Revised July 9, 1997

Do Criticisms of the Treasury Department’s
Distributional Analyses Have Merit?

Why Treasury and Joint Tax Committee Analyses Differ

by Robert Greenstein and Iris J. Lav


In recent weeks, House Ways and Means Committee Chairman Bill Archer has issued press releases, sent a letter to the President, and made statements claiming that the Treasury Department has misused data on family incomes to portray the House tax bill inaccurately as being tilted toward the rich. For example, a June 12 Archer press release was entitled "The Trick in Treasury's Tables" and subtitled "Treasury Uses Misleading Method to Cook Their Books." Rep. Archer has charged that the Treasury is defining income in such a way as to increase families' incomes artificially, making families look richer than they actually are and the House and Senate tax bills appear more favorable to the wealthy than they really are. On July 6, House Speaker Newt Gingrich echoed the Archer changes on CBS' "Face the Nation," claiming the Treasury is "deliberately rigging the numbers in a way that is weird."

In his press releases, letter and statements, Mr. Archer has contended that the House bill provides the bulk of its tax benefits to the middle class, not to the well-to-do. He has claimed the bill would provide 71 percent of its tax cuts to people with incomes between $20,000 and $75,000. By contrast, the Treasury analyses find that two-thirds of the tax cuts in the House and Senate bills would go to the top 20 percent of households when the tax cuts are fully in effect and that the top one percent of households would receive nearly as much or more in tax cuts as the bottom 60 percent combined.

These sharp differences raise important questions. Are Mr. Archer's criticisms of the Treasury analyses valid? What accounts for the very large differences between the Treasury analyses and the figures Mr. Archer cites, which portray the Congressional tax bills as primarily middle-class tax cuts?

As this analysis explains, an examination of Mr. Archer's charges concerning the Treasury analyses shows the charges to be without foundation. Analysis also indicates that the principal reason the Treasury figures and Mr. Archer's figures differ so sharply is that the figures Mr. Archer is using on the House bill's impact on different income groups distort the bill's effects in large ways.

Different Income Definitions Do Not Fundamentally Affect Distributional Analyses

There are a number of different ways to measure income when conducting distributional analyses. The three governmental entities that examine income distribution issues and the impact of tax and budget changes on income — the Congressional Budget Office, the Joint Committee on Taxation, and the Treasury Department — each employ a different income definition. Some definitions count more non-cash income than others. But whatever income definition is used, the division of families (or taxpaying units) into income fifths, or quintiles, is accomplished the same way — families are ranked by income and then divided into fifths. In most cases, the families in the top fifth of the Treasury distribution and the families in the top fifth of the Joint Tax Committee distribution are the same or similar families. (This also holds true for other income quintiles.) As a result, the effects of tax policy changes on the various income quintiles, and on the top one percent and top five percent of the population, are not affected much by the income definition used.

In defining income, both the Joint Tax Committee and the Treasury Department count various items as income that the Census Bureau does not count. The Joint Committee counts contributions that a family's employer makes for the family's health and life insurance as well as for payroll taxes. The Joint Committee also counts as income the insurance value of Medicare. Treasury counts the employer contributions as income as well and also counts several additional items such as employer contributions to pension plans, profit-sharing payments, stock options, and the build-up of untaxed earnings in IRA and Keogh plans.1 There is little debate among economists that these items have value to families and increase families' net wealth.

Reasonable people can differ about which of these non-cash income sources should be counted as income when assessing the impact of tax changes on different income groups. But such differences have little effect on distributional analyses of tax cuts. The proportion of the House tax cut that would go to each income quintile will be broadly similar regardless of how narrow or expansive the definition of family income that is being used.

If the Joint Committee on Taxation had looked at the impact of the House tax provisions when the provisions would be fully effective instead of just looking at their impact through 2002, and had used the traditional methodology to measure the impact of capital gains tax cuts on different income groups (a methodology the Joint Committee used until late 1994), the Joint Tax Committee tables would look similar to the Treasury tables. Furthermore, a recent Center on Budget and Policy Priorities/Citizens for Tax Justice study used the same definition of income that the Congressional Budget Office employs; this is a less expansive income definition than the Joint Tax Committee uses and the definition closest to the Census income definition that Mr. Archer apparently regards as the most appropriate income measure. The Center/CTJ study found the House tax bill to be somewhat more skewed toward higher-income taxpayers than the Treasury study did.

1 Both the Joint Tax Committee and the Treasury count capital gains income as part of family income but do so in a somewhat different way. The Joint Committee counts capital gains income when it is realized, i.e., when assets are sold. Treasury counts the annual accrual of capital gains as income.

Rep. Archer repeatedly contends, based on a Joint Tax Committee table, that 71 percent of the tax cuts the House bill provides would go to people with incomes between $20,000 and $75,000. Such a finding is sharply at odds with the Treasury analysis. But the reason that the Treasury's distributional analyses and the Joint Tax Committee's analyses differ so markedly is not that they use somewhat different definitions of family income. The sharp differences are instead the result of serious deficiencies in the Joint Committee's tables and methodology. These deficiencies cause the Joint Committee's tables to distort the effects of the tax legislation on different income groups.

Leading Analysts Reject New Joint Tax Methodology on the Distribution of Capital Gains Tax Benefits

Many of the leading analysts in the field reject the new Joint Tax Committee method as producing severe distortions in the distribution of the benefits that a capital gains tax cut produces. Among those rejecting the new Joint Tax Committee approach are: Robert Reischauer, former director of the Congressional Budget Office; Henry Aaron, senior fellow at the Brookings Institution; and Jane Gravelle, the Congressional Research Service's leading tax expert and analyst. In addition, several years ago Gravelle co-authored an article on this matter with Lawrence Lindsey, a noted conservative economist who served until recently on the Federal Reserve Board and who supports a capital gains tax cut. In their article, Lindsey and Gravelle explicitly rejected the methodology the Joint Tax Committee has now adopted.

As Aaron has observed, investors who respond to a capital gains tax cut by selling more assets are people who face one set of opportunities under the current capital gains tax rates — and find it financially advantageous not to make additional asset sales — but face a more generous set of opportunities when capital gains tax rates are reduced and choose to follow a different course. "Since they have the option of doing what they did before (i.e., not selling additional assets), but the new, more favorable tax rates induce them to do something else, they must be better off," Aaron explains. "It is logically absurd to count them as worse off in any way whatsoever."

Aaron's view is supported by an article Gravelle and Lindsey co-authored in 1988 before Lindsey joined the Fed. In the article they stated:

"... suppose a reduction in the capital gains tax rate led to substantially more capital gains realizations [i.e., more sales of assets] and actually increased the tax revenue paid by upper-income would be totally inappropriate to say that their tax burden had increased. After all, with a lower tax rate, these upper-income taxpayers are less burdened than they were before, even though they pay more taxes."(8)1

In addition, in a more recent analysis examining the new Joint Tax Committee methodology, Gravelle notes that the standard methodology, if anything, understates the benefits that investors would secure from a capital gains tax cut because it does not reflect the tax benefits they would receive when they voluntarily sell more assets to take advantage of a lower capital gains tax rate. She also observes that economists generally would reject the new methodology.


1 This quote is from Jane G. Gravelle and Lawrence B. Lindsey, "Capital Gains," Tax Notes, January 25, 1988, p. 399. Gravelle included this quote in Jane G. Gravelle, "Distributional Effects of Tax Provisions in the Contract with America as reported by the Ways and Means Committee," CRS Report for Congress, April 3, 1995.

The Treasury analysis finds that the House tax bill would provide 67 percent of its benefits to the top 20 percent of the population. It also finds that the wealthiest one percent of the population would get nearly 1 times as much in tax cuts as the bottom 60 percent of the population.(9)

In its paper, the Congressional Research Service supports the Treasury analysis. CRS states, "Using measures more consistent with conventional economic analysis and the permanent provisions of the bills, the Treasury OTA's estimates indicate that, by any distributional measure, the tax cuts favor higher income individuals in the House and Senate bills, with the effects more pronounced in the House bill."(10)

If Mr. Archer disagrees so strongly with the Treasury analysis and the Treasury's income definition, there is a step he can take. He can ask the Joint Committee on Taxation to conduct an analysis of the distributional effects of the House tax bill when the bill is fully in effect (not in 2002), using the generally agreed-upon methodology for assessing the impacts of capital gains cuts (i.e., the methodology the Joint Committee itself used to use). The results of such a Joint Tax Committee analysis — using the Joint Committee's definition of income — would almost certainly mirror the results of the Treasury analyses that Rep. Archer has so sharply castigated.

End Notes

1. Department of the Treasury, Tax Reform for Fairness, Simplicity, and Economic Growth: The Treasury Department Report to the President, November 1984, pp. 57-59.

2. Jane G. Gravelle, Distributional Effects of the Proposed Tax Cut, Congressional Research Service, July 2, 1997, p. 3. Gravelle is CRS' leading expert on these aspects of federal tax policy and is recognized nationally for her tax analysis work.

3. CRS, p. 4.

4. CRS, p. 3.

5. CRS, p. 3, 4.

6. CRS, p. 4.

7. For example, the Treasury methods take into account the future value of backloaded IRAs to taxpayers, while the Joint Committee analysis includes virtually none of the value of the proposed accounts.

8. This quote is from Jane G. Gravelle and Lawrence B. Lindsey, "Capital Gains," Tax Notes, January 25, 1988, p. 399. Gravelle included this quote in Jane G. Gravelle, "Distributional Effects of Tax Provisions in the Contract with America as reported by the Ways and Means Committee," CRS Report for Congress, April 3, 1995.

9. As noted earlier, neither the Joint Tax Committee tables nor the Treasury tables account for the effects of the estate tax cuts. An analysis conducted by the Center on Budget and Policy Priorities and Citizens for Tax Justice, which differs from the Treasury analysis in that it includes the effects of the estate tax cuts and the full, long-term effects of capital gains indexing, estimates that an even larger proportion of the House tax cuts — close to 80 percent — would go to the top 20 percent of the population when the tax cuts are fully in effect.

10. CRS, p. 6. OTA refers to the Office of Tax Analysis at the Treasury Department.