Explore Harvard's Nieman network Nieman Fellowships Nieman Lab Nieman Reports Nieman Storyboard

Tax code helps redistribute income – at the top

COMMENTARY | May 04, 2006

From 1993 to 2003, average pay for S&P 500 CEOs rose $5.4 million. Meanwhile, middle class wages have been stagnant for 20 years.

By Henry Banta

It is now widely recognized that there has been a huge explosion in corporate management compensation, thanks, in part, to the spectacular generosity of Exxon-Mobil to its former CEO.  Even the popular media are beginning to understand that this increase in the compensation of top corporate managers has been a major element in the redistribution of income in the country. But the extent of this redistribution has been largely buried by the outrage over the more extreme examples of executive overreaching.

In general terms, the income of the American middle class has stagnated for over twenty years, while virtually the entire gain in GDP went to the top one percent, or less. Over the last twenty years this massive gain in income came increasingly from wages and salaries while in previous decades the income of the very rich came largely from returns on investments. A new phenomenon was making some people very rich: corporate management compensation.

By almost every measure executive compensation has increased at a dramatic rate. For example, among the S&P 500 firms average CEO compensation went from $3.7 million 1993 to $9.1 million in 2003; average top-five executive compensation increased from $9.5 million in 1993 to $21.4 million in 2003.

In an era when the "market" has become a sacred icon, how could this happen? To the extent that the explosion in management compensation has been explained, it is usually in terms of a failure of "corporate governance." Boards of directors, compensation committees, compensation consultants, human resources departments are, to one degree or another, identified as culprits. Lucian Bebchuk and Jesse Fried in their book, "Pay without Performance," make a persuasive case that the problem lies in the power relationship between top managers and their boards. This explains the behavior of individual corporations, but it does not explain industry-wide or economy-wide behavior. How could this have happened in industry after industry across the entire economy? Why didn't competition restrain such profligate behavior?   

What is generally ignored is that this huge transfer of income has been, in large part, promoted by the tax code. Stock options have long been an element in management compensation. Beginning in the early 1990s their use dramatically increased. Bebchuk and Yaniv Grinstein (in an article the Oxford Review of Economic Policy) reported that equity-based compensation (stock options and restricted stock) in 1993 was 37 percent of the total compensation paid to the top five executives of S&P 500 firms. By 2003 this had risen to 55 percent.

At the same time a very significant gap developed between the book income of corporations and their tax income. Book income is what corporations report to their shareholders; tax income is what they report to the IRS. In March of 2002, Martin A. Sullivan, writing in Tax Notes, pointed out that when employees cash in their stock options, corporations may take deductions on their tax returns equal to their employees' gain on the options without reducing their book income. Sullivan estimated the value of the tax benefits to corporations provided by stock options was $27.6 billion in 1998, $42.0 billion in 1999, and $56.4 billion by 2000. He noted that the treatment of options explained why Enron paid very little taxes on the profits it was reporting, and that it was unlikely, based on their grants of stock options, that other firms such as Microsoft, Cisco, Dell, America Online, Amgen, and Sun Microsystems, paid any or very much federal income tax.

In an April 2002 National Bureau of Economic Research working paper Mihir A. Desai found the increasing gap between book income and tax income could be partly explained by corporate tax sheltering activity, the most important of which is the use of stock options for executive compensation. He noted, "For the largest public companies, proceeds from option exercises equaled 27 percent of operating cash flow from 1996 to 2000 and that these deductions appear to be fully utilized thereby creating the largest distinction between book and tax income." 

The Congress and the press have focused on the debate over the accounting treatment of stock options. Many critics in the financial world claim that the tax code is correct in treating the employee's gain on a stock option as a cost to the firm and that its book accounting treatment creates illusory profits. But the focus has been on the notion that the accounting was unfair to shareholders in that it pretended that options were not a cost to the firm. (Of course, one man's illusion is another's reality. This illusion is clung to and defended with tenacity that can be truly called heroic.) The effect on middle class employees who did not get generous options has gone largely without comment.

Let's put all this in context. Virtually every corporation in America is claiming that the inexorable forces of the competitive market are forcing massive retrenchments on the pay and benefits of their employees. Pension rights, health care benefits, generous vacations and job security are disappearing. The business pages regularly recount the drastic measures firms are forced to take to cut costs, the sole exception being management's pay. But no attention is paid to the broad consequences of the tax treatment of stock options. Except for Messrs. Sullivan and Desai, the role of corporate taxes has gone virtually without comment in the debate over executive pay.

The notion that a tax provision, making a major contribution to one of the most massive redistributions of income in the history of any industrialized country, could go essentially without public notice, much less debate, is frightening.

Economists have displayed admirable imagination in developing clever and elegant theories to justify the momentous increases in corporate executive compensation. Virtually all can be summed up with the claim: "They earned it." Most of the literature on the other side can be summarized with: "No, they didn't." At this stage, a review of this debate hardly seems worthwhile. The evidence demonstrating the lack of connection between executive pay and performance is overwhelming. (Anyone interested in pursuing the rise and fall of the various theories can consult "Pay without Performance" and the extensive bibliography therein.)

More than anything else, it is the sheer scale of the income distribution resulting from the increase in management compensation that has done irreparable damage to the various theories seeking to justify the increases. In standard neo-classical economic theory, executive compensation should be determined by marginal productivity, i.e. an increase in pay should be proportionate to one's contribution to the increase in output.

Let us leave aside the fact that management's marginal productivity is well beyond anyone's ability to measure. If it were true that productivity growth justified management's increasing share of national income, a corollary would also have to be true: the marginal productivity of virtually the entire middle class has stagnated for over twenty years. The growth of our national income would have been a result of the efforts of less than one percent of the population. This proposition must raise problems even for the most devout believer in the free market. Besides, when did they start thinking that tax shelters were part of the free market?

The NiemanWatchdog.org website is no longer being updated. Watchdog stories have a new home in Nieman Reports.