The answer to the question of whether American CEOS are overpaid is clearly "yes" for those who earn large bonuses and generous stock options when their companies are doing badly, either absolutely or relative to competitors. Business Week has had an annual list of the most overpaid CEOS relative to the performance of the companies they head. A number of well-known companies usually top that list.
But the concern in the media and in Congress over CEO pay is not motivated by some bad apples like these, but by the huge increase in the typical CEO pay in the US during the past 25 years. The total real compensation (that is, compensation adjusted for increases in the price level) of CEOS in larger publicly traded companies during this period grew a remarkable six fold, where compensation adds together regular pay, bonuses, stocks awarded, the value of stock options, and payouts from longer term pay programs. A big but not the only component of the increase is due to much greater use of stock options. Since median fulltime real earnings during the same period only just about doubled, the gap between pay at the top and the average pay of employees widened enormously. It is hard to resist the widespread perception from these trends that CEOS and other top executives are being increasingly overpaid.
The case against the pay of American CEOS looks even more powerful by recognizing that the typical American company head receives greater total compensation than company heads in Great Britain, Canada, Japan, Spain, and in pretty much all developed countries. Clearly, American CEOS are much better paid than CEOS elsewhere, even when per capita incomes of the countries do not differ by very much.
Yet competition for top management can explain the rapid rise over time in the pay of the average American CEO. To understand how competition works in the management market, consider the strong and stable relation at any moment between the total compensation of CEOS at publicly traded companies, and the size of the companies they head. For every 10 per cent increase in firm size, measured by the market value of assets, by sales, or by related variables, compensation increases by about 3 per cent. This "30 per cent" law held during the 1930's, and has held for every succeeding decade, including right up to the present. Note that stock options and other forms of compensation than salaries and bonuses were unimportant until the 1970's, so this relation is not due to the rapid growth of options and compensation through shares of stock.
The usual explanation given by economists for the positive relation between compensation and firm size is that the largest companies attract the best management. Therefore, bigger companies have to pay their CEOS better in order to discourage them from going to head smaller companies. It is also socially efficient to have the best mangers run the largest companies because their greater skills then have a bigger influence since they would manage a larger amount of labor and capital. The efficient combining of better managers with larger companies in a competitive market for top managers would imply a positive relation between firm size and the total compensation package. This analysis does not explain why the 30 per cent rule holds, but it suggests that the relation between pay and size is likely to be sizable, even when top management in different sized companies do not differ greatly in skills and abilities.
We need two additional facts to explain the sharp rise in pay over time, and the much higher pay in the United States than other countries. The first is that the average size of large American companies has grown in real terms about six fold during the past twenty-five years, regardless of how "large" is measured, as long as the same measure is used consistently over time. The other important fact is that the largest 50, 100, or 500 American publicly traded companies are much bigger than the largest companies in other countries.
Clearly, if large companies pay more, and if the average size of companies has grown sharply over time, average compensation would also grow, even if the value of the increasingly generous granting of stock options and equity shares were fully understood by stock markets and boards of directors. It is also possible to understand why average compensation grew about as rapidly as average company size, although the argument here is more complicated (for the details of this argument, see Xavier Gabaix (MIT) and Augustin Landier (NYU), "Why Has CEO Pay Increased So Much?" unpublished, April 17, 2006). The allocation of better managers to larger firms, and competition for these managers among companies of different sizes, means that companies in say 2006 would have to pay more for their CEOS than even the same sized companies did in 1980, although much less than six times as much. The reasoning is that the 2006 companies of a given size are competing against relatively larger companies than comparable size companies did in 1980. Using this analysis, Gabaix and Landier are able to explain why total compensation of the average CEO of larger companies grew about six fold along with the six-fold growth in average company size during the past several decades.
The same argument explains why compensation of American CEOS is much higher than that of CEOS in other countries. Since average firm size is much lower elsewhere, their pay would be more like that of pay in the US in 1980 or 1990 than the pay of CEOS in today's much larger American firms. As the market for top executives becomes increasingly global, the pay of CEOS in other countries would rise, and that of CEOS in America might fall. For example, to attract Carlos Ghosn, a Brazilian working in France, to turn around Nissan, a seriously ailing company, Nissan had to pay him not at the low Japanese CEO levels, but at the much higher levels found in other countries.
I believe that the explanation based on the allocation of CEO talent largely is behind the explosion in compensation of American CEOS during the past several decades. Yet at the same time, some American CEOS are obviously grossly overpaid since they have mismanaged their companies, and still receive exorbitant compensations. But mismanagement is not new and probably has not become so much more important over time. So I am suggesting that the rapid growth of compensation of American CEOS, and its premium over compensation of CEOS in other countries, is not mainly due to a growth in the degree of excess payment of executives in the United States. Rather, on this interpretation, the main cause of the increase in pay is the greater challenges and opportunities facing executives who manage much larger combinations of resources.