Dynamic scoring and the power of numbers

In one of their first moves upon taking power in Congress, the Republicans instructed the Congressional Budget Office (CBO) to adopt “dynamic scoring” when evaluating the budgetary impacts of new tax legislation. This is the type of issue that only a Beltway insider could love. But the priority that Republicans put on this move indicates the high value that politicians place on controlling the numbers attached to the assessment of public policies. This is particularly true of numbers that are produced with the veneer of neutrality.

{mosads}Under dynamic scoring, the CBO would estimate changes to the macroeconomy as a result of the tax change, and incorporate those changes in their estimates of the budgetary impact of the new tax policy. Under the previous system, static scoring, the changes to the larger economy were ignored when estimating budgetary impacts. Now many policy changes will affect crucial variables such as employment, economic growth and productivity. So technically static scoring is fated to contain errors by not taking into account those changes.

The problem arises when economists are asked to correct these errors. Doing so requires making assumptions about exactly how the economy will react to policy changes. Take, for example, the policy area that Republicans hope will be most helped by dynamic scoring. Conservatives want the CBO to assume that tax cuts will inevitably lead to growth and that such growth will lead to lower deficits. Unfortunately, many (not all) economists disagree with this premise unless the tax cuts are targeted at those at the bottom of the economic spectrum. Gov. Sam Brownback, who cut taxes in Kansas, has also learned this lesson the hard way.

Economics is not a science. It is a social science and to reach conclusions, it inevitably relies on assumptions about human behavior. When a neutral body like the CBO produces economic estimates however, the public often ascribe to them the status of scientific fact. No matter how much the good folks at CBO qualify those numbers, people will see them as much more certain than they really are, particularly if politicians take those numbers out of context and use them to advance policy goals.

And unfortunately, there is every reason to believe that this usage is what Republicans who support dynamic scoring have in mind. Why else would they require dynamic scoring for tax changes, but not for spending? If dynamic scoring is the correct approach to assessing budgetary impacts, then it should be correct for all policy changes. For example, recent evidence has shown that Medicaid spending on children may pay for itself over the long run (by ensuring that the children grow up into tax-paying adults). True, these results are not conclusive, but they are certainly at least as conclusive as the hypothesis that tax cuts will increase tax revenue.

At best, the change to dynamic scoring will affect a very small number of bills each year. At worst, it will demean the reputation of an excellent institution as politicians use the resulting numbers to score political points. Static scoring is a humble approach to public policy analysis. It acknowledges that there are limits to what we can predict. The implementation of dynamic scoring (particularly in the one-sided manner in which it has been done) ignores those limits and instructs CBO to make predictions in areas where there is great uncertainty. Republicans have done this fully aware that when the predictions are made, the uncertainty will be ignored, and they will have powerful numbers to use in political battles.

Shapiro is an associate professor and director of the Public Policy Program at Rutgers University and a member of the Scholars Strategy Network.