In a policy report published on February 6, 2001, Robert Greenstein of the Center on Budget and Policy Priorities (CBPP) claimed that President Bush's tax cut plan would "cost" the federal Treasury $2.5 trillion over the 11-year period 2001 through 2011. That sum is $1.2 trillion more than Congress's Joint Committee on Taxation (JCT) estimated the plan will save taxpayers and $900 billion more than the tax relief estimated by the President's own economists.¹

The problem is that the CBPP analysis assumed features of the Bush plan that are not in it and assumed also that the economy would not respond to the largest tax cut in 20 years. Table 1 shows what the CBPP included to arrive at its cost of President Bush's tax proposal. It also compares those costs to estimates of taxpayer savings made by the Center for Data Analysis (CDA) of The Heritage Foundation.

Table 1
Comparison of CBPP & Heritage's Tax Plan Estimates²
(in trillions of dollars)

What does the CBPP do to arrive at these estimates, and how are they different from Heritage's work and the estimates of the economists at the Joint Committee and the White House?

1. The JCT estimated the cost of the plan over the period 2001 through 2010. The CBPP estimates cover the period 2002 through 2011, and the larger tax base in 2011 adds to the revenue impact of the plan (around $200 billion). The CDA estimates are $60 billion less, or $140 billion.

2. Mr. Greenstein argues that the entire tax system will have to be fixed for growth in the Alternative Minimum Tax, since it is expanding beyond anyone's expectations. So, he adds an additional $200 billion to the cost of the Bush plan for adjusting the incidence of the AMT. But the Bush plan does not include such a provision. The CDA estimates only reflect what President Bush has actually proposed. Mr. Greenstein, however, has included this provision as though it were part of the President's plan.

3. Lower revenues under President Bush's plan mean to the CBPP that debt service costs will rise because less publicly held debt will be retired. Thus, Mr. Greenstein adds $400 billion to the 10-year cost to cover higher than expected interest costs. The CDA dynamic estimate of $940 billion, however, contains debt reduction and debt service costs associated with the tax plan. Interestingly, the adoption of President Bush's tax plan adds only one more year to the time needed for all publicly held debt to be paid back.

4. The CBPP adds another $400 billion for making the tax cut retroactive, despite the fact that President Bush's plan currently contains no retroactive components. Some politicians do recommend making portions of the tax plan effective from January 1 of 2001. Thus, the CBPP is right in assuming that a retroactive version of the plan would cost more than the one the President just sent to Congress. However, it is wrong to assume that every provision would be fully effective on January 1, 2001, and that none of the provisions would be phased in.

If the plan becomes law on July 1 and its tax rate, marriage penalty, child tax credit, and charitable deduction provisions are made effective beginning on January 1, then several revenue changes occur. First, withholding schedules are changed, which reduces collections for one quarter in fiscal year 2001 (the third calendar quarter) and the first quarter of fiscal year 2002 (the fourth calendar quarter of 2001). Second, quarterly taxpayers may file amended returns for the first two quarters of calendar year 2001 that might reduce collections in the third and fourth calendar quarters of the year. Third, regular taxpayers will file their final returns after January 1, 2002 and claim refunds for overpayment of taxes in 2001. The CDA estimates that these changes lower revenues by an additional $163 billion in 2001 and 2002.

5. Finally, the CBPP assumes that the economy does not react at all to the largest tax cut in 20 years. Thus, the 10-year JCT score (itself a nearly static estimate of the tax plan) contains no economic feedback. The CDA estimate begins with a static tax estimate of $1.7 trillion over 10 years but then makes the very reasonable real-world assumption that taxpayers will react to lower taxes on their labor and capital. Thus, the "cost" of the tax plan after a stronger economy has increased the tax base of wages and salaries is estimated at $940 billion, or 55 percent of the CDA's static estimate.³

1. Robert Greenstein, "Cost of the Bush Tax Cut Rises: Making Rate Cuts Retroactive Adds $400 Billion," Center on Budget and Policy Priorities, February 6, 2001.

2. The Heritage Center for Data Analysis follows the practice of estimating changes in revenues first with a static model and then with a behavioral model that uses output from the static model as its starting point. The difference between the Heritage static and dynamic totals in Table 1 reflect the economic reflows to the tax base, and thus to federal revenues.

3. For details on the CDA analysis of President Bush's tax plan, see D. Mark Wilson and William W. Beach, "The Economic and Budgetary Effects of President Bush's Tax Relief Plan, " Heritage Foundation Center for Data Analysis Report, No. 01-01, February 19, 2001.