Does Atlas Shrug? The Economic Consequences of Taxing the Rich.

• Author: Avi-Yonah, Reuven S.

Edited by Joel B. Slemrod. * Cambridge: Harvard University Press, 2000. Pp. 524. $57.95.

In Greek mythology, Atlas was a giant who carried the world on his shoulders. In Ayn Rand's 1957 novel Atlas Shrugged, Atlas represents the "prime movers"--the talented few who bear the weight of the world's economy. (1) In the novel, the prime movers go on strike against the oppressive burden of excessive regulation and taxation, leaving the world in disarray and demonstrating how indispensable they are to the rest of us (the "second handers").

Rand wrote in a world in which the top marginal federal income tax rate in the United States was 91% (beginning at taxable income of $400,000). (2) This is an unimaginably high rate by today's standards, when the dominant view in Washington is that a marginal rate of 39.6% (the top rate from 1993 to 2001) is too high. (3) The key turning point in the process of abandoning high marginal tax rates occurred in the presidency of Ronald Reagan. When Reagan became President in 1981, the top marginal federal income tax rate was 70%; when he left office in 1989, the top rate was 28%. (4)

The reduction of marginal tax rates in the Reagan years was driven by a new policy consensus that still persists today. That consensus is that high marginal tax rates on the rich come with an unaffordably high price for the U.S. economy in the form of reduced incentives for the rich to work and to save, and increased incentives to engage in socially wasteful tax planning. And yet 1957, when Rand wrote Atlas Shrugged and the top income tax rate was 91%, falls in the middle of the period from 1951 through 1963. Those were the golden years of the U.S. economy, in which the average annual rate of productivity growth was 3.1% (compared with about 1.5% after 1981). (5) Of course, the growth might have been even faster had the marginal tax rates been lower, but the coincidence of high rates and high productivity raises challenging questions for those who believe that high marginal tax rates carry an unacceptable cost. (6)

Thus, the question of whether high marginal tax rates come with an unaffordably high cost to the U.S. economy remains unsettled. Does Atlas Shrug?, a recent collection of papers written mostly by public finance economists and superbly edited by Joel Slemrod, represents the most recent attempt to answer this question. Unfortunately, no clear-cut answer is forthcoming in the book, and the debate is sure to rage on. (7)

This Review is divided into three Parts. In Part I, I summarize the main findings of Does Atlas Shrug?, emphasizing their contribution to the debate on taxing the rich. In Parts II and III, I discuss a question that is only briefly touched on in the book: Why should the rich be taxed? Part II surveys the existing--and to me incomplete--legal literature on this issue, while Part III begins to outline some tentative alternative answers. In my view, the debate about the economic consequences of taxing the rich has obscured this fundamental normative question, and answering it is essential to assessing the merits and relevance of the findings contained in Slemrod's book.

1. THE ECONOMIC COST OF TAXING THE RICH: DOES ATLAS REALLY SHRUG?

   Slemrod's book is divided into three parts. Part I provides history and background. Part II is the heart of the book and contains nine new empirical studies on the behavioral responses of the rich to taxation. Part III is called "alternative perspectives" and discusses taxation of the rich on the basis of assumptions that deviate from the standard ones underlying the economic models of Part II.

   Part I begins with an excellent historical survey by W. Elliot Brownlee of the rates facing the rich from the beginning of the U.S. income tax in 1913 to the present. In terms of the top marginal rate, the income tax began modestly at 7%, but climbed precipitously to 77% during World War I. Post-war reductions brought the rate down to 25% (in 1925), but during the Hoover Administration it went back up to 63% (in 1932). The rise in rates continued through the New Deal and culminated during World War II, when the top marginal rate reached 94% (in 1944-1945). The top rate remained at 91% until 1964, when it was reduced to 77%, and then to 70% in 1974. The Reagan tax cuts sharply reduced the top rate to 28% by 1986. The rate then grew to 31% in the "read my lips" tax hike of 1990, and to 39.6% in the Clinton tax increase of 1993, before being reduced again to 35% (phased in to become fully effective in 2006) in the Bush tax cut of 2001. (8)

   Statutory marginal tax rates are important for their symbolic significance and incentive effects, but from an economic perspective, it is just as important to determine the effective tax rate facing the rich. The effective rate is the rate the rich pay after taking into account lower rates for lower brackets of income and the available deductions, credits, and other methods of narrowing the tax base (i.e., reducing the taxable income on which the marginal rate is imposed). Brownlee helpfully provides estimates of the historical effective rates for the richest one percent of households as well. He indicates that effective rates during the high marginal rate years of World War I reached 15.8%, and that during the high marginal rate years of World War II they reached an astonishing 58.6% in 1944. (9) After the war, while the top marginal rate remained extremely high at 91%, the effective rate for the rich declined to 32.2% in 1952, then 24.6% in 1963, rising to 28.9% when Ronald Reagan took office and declining to 22.1% following the 1986 tax reductions. (10) More recent estimates for the Clinton years are not yet available. The conclusion drawn by Brownlee is that the rich can be taxed at very high effective rates during times of national emergency, but that at other times their political clout ensures that effective rates are much lower than marginal rates. It turns out that when Ayn Rand was writing Atlas Shrugged, the actual burden borne by the "prime movers" was not so high after all; by the late 1940s the rich had "largely succeeded in removing the redistributional fangs from the movement for progressive taxation." (11)

   Another introductory chapter by Edward N. Wolff attempts to identify "the rich" by looking at demographic data from 1983 to 1992. He focuses on the top one percent of U.S. households (by wealth) and shows that in 1992 they owned 35.9% of total wealth (up from 32.6% in 1983) and earned 15.7% of total income (up from 12.8% in 1983)--a remarkable rise, which continued through the 1990s. (12) Wolff also shows how the demographics of the richest one percent changed from 1983 to 1992. The rich became younger and earned more labor income and less income from property. (13) The rich were also more likely to be self-employed and less likely to be salaried managers or professionals. (14)

   The final introductory chapter by Douglas A. Shackelford analyzes the current tax environment facing the rich, which includes both income and transfer taxes (the estate and gift tax). Shackelford shows that under the most tax-disadvantaged form of earning income (in which the income is taxed as ordinary income when earned and interest on the after-tax income is taxed again, and what remains is subject to the estate tax), the marginal tax rate facing the rich can be as high as 91%. He also shows, however, that various forms of legal tax avoidance (e.g., transforming ordinary income into capital gains, postponing realization of capital gains, making tax-free gifts, and transferring future interests to heirs) can reduce the effective tax rate to close to zero. (15)

   The core of the book is the nine empirical studies of Part II. In general, they provide a mixed answer to the question of what the economic costs are of taxing the rich. While there is some evidence of behavioral responses, it is quite limited and seems to depend crucially on the authors' chosen methodology. Importantly, most of the findings of behavioral response relate to the use of various tax avoidance techniques--and even there the evidence is mixed, with some obvious techniques being used less than they should be in a world in which tax minimization is very important to the rich. Real behaviors, such as labor and saving, seem much less affected by taxation. This distinction is important because while both tax avoidance techniques and real behavioral changes cause deadweight losses, the former can be partially prevented by changing the law, while the latter are less amenable to legal change since one cannot force the rich to work or save more.

   The papers in Part II can be divided into three broad groups. The first group contains studies that address the limitations of the previous literature on the economic consequences of taxing the rich. Thus, the chapter by Austan Goolsbee responds to the "new tax responsiveness" (NTR) literature pioneered by Lawrence Lindsey (President George W. Bush's chief economic adviser) and Martin Feldstein. The traditional economics literature found little evidence that high tax rates discourage labor supply by the rich. The NTR literature focused instead on the elasticity of taxable income and, by using "natural experiments" (tax rate increases and decreases), found very high responsiveness (elasticities exceeding one). (16) Goolsbee argues, however, that this literature is fatally flawed because it assumes that the rich are similar to other income groups except for their higher tax rates. In fact, the rich are different in at least three ways. First, their incomes have recently been trending upward at a rate that is faster than others' incomes, which, in a time of tax cuts for the rich, can appear as tax responsiveness. (17) Second, the rich are more sensitive to demand conditions than others, and therefore their incomes tend to surge in good times that also happen to coincide with tax cuts. (18) Finally, the compensation of the...