On February 27 the appointment of Keith Hall as the new director of the Congressional Budget Office was announced. Although we overlapped only briefly as scholars at the Mercatus Center, I know Hall as a serious research economist, a good administrator, and a cool, fair, and open-minded analyst. His prospects for success in this new and challenging role are excellent. I extend to him my congratulations and best wishes. In this regard, I am sharing here evidence on the need for improvements in CBO scoring and analysis; constructive suggestions for ways to achieve more accuracy and transparency; and how government resources can be more effectively used and extended for the CBO’s work. 

A. Three Instances of CBO Errors 

The CBO produces a large volume of regular and one-time studies, reports, and cost estimates (scores), on a wide range of topics, in response to continual congressional legislative demands and interests, often on tight deadlines and in difficult political situations. That its output is nonetheless readable and perceived as balanced is therefore a great accomplishment. 

However, that does not say that the work of the CBO is always correct or even necessarily well done. Of course, it would be easy to show that the CBO’s annual macroeconomic forecasts are wrong — all such forecasts will be wrong soon after their publication because the global economy is so complex, and unexpected conditions and situations always arise. It is a cheap shot to criticize the CBO on that basis alone. But still it is plausible and fair to claim that particular CBO scores or analyses are in error at the time of publication because they ignored relevant and available data or information; their authors did not think deeply or creatively about the context or effects of the proposed legislation; or they hued too literally to the mainstream viewpoint despite evidence pointing to contrary outcomes. Moreover, because the CBO is central to the legislative process, its work, especially the scoring of legislative proposals in critical areas, is very important — and therefore, so are the consequences of errors. It is essential to get everything right. 

I myself have found three recent instances of CBO errors, not because I was looking for them, but because I was researching the relevant topics — income inequality, long-term care insurance, and employer-provided health insurance — and came across the relevant CBO analysis or score, either contemporaneously or not long after it was published. Although not completely random picks, my negative reviews are a somewhat worrisome indicator of the overall quality of CBO work. To learn whether my inference of more widespread inadequacies is correct, or in any case that an overdue external audit of the CBO is needed, I recommend that a more complete statistical sample of recent CBO scores and analyses should be examined methodically, comprehensively, and thoroughly by a committee of external experts that includes economists and accountants. This committee would include both academics and practitioners who use a wide range of methodological approaches and have different viewpoints. The installation of a new CBO director is an excellent time to embark on this comprehensive review. 

Continue reading