In March 25, the Congressional Budget Office released an important study of President Bush’s budget proposal. What was novel about this study is that the CBO attempted to calculate the impact of the proposal on the economy as a whole. Normally, it assumes that even large changes in taxing and spending will have no effect whatever on such things as the unemployment rate, real growth in the economy, inflation, and interest rates, among other things.This method of analysis has long frustrated advocates of supply-side tax cuts. They believe that holding the economy constant when calculating the effects of such tax changes exaggerates their budgetary cost, thereby decreasing their chances of being enacted by Congress. After all, the whole point of a supply-side tax cut is to increase economic growth by stimulating work, saving, and investment.
Some people believe that the inclusion of macroeconomic effects in revenue estimates is some kind of trick to make tax cuts appear costless. It is often alleged that Ronald Reagan played such a trick on the American people in 1981 by saying that the big tax cut that year would not reduce federal revenue. This is nonsense. The Reagan administration always said that the 1981 tax cut would lose large revenues and its estimates were comparable to those made by independent analysts.
Furthermore, supply-side economists who made private estimates of the revenue impact of the Reagan tax cut — estimates that did incorporate growth effects — also showed large revenue losses. For example, an estimate by economist Norman Ture of an early version of the Reagan plan showed large net revenue losses even 10 years after enactment. Economist Michael Evans came to similar conclusions.
What supply-siders always said is that the Reagan tax cut would not lose as much revenue as conventional (static) estimates predict. Economist Lawrence Lindsey, then at Harvard, concluded that when all was said and done the net revenue loss from the 1981 tax cut was about a third less than official estimates predicted. A CBO study found that it was about 25% less.
Supply-siders believe that a dynamic analysis of President Bush’s tax plan would show approximately the same thing — that the net revenue loss will be between 25% and 33% less than a static estimate would show.
Unfortunately, because of political and bureaucratic resistance, techniques for calculating the dynamic effect of tax cuts are not very far along. So the new CBO effort must be viewed as very preliminary. Nevertheless, it does support basic supply-side theory. It shows that marginal tax-rate reductions will increase aggregate hours worked, and that elimination of the double taxation of corporate profits will increase investment and productivity.
Because supply-side theory is not yet well understood at the CBO, some of the mathematical models used to estimate the effects of President Bush’s proposal are rather primitive. And some of the assumptions that the CBO made are quite unrealistic. However, under the leadership of its newly appointed director, economist Douglas Holtz-Eakin, I expect that the analysis will improve with time.
Consequently, the best way of measuring supply-side effects may be with commercial econometric models. Corporations use them to calculate the impact on sales of changes in economic growth, interest rates, and other economic variables. These models are based on past economic data and assume that people will behave in the future as they have in the past to changes in economic conditions.
CBO used two commercial models to look at the Bush plan. The best known of them, the Global Insight model (formerly the DRI-WEFA model) showed continuing positive growth effects from the tax cut. A larger economy recoups about 30% of the static revenue loss, it estimates, which sounds about right to me.
One problem with the CBO analysis is that it looked at all provisions of the president’s budget, including higher spending and those tax cuts that clearly will have no growth effect. The higher spending retards growth, while the non-supply-side tax cuts inflate the revenue loss without producing any economic feedback. If the analysis were limited only to the supply-side features of the tax plan, all of the models would show strong growth effects.
In the future, I hope the CBO will do more to break out specific provisions of tax-and-spending proposals and look at them individually. This will show that some tax cuts significantly raise growth, while others have no impact whatsoever. Hopefully, this will encourage Congress to enact more of the former and less of the latter.
In any case, the CBO is to be commended for taking the first baby step toward dynamic revenue estimating. I look forward to its future efforts.