CBO Provides New Evidence That the 2001 And 2003 Tax Cuts Have Only Modest Economic Effects and Do Not Pay For Themselves
April 13, 2007
A new analysis by the Congressional Budget Office finds that extending the 2001 and 2003 tax cuts could result in a modest increase in the number of hours that people work. However, any "dynamic" revenue gains associated with the resulting increase in wages and salaries would pale in comparison with the cost of extending the tax cuts.
Those findings should not be surprising because they are consistent with earlier dynamic analyses of tax cuts by CBO, the Joint Committee on Taxation (JCT), and even the President's own Treasury Department. Claims that tax cuts pay for themselves or that standard estimates of the cost of tax cuts are substantially overstated fly in the face of the accumulating evidence from studies such as these.
The CBO Analysis
CBO's latest analysis, a background paper on The Effect of Tax Changes on Labor Supply in CBO's Microsimulation Tax Model, issued April 12, provides extensive detail on how extending the 2001 and 2003 tax cuts would affect labor supply. It is not a complete macroeconomic policy analysis and does not take into account how the tax cuts would be financed.
In effect, the analysis puts the best possible light on how large the effects would be, because it implicitly assumes that the tax cuts will be financed with offsets that do not introduce any negative effects of their own. As CBO notes, if the tax cuts are financed through budget deficits, there would be negative effects on national saving. Those would erode and could even outweigh the positive effects included in the analysis.
CBO finds that extending the provisions of the Economic Growth and Tax Relief Reconciliation Act of 2001 (EGTRRA) and the Jobs and Growth Tax Relief Reconciliation Act of 2003 (JGTRRA), as well as extending relief from the alternative minimum tax (AMT), would increase the number of hours that Americans work by enough to raise total earnings in 2011 by about $42 billion (or 0.55 percent).
CBO estimates that the additional income tax generated by that $42 billion of earnings would offset about 2.3 percent of the static revenue loss from cutting taxes on earnings, and the additional payroll taxes collected would offset another 1.7 percent of the static revenue loss. In other words, CBO estimates that only about 4 percent of the revenue loss from cutting taxes on earnings would be offset by the additional revenues generated by an increase in hours worked and the resulting increase in earnings.
It is worth reiterating that even these modest results depend on the implicit assumption that the cost of extending the 2001 and 2003 tax cuts is paid for (by increases in other taxes, or reductions in spending, that have no adverse economic effects). CBO carefully notes that, "If the extension of those tax provisions was financed through larger deficits or smaller surpluses, the overall macroeconomic consequences would be less positive than an analysis of labor-supply effects alone would suggest."
Results from Other Dynamic Analyses
CBO's finding that tax cuts can have modestly positive impacts on economic activity, but that the "dynamic feedback" effects on revenue are likely to be small, is typical of the findings of dynamic analyses. More complete macroeconomic analyses of the President's budgetary proposals conducted earlier by CBO reach similar conclusions, as do CBO and JCT analyses of generic tax cuts.
Interestingly, last year the Treasury Department issued A Dynamic Analysis of Permanent Extension of the President's Tax Relief, to accompany the July 2006 mid-session review of the budget, and that study produced similar results. Treasury's complete macroeconomic analysis of a favorable case, in which the tax cuts were fully paid for with non-distorting offsets, produced results that would generate revenues equal to only about 10 percent of the static revenue loss. In other words, even under optimistic Administration assumptions about the tax cuts' effects and financing, 90 percent of the revenue loss would remain.
In contrast to the new CBO analysis, which focuses only on labor supply effects, last year's Treasury analysis included effects on saving and investment as well. This may be one reason why the 10 percent revenue offset reflected in the Treasury study's numbers is larger than CBO's estimate for labor supply alone of a 4 percent revenue offset, before taking into account the negative effects of any deficit financing. But even so, it throws cold water on claims that tax cuts come remotely close to paying for themselves.
CBO's latest analysis of how tax cuts affect economic behavior adds to the growing literature showing that tax cuts can have a modest positive impact on economic performance if they are paid for. But, it also adds to the growing literature showing that even under the best of circumstances, the additional revenue generated by economic activity stimulated by tax cuts offsets only a small percentage of the revenue loss that the tax cuts cause. If tax cuts are deficit financed, the net long-term economic effect could actually be negative, and the revenue losses could be larger, rather than smaller, than the static revenue estimates indicate.
 CBO states: "...if tax changes are financed through higher budget deficits or lower budget surpluses, national saving would be reduced, creating a drag on economic growth that is not accounted for in this analysis."
 This increase would be composed of a "substitution effect" from lower marginal tax rates that would raise earnings by 0.80 percent and an "income effect" from higher after-tax income that would reduce earnings by 0.25 percent.
 Jason Furman, "Treasury Dynamic Scoring Analysis Refutes Claims by Supporters of the Tax Cuts," Center on Budget and Policy Priorities, August 24, 2006.