that would be name calling and I shouldn't stoop that low. Oh, what the hell.

1 comment:

Anonymous said...

I'd be all for the use of dynamic models in forcasting the "cost" of tax policies if someone could show me an accurate dynamic model. No such creature exists. I have no idea why these people calling for dynamic scoring think dynamic models are any more accurate than static anayisis.

[Actually, I doubt they think they are more accurate, but they know that you can get a dynamic model to say whatever you want by adjusting the variables. Want a $250 million cut to show an increase in tax revenues? Just tweak the labor supply response in the model. There. See? We can cut taxes and increase revenue. This dynamic model proves it.]