WikiLeaks Document Release http://wikileaks.org/wiki/CRS-RS22020 February 2, 2009 Congressional Research Service Report RS22020 Dynamic Revenue Estimating: A Brief Overview Jane G. Gravelle, Government and Finance Division January 6, 2005 Abstract. Dynamic revenue estimating taking into account macroeconomic feedback effects from the economy was initiated in the 108th Congress, with the first analysis of the feedback effect provided during consideration of the 2003 tax cut. The effects varied in magnitude and direction, depending on the model used, and have not yet been incorporated into official estimates. To do so would be difficult given the lack of consen- sus about model and behavioral specifications; at present any analysis is provided to supplement official estimates. Order Code RS22020 January 6, 2005 CRS Report for Congress Received through the CRS Web Dynamic Revenue Estimating: A Brief Overview Jane G. Gravelle Senior Specialist in Economic Policy Government and Finance Division Summary http://wikileaks.org/wiki/CRS-RS22020 Dynamic revenue estimating taking into account macroeconomic feedback effects from the economy was initiated in the 108th Congress, with the first analysis of the feedback effect provided during consideration of the 2003 tax cut. The effects varied in magnitude and direction, depending on the model used, and have not yet been incorporated into official estimates. To do so would be difficult given the lack of consensus about model and behavioral specifications; at present any analysis is provided to supplement official estimates. This report will not be updated. A long-standing issue of estimating the effect of tax cuts on economic output, and incorporating these macroeconomic feedback effects into future revenue estimates, became more prominent in the 108th Congress. For some years, the Joint Committee on Taxation (JCT) had studied the issue,1 and had moved toward developing economic models for dynamic revenue estimating. (Official JCT estimates incorporate microeconomic behavioral responses, but keep aggregate output fixed.) When organizing for the 108th Congress, a House rule requiring such a macroeconomic analysis to be provided for any tax bill considered in the House was adopted. The first instance of the application of the House rule was during consideration of the 2003 tax cut: H.R. 2, the Jobs and Growth Tax Relief Reconciliation Act of 2003. Although the Joint Tax Committee provided an analysis,2 a significant range of results 1 Joint Committee on Taxation Tax Modeling Project and 1997 Symposium Papers, Joint Committee Print, JCS-21-97, Washington, DC, U.S. Government Printing Office, Nov. 20, 1997. This project brought together nine modelers to estimate the economic effects of fundamental tax reform. The committee print provides an overview of their findings, their individual papers, and discussants' comments. 2 This analysis was inserted in the Congressional Record, May 8, 2003, pp. H3829-H3832. The committee subsequently published a background document that discussed modeling details: Joint Committee on Taxation, U.S. Congress, Overview of Work of the Staff of the Joint Committee on (continued...) Congressional Research Service ~ The Library of Congress CRS-2 was reported, depending on the model type and assumptions used. And, it is generally recognized that the direction, as well as the magnitude, of results is sensitive to a number of economic assumptions. Issues have also been raised about the appropriateness of including macroeconomic feedback effects for revenue bills, but not for spending bills. The Congressional Budget Office (CBO) had earlier in 2003 provided a range of macroeconomic effects for the President's FY2004 budget package, which included spending as well as tax changes.3 The revenue effects with feedbacks were not incorporated in the official scoring of the bill, and the Senate does not have a requirement for macroeconomic analysis. A similar analysis was prepared for the FY2005 budget proposal.4 States had also experimented with dynamic revenue estimating, although only a handful have developed dynamic models. California appears to be the only state that frequently performs dynamic analysis.5 The underlying policy problems are several. First, to what extent should the feedback effects of revenues be used to assess the desirability of alternative tax revisions? http://wikileaks.org/wiki/CRS-RS22020 And what sort of feedback effects should be considered? Feedback effects of a tax cut can include those arising from the short run economic stimulus (where increased income results in offsetting revenue gains), the effects of deficit finance which crowd out private investment and magnify the revenue costs, and the supply side effects (primarily labor supply effects) which could be positive or negative. Some have argued that demand side effects are not appropriate for inclusion. Is there sufficient evidence and a sufficient consensus on the magnitude and direction of these effects? Economists have disagreements, sometimes pronounced, about the appropriate models to use in providing macroeconomic feedback effects. Thus far, the approach of the Joint Tax Committee and the Congressional Budget Office has been to use a range of different models. Economic research has also produced a significant range of empirical estimates for both demand induced (stimulus) effects on output and supply- side effects. Some models produce results that are heavily dependent on other, unspecified, policy decisions (either legislative or actions of the Federal Reserve Board). The range of models and results can produce estimates that vary not only in the magnitudes, but the direction, of feedback effects. The CBO study of the President's FY2004 budget package, for example, reported a wide range of potential effects -- from 2 (...continued) Taxation to Model the Macroeconomic Effects of Proposed Tax Legislation to Comply with House Rule II.i.(h)(2), 2003, Dec. 2003 (JCX-105-03). 3 See Congressional Budget Office, An Analysis of the President's Budget Proposal for FY2004, Mar. 2003. A subsequent study that provided technical details of the assumptions used, How CBO Analyzed the Macroeconomic Effect of the President's Budget, was published in July 2003. 4 Congressional Budget Office, An Analysis of the President's Budget Proposals for FY 2005, Mar. 2004. 5 See Jay Wortley, "Dynamic Revenue Estimating: A State Perspective," Proceedings of the National Tax Association 96th Annual Conference 2003, Washington, DC, National Tax Association, 2004, pp. 318-324. CRS-3 a 30% decrease in revenue (and other budgetary) costs to a 15% increase. These differences reflect the types of effects included, which of the four types of models were used, and the behavioral responses. The range of results in the CBO study would be even larger if further sensitivity analysis for supply response were undertaken; in particular, such sensitivity analysis would probably cause larger additional costs (rather than revenue offsets) from feedback effects. Does the macroeconomic analysis of tax cuts, but not of spending increases, create a more favorable environment for tax cuts than spending increases to stimulate the economy? Spending increases have similar, or even larger, effects in stimulating the economy in the short run. Does an estimate of revenue effects alone, without considering related costs (such as increased interest costs arising from increased debt) cause difficulties in understanding the budgetary effects of tax changes? Finally, should feedback effects be incorporated into official scoring of costs of legislation? There are also a number of institutional and political considerations. There is disagreement about the desirability of officially produced dynamic revenue estimates. Supporters argue that such information is needed to evaluate tax legislation, and is an http://wikileaks.org/wiki/CRS-RS22020 important means of differentiating among different tax proposals. Critics fear that these estimates might eventually become part of official scores. They are also concerned that political pressure by the majority party could be brought on preparing these estimates. Some critics are also concerned that the analysis will create a bias in favor of tax cuts rather than spending, or favor tax cuts for wealthy individuals. Some conservatives are also concerned that the practice will spread to spending programs and encourage too much spending.6 6 CRS Report RL31949, Issues in Dynamic Revenue Estimating, by Jane G. Gravelle provides background and analysis of the basic issues, including types of effects and models, and surveys of evidence on behavioral responses. For further discussions of the political and economic issues, see Joint Economic Committee, "Understanding the CBO's Dynamic Analysis," Apr. 1, 2003; Gene Steuerle, "Making the Right Case for Dynamic Analysis," Tax Notes, Apr. 21, 2003, pp. 417-419; John W. Diamond and Pamela H. Moomau, "Issues in Analyzing the Macroeconomic Effects of Tax Policy," National Tax Journal, vol. 56. Sept. 2003, pp. 447-462; "AEI Conference Examines the Future of Revenue Estimating," Tax Notes, Nov. 10, 2003, pp. 683-686; Jane G. Gravelle, "Models and Elasticities in Dynamic Revenue Estimating," Proceedings of the National Tax Association 96th Annual Conference 2003, Washington, DC, National Tax Association, 2004, pp. 306-317; William G. Gale and Peter R. Orszag, "Bush Administration Tax Policy: Effects on Long-Term Growth," Tax Notes, Oct. 18, 2004, pp. 415- 423; and Martin A. Sullivan, "Practical Aspects of Dynamic Revenue Estimation," Tax Notes, Nov. 29, 2004. This last paper has an extensive history of events in the 1990s, as well as a brief discussion of state practices and a review of economic issues.