Budget Plans Should Not Rely on "Dynamic Scoring"
Estimates Are Uncertain and Subject to Manipulation

PDF of this report (6pp.)

By Paul N. Van de Water

Revised June 21, 2012

Some Members of Congress and outside groups are calling for the use of "dynamic scoring" to estimate the budgetary effects of major legislation, notably tax reform proposals.  In February, for instance, the House passed a bill (H.R. 3582) requiring the Congressional Budget Office (CBO) and the Joint Committee on Taxation (JCT) to prepare "dynamic" estimates of major legislation.  Congress should reject the temptation to use dynamic scoring, which would include estimates of how changes in tax policy would generate changes in the overall economy — such as, the size of the economy as measured by gross domestic product (GDP).

Contrary to the widespread misunderstanding among policymakers and others, the standard estimates of tax and spending proposals that the Congressional Budget Office and other federal agencies prepare are not "static."  They incorporate many changes in individual and business behavior that occur in response to changes in tax rates and other policies.

They do not, however, include estimates of "macroeconomic feedbacks."  That is, they do not attempt to estimate whether, and by how much, a change in tax or spending policy would affect the overall economy.  They do not, for instance, include estimates of how a proposal would affect GDP growth and, therefore, do not include an estimate of how any change in GDP would affect revenues.

There are very good reasons that federal agencies do not use dynamic scoring:

  • Estimates of the macroeconomic effects of tax changes are highly uncertain.  Economists do not agree on the size of macroeconomic feedbacks from reducing marginal income tax rates or other tax changes.  According to most studies, however, they would likely be small and not have large enough effects on revenue estimates to justify the problems that dynamic scoring would create.
  • Dynamic scoring would impair the credibility of the budget process.  Because the estimates of macroeconomic feedbacks are so uncertain, including them in revenue estimates would be highly controversial and inevitably viewed as biased and politically motivated.  Including macroeconomic feedbacks for the first and only time to date in estimating a tax reform proposal would appear arbitrary and would seem like a budgetary gimmick.

Finally, if Congress can craft a package of tax changes that would expand the income tax base by eliminating various tax preferences, the economy would benefit more from using the additional revenues to reduce budget deficits rather than to cut marginal tax rates.  When economic resources are fully employed, deficits reduce saving, investment, and economic growth and, according to most estimates, reducing deficits will have a larger effect in spurring economic growth than using budget savings to cut marginal tax rates.

Revenue Estimates Already Include a Wide Range of Behavioral Responses

Congress's Joint Committee on Taxation  and Treasury's Office of Tax Analysis (OTA) both produce estimates of  changes in tax laws would affect federal tax receipts.  (CBO includes JCT's estimates of tax-law changes in the cost estimates it provides to Congress.)  They measure these changes relative to a revenue baseline that is a projection of the receipts that would be collected under current tax laws or policies.

The term "dynamic scoring" is misleading, since the standard revenue estimates of JCT and OTA are not static but incorporate a wide range of behavioral changes in response to changes in economic incentives.  According to JCT, "Such behavioral effects include shifts in the timing of transactions and income recognition, shifts between business sectors and entity form, shifts in portfolio holdings, shifts in consumption, and tax planning and avoidance strategies."[1]  

For example, if a bill increased the excise tax on cigarettes or gasoline, the revenue estimate would take account of the decrease in smoking or driving that would stem from the expected increase in the price of the taxed product.  By the same token, if a bill decreased or increased the tax rate on capital gains, the revenue estimate would take account of the resulting increase or decrease in capital gains realizations, as well as changes in the timing of realizations.  If the change in realizations is partly offset by a shift away from or towards other forms of income, such as dividends, the estimate would also take account of the change in revenues resulting from the shift.[2]

Revenue and spending estimates, however, traditionally do not include macroeconomic responses, such as changes in the overall level of economic activity.  As explained below, there are sound reasons for this omission.  These excluded macroeconomic responses are of two types: cyclical and structural, or short-run and long-run.  "Some tax or spending bills," explains CBO, "might affect aggregate demand (that is, total spending in the economy) and, if not offset by changes in monetary policy, restrain or stimulate the economy in the short run.  Other proposed legislation could alter the supplies of labor, capital, or technology that determine the potential growth of the economy in the long run."[3]   Current budget estimates exclude both types of macroeconomic feedbacks, although JCT and CBO provide supplemental analyses of the macroeconomic effects of legislation.

Estimates of Macroeconomic Feedbacks Are Uncertain and Small

Estimates of the macroeconomic effects of tax changes depend critically on the assumptions and methods employed, including the choice of economic model, the specific parameters assumed for that model, and the assumed response of the Federal Reserve to changes in fiscal policy.  Different models and assumptions can produce estimates that vary widely — sometimes even in the direction of their effect (for example, in whether a given tax change increases or reduces economic growth).

The relationship between taxes and labor supply illustrates the uncertainty with respect to one such assumption.  A reduction in marginal tax rates (the tax imposed on an extra dollar of income) has two different effects on labor supply that work in opposite directions.  On the one hand, cutting marginal tax rates increases the after-tax compensation that a person receives for an additional hour of work; this makes additional work more attractive.  On the other hand, cutting tax rates increases total take-home pay from a person's current hours of work; this allows a person to maintain his or her current level of after-tax income while working fewer hours.  Economists call these two effects the "substitution effect" and the "income effect," respectively, but economic theory alone cannot say which one is larger, and analysis of actual data is required. 

Analyzing people's decisions about how much to work is not a simple matter.  Many factors impinge on how much a person works, including employers' demand for labor as well as individual preferences, and different categories of people respond differently to changes in after-tax wage rates.  Disentangling all of these factors requires sophisticated data analysis.  Simply comparing revenues in different years is inadequate, since tax collections are highly volatile and can swing from year to year for many reasons unrelated to tax law changes.[4]

Measurements of the responsiveness of labor supply to changes in after-tax wages (termed the labor-supply "elasticity") not surprisingly prove to be very uncertain.  A CBO analysis concludes that "Academic studies reach no consensus about the size of the elasticity, reporting a wide range of estimates."[5]   Depending on the elasticity assumed, CBO estimates that the effect of extending the 2001 and 2003 tax cuts and relief from the alternative minimum tax (AMT) could vary by a factor of six — from an increase of 0.15 percent in labor supply to an increase of 0.90 percent.[6]

The range of estimates of macroeconomic feedbacks is even wider when all sources of uncertainty are taken into account.  CBO finds that the effect of permanently extending the 2001 and 2003 tax cuts and AMT relief on gross national product (GNP) ten years in the future "varies substantially — ranging from a reduction of 1.6 percent to an increase of 0.1 percent — depending on the model and the assumptions used."[7]   Using other models and assumptions, JCT also finds substantial uncertainty in the size of macroeconomic feedbacks.[8]

An important but often overlooked technical issue is that estimating macroeconomic feedbacks may require making assumptions about how and when proposed policies will be paid for.  "Some approaches to dynamic scoring, particularly forward-looking growth approaches, simply will not work (i.e, the computer algorithms will not function) when the government budget is on an explosive debt trajectory," explains former CBO Director Douglas Holtz-Eakin.  Since legislative proposals that reduce taxes or increase spending will produce debt that will grow explosively, analyzing the macroeconomic effects of such policies in certain models requires the introduction of an offsetting budget policy at some point in the future.[9]   Consequently, the results of these models often reveal more about the effects of the assumed offsetting policy than they do about the effects of the proposed policy itself.[10]

Because of the uncertainty of the estimates, it is entirely appropriate that estimates of the effects of tax legislation on revenues exclude macroeconomic feedbacks.  As former CBO Director Rudolph Penner has written:

The fact of the matter is that economists differ significantly in their assessment of the effects of tax cuts. . . . There may come a day when there is sufficient agreement about dynamic effects to automate the process using powerful computers.  But we are many decades from such technology.[11]

Moreover, most studies find that excluding macroeconomic feedbacks has only a small effect on revenue estimates.  And budget estimators have consistently stated that tax cuts come nowhere close to paying for themselves through stronger economic growth.  Even under a very optimistic economic scenario, for example, President George W. Bush's own Treasury Department estimated that the macroeconomic effects of his tax cuts would offset less than 10 percent of their conventionally estimated cost.[12]   Former CBO Director Penner concludes that leaving out macroeconomic feedbacks "is not so bad.  The CBO's dynamic analysis suggests that static scoring is usually pretty accurate."

Dynamic Scoring Would Impair the Credibility of the Budget Process

Congressional cost estimates have always left out macroeconomic feedbacks and have done so for good and sufficient reasons.  First, as explained earlier, there is little consensus among economists on the size of macroeconomic feedbacks, and they are likely to be small.  Second, to assure fairness and consistency in the budget process, estimates of macroeconomic feedbacks would have to be prepared for spending bills as well as tax legislation.  (Government investments in infrastructure, education, and basic research can boost long-term economic growth, just like private investment.)  Third, if macroeconomic feedbacks were taken into account, spending bills could affect revenues (and the reverse), and the congressional budget process (which now has separate procedures for controlling revenues and spending) would have to be changed to recognize that fact.  Fourth, estimating macroeconomic feedbacks can be complex and sometimes could not be completed within the time that the legislative process now requires.[13]

Including macroeconomic feedbacks in the cost estimate for a budget proposal — or for a subsequent tax reform bill — would break with longstanding practice and impair the credibility of both the proposal and the budget process itself.  Congress could exert pressure on the JCT, whose staff reports to the chairmen of the Senate Finance and House Ways and Means Committees.  Including macroeconomic feedbacks in the revenue estimate for this one bill would also appear arbitrary and inconsistent and would smack of budgetary gimmickry.  For these reasons, people would inevitably view the revenue estimate as biased and politically motivated.

Reducing Deficits Is More Beneficial Than Cutting Tax Rates

Some members of Congress have proposed tax changes that would broaden the income tax base by eliminating various tax preferences and use the resulting increase in revenues to cut marginal tax rates.[14]   Such a package might be constructed so that it would be "revenue-neutral" (that is, have no net effect on revenues) using standard estimating methods but could increase revenues if certain assumptions (designed to produce that result) are made about its macroeconomic feedbacks.

Revenue-neutral tax reform, however, would not necessarily produce significant positive macroeconomic feedbacks.  Any legislation that phases out major tax preferences is likely to include transitional provisions to cushion the blow — for example, to moderate the drop in house prices that would result from eliminating the income tax deductibility of mortgage interest and property taxes.  The complexity that these transition rules and other legislative details would add to the tax code could easily offset some or all of the benefits from reducing marginal tax rates.

As a case in point, the Tax Reform Act of 1986 (TRA86) — the last major effort to broaden the tax base and reduce rates — appears to have produced little real impact on economic growth.  Summarizing the results of an assessment sponsored by the University of Michigan, one prominent economist wrote, "Most of the papers presented at this conference reinforce the casual observation that TRA86 has had little effect on the broad measures of real economic activity in which most economists are interested."[15]

Even if revenue-neutral tax reform might produce some small economic growth benefits, it would be far more economically beneficial to use the additional revenues gained from limiting tax preferences to reduce budget deficits rather than cut marginal tax rates.  When the economy is operating near or at its capacity (unlike the situation today), federal budget deficits reduce total saving in the economy, crowd out capital investment, and reduce the economy's potential rate of growth.  Most economists believe that the adverse effect of higher deficits dominates the effect of higher tax rates.  For example, the Congressional Budget Office finds that permanent extension of the 2001 and 2003 tax cuts and AMT relief would reduce output in the long run if the extension is deficit-financed.[16]   Conversely, reducing the deficit will spur economic growth even if it requires higher tax rates.


Given our nation's projected long-term budget deficits, the single most important goal of tax reform should be to raise substantial revenue, in a progressive manner, as part of a balanced deficit-reduction plan that also includes reductions in projected spending.  If not done carefully, however, tax reform could increase the deficit and threaten the progressivity of the tax code.  A tax reform approach that specifies rate cuts but not tax expenditure cuts in advance is one pitfall to avoid.  When lawmakers realize the extent of cuts they will have to make in popular tax benefits to finance the rate cuts and meet the revenue target, they will likely balk — and abandon the revenue target.[17]

Dynamic scoring poses yet another risk.  In that case, speculative revenue gains that assume macroeconomic feedback which may never occur would replace real changes in tax policies, and deficit-reduction would again lose out.  Congress should therefore reject proposals to use dynamic scoring to evaluate tax reform or other proposed legislation.

End notes:

[1] Thomas A. Barthold, Chief of Staff, Joint Committee on Taxation, Testimony of the Staff of the Joint Committee on Taxation before the House Committee on Ways and Means Regarding Economic Modeling, September 21, 2011.

[2] Congressional Budget Office, Budget Estimates: Current Practices and Alternative Approaches, January 1995.

[3] CBO, Budget Estimates, p. 3.

[4] Jason Furman, A Short Guide to Dynamic Scoring, Center on Budget and Policy Priorities, August 24, 2006, http://www.cbpp.org/files/7-12-06bud2.pdf.

[5] Congressional Budget Office, The Effect of Tax Changes on Labor Supply in CBO's Microsimulation Tax Model," CBO Background Paper, April 2007, p. 7.

[6] CBO, The Effect of Tax Changes on Labor Supply, p. 9.

[7] Douglas W. Elmendorf, Director, Congressional Budget Office, The Economic Outlook and Fiscal Policy Choices, Statement before the Committee on the Budget, United States Senate, September 28, 2010.

[8] Barthold, Testimony.

[9] Douglas Holtz-Eakin, Economic Models Available for Analyzing Tax Reform Proposals, Testimony before the Committee on Ways and Means, U.S. House of Representatives, September 21, 2011, pp. 4-5.

[10] For examples, see John L. Buckley, Dynamic Scoring: Will S&P Have Company?, Tax Analysts, February 28, 2012, http://www.taxanalysts.com/www/features.nsf/Articles/43736B49FCB019E3852579B5006E1933?OpenDocument.

[11] Rudolph G. Penner, "Dynamic Scoring: Not So Fast!,"  The Ripon Forum, April/May 2006.

[12] Furman, A Short Guide to Dynamic Scoring, pp. 5-6.

[13] CBO, Budget Estimates, pp. 13-20.

[14] James R. Horney, Chye-Ching Huang, Edwin Park, and Paul Van de Water, Toomey Budget Similar to House-Passed Ryan Budget, Center on Budget and Policy Priorities, May 9, 2012, http://www.cbpp.org/cms/index.cfm?fa=view&id=3771.

[15] Henry J. Aaron, "Lessons for Tax Reform," in Joel Slemrod (ed.), Do Taxes Matter? (Cambridge: MIT Press, 1990), p. 322.

[16] Elmendorf, The Economic Outlook and Fiscal Policy Choices, p. 32.  See also Congressional Budget Office, The Long-Term Budget Outlook, June 2011, pp. 28-31.

[17] Chuck Marr and Chye-Ching Huang, Tax Reform Holds Promise, But If Not Done Carefully, Could Increase the Deficit and Inequality and Harm the Economy, Center on Budget and Policy Priorities, June 8, 2012, http://www.cbpp.org/cms/index.cfm?fa=view&id=3792.

  1. Jobs
  2. RSS
  3. Contact Us

Sign Up for E-Mail Alerts

RSS Feeds


Browse Reports