FairShare

Part One · The Case

Why the Case for Top Pay Fails

In 1965 the chief executives of America’s largest companies were paid 21 times what a typical worker earned. In 2023 they were paid 290 times. Between 1978 and 2023 their pay rose by 1,085 percent, and a typical worker’s by 24 percent.1

Most of us have met figures like these before, and many of us are unmoved by them. That is not stubbornness. Nobody disputes that the gap exists. The dispute is over whether it is deserved, and the belief that it is deserved runs deep. It is a natural thing to believe. We watch a person rise to the head of a great enterprise and we conclude that they must be worth what they are paid, and that the pay is the measure of what they add.

A belief held that widely deserves to be taken seriously, and the way to take a belief seriously is to test it. This is the right place to do so. If one person’s contribution can be measured anywhere, it should be at the top, where the stakes are largest and where more effort has gone into the measuring than anywhere else. So this chapter sets out the five strongest arguments for top pay, as their best defenders make them, and looks at each in turn.

Nothing here says that leaders do not matter. They do, and in the model this paper proposes a chief executive sits at the top of the salary ladder. The question is whether a multiple in the hundreds measures anything at all.

The argument from scale

The claim. A chief executive’s decisions are multiplied by the size of the company. A decision that is one percent better at a company worth $100 billion is worth $1 billion. It is therefore rational to pay a great deal for a slightly better decision-maker, and as companies have grown, pay has grown with them. The economists Xavier Gabaix and Augustin Landier gave this argument its definitive form: the sixfold rise in American chief executive pay between 1980 and 2003, they showed, matches the sixfold rise in the market value of large companies.2

The test. This is the best argument there is, and it repays a careful look at what it concedes. Gabaix and Landier worked out how much the top chief executives differ in talent, and their answer is remarkable. The difference is tiny. Replace the 250th-best chief executive in the country with the very best, and by their own calculation the value of the company rises by 0.016 percent. Yet the best is paid more than 500 percent more than the 250th.

The model is consistent. On a company worth $100 billion, 0.016 percent is $16 million, and sums of that kind are what the bidding is about. But notice what has been conceded along the way. The people are close to interchangeable. The pay differs because the companies differ in size. The money follows the seat.

Two further facts tell against treating this as a law.

  • It did not hold for thirty years. From the late 1940s to the 1970s American companies grew enormously and executive pay stayed almost flat. The link between company size and pay in those decades was between a tenth and a third as strong as it later became.3 The companies of 1960 were not badly run for want of pay. Something other than scale changed afterwards.
  • Scale multiplies everyone. If the size of the company magnifies a small edge in the chief executive, it magnifies a small edge in the engineer whose design ships to a billion devices, and the cost of a mistake by the person who maintains the data center. The scale belongs to the company, which thousands of people built. The argument hands the whole multiplier to one position.

The argument from performance

The claim. Most top pay is in shares. The chief executive gets rich only if the shareholders do, so the pay is a share of value the person demonstrably created.

The test. A share price moves for a great many reasons, and most of them are not the chief executive. Marianne Bertrand and Sendhil Mullainathan found an ingenious way to check. They looked at luck: changes in a company’s fortunes that its boss plainly did not cause, such as a rise in the price of oil for an oil company. Chief executives, they found, were paid as much for lucky gains as for any other kind.4 Later work found that the luck ran one way. Executives were rewarded for good luck and not penalized to match for bad.5

Pay in shares, then, rewards holding the position while the market rises. That is a different thing from producing the rise. And where the rise was produced, the whole company produced it.

Apple offers a clean example. Tim Cook was paid $99.4 million for 2022, about $83 million of it in share awards. Only 64 percent of shareholders voted to approve, a poor result for such a vote. For 2023 the board cut his target pay by more than 40 percent, to $49 million, and it did so on Cook’s own recommendation.6 He was paid $63.2 million for that year. If $99 million had measured what the work was worth, a cut of a third would have been a loss to someone. The board, and the man himself, judged that the job could be done for much less.

The argument from scarcity

The claim. Almost nobody can run a company of 150,000 people. The pay is what it takes to get one of the few who can.

The test. The scale argument has already answered part of this: on its own figures, the best candidates are nearly indistinguishable. The other part is how much difference any chief executive makes. Researchers have tried to measure the share of a company’s performance that belongs to its chief executive, as distinct from its industry, its history and the year. They disagree. One analysis found that only 4 to 5 percent could be clearly separated from chance; others dispute the method and find a larger effect.7 Even on the larger estimates, most of what a company achieves is owed to something other than the person who leads it.

There is also a way to test the claim that needs no statistics. Take half of what Apple paid its chief executive for 2022, about $50 million. That sum would pay for more than thirty-five senior people, each given a million dollars’ worth of the best training available and a salary of $350,000. It would pay for another thirty-five the next year, and the year after that.

The usual reply is that a company cannot have dozens of chief executives, because the final decision does not divide. That is true, and the test does not need it. Apple keeps its chief executive, at $50 million. The claim being tested is that the second $50 million does more good in that one person’s pay than it would do anywhere else in a company of about 160,000 people. Put that way, it is a hard thing to believe.

The argument from accountability

The claim. The chief executive carries the risk. When the company fails, they are the one who is fired and blamed, and high pay is the price of bearing that.

The test. Look at what happens when they are fired. Dennis Muilenburg was removed as chief executive of Boeing in December 2019, after two crashes of the 737 MAX killed 346 people. He received no severance and gave up $14.6 million in awards. He left with $62.2 million in pension, deferred pay and long-term awards.8

One case proves little. But the one-sidedness of luck, described above, is the general pattern. Top pay rises with good fortune and does not fall to match with bad. The people who carry the real risk of a company’s failure are the ones with a single income, no package and a mortgage. Nobody pays them a premium for it.

The claim. It is the shareholders’ money. They agree to pay it, they could stop, and nobody else has standing to object.

The test. Consider how the figure is arrived at. A committee of the board hires a consultant, who assembles a list of comparable companies. The committee then typically sets pay at or above the middle of that list. Research on these peer groups finds that they are built in ways that push pay upward.9 And when every board aims at the middle or higher, the middle rises every year.

Shareholders do vote, but the vote is advisory and comes after the fact. In 2024 the average pay package at a large American company was approved by about 90 percent of votes, and around one in a hundred failed.10 An owner of a fraction of a percent of a company has little reason to study its pay report, and less to fight it.

This explains why nobody stops top pay from rising. It does not show that the pay measures anything. And the consent involved belongs to the people who own the shares. Those who did the work were not asked.

“They earned it”

Beneath all five arguments lies something simpler and more human: the feeling that the people at the top worked hard, gave things up, and deserve what they have.

Many of them did work hard. So did the nurse on nights and the driver on a double shift. Effort and sacrifice are found at every level of a company, and nobody has 290 times as many hours in the day. If pay followed effort, the ladder would be nearly flat.

Desert also depends on being able to say what a person is responsible for. That returns us to the software, and the chord.

What is left

Leaders matter, and nothing in this chapter says otherwise. A good chief executive is better than a bad one, and belongs at the top of the ladder.

What the evidence does not support is the size of the gap. On the best available account, the people competing for the top jobs are nearly identical in ability. Their pay follows the size of the company, the direction of the market and the way boards set pay by looking at one another. Thirty years of flat pay show that none of this is a law of nature. It is a habit, and habits can change.

A chief executive’s signature goes on decisions that thousands of people prepared, tested and will carry out. The value is made together, and the pay is assigned to one name.

If one person’s contribution cannot be measured at the top, where the most effort has gone into measuring it, it cannot be measured anywhere. Chapter 5 takes the same argument to the other end of the ladder. First, Chapter 3 shows that the gap in pay understates the gap in what people end up with.

Notes

  1. Economic Policy Institute. (2024). Report on chief executive compensation at the 350 largest U.S. public firms in 2023. https://files.epi.org/uploads/288934.pdf
  2. Gabaix, X., & Landier, A. (2008). Why Has CEO Pay Increased So Much? The Quarterly Journal of Economics, 123(1), 49–100.
  3. Frydman, C., & Saks, R. E. (2010). Executive Compensation: A New View from a Long-Term Perspective, 1936–2005. The Review of Financial Studies, 23(5), 2099–2138.
  4. Bertrand, M., & Mullainathan, S. (2001). Are CEOs Rewarded for Luck? The Ones Without Principals Are. The Quarterly Journal of Economics, 116(3), 901–932.
  5. Garvey, G. T., & Milbourn, T. T. (2006). Asymmetric Benchmarking in Compensation: Executives Are Rewarded for Good Luck but Not Penalized for Bad. Journal of Financial Economics, 82, 197–225.
  6. Apple Inc. (2023). Proxy statement for the 2023 annual meeting of shareholders. Reported in: Apple CEO Tim Cook Taking Substantial Pay Cut in 2023 After Earning Nearly $100 Million Last Year. (2023, January 12). MacRumors. https://www.macrumors.com/2023/01/12/tim-cook-taking-pay-cut-in-2023/
  7. Fitza, M. A. (2014). The Use of Variance Decomposition in the Investigation of CEO Effects: How Large Must the CEO Effect Be to Rule Out Chance? Strategic Management Journal, 35(12). For the reply: Quigley, T. J., & Graffin, S. D. (2017). Reaffirming the CEO Effect Is Significant and Much Larger than Chance: A Comment on Fitza (2014). Strategic Management Journal, 38(3).
  8. Boeing says no severance package for ousted CEO Muilenburg. (2020, January). Agence France-Presse. https://techxplore.com/news/2020-01-boeing-severance-package-ousted-ceo.html
  9. Bizjak, J., Lemmon, M., & Nguyen, T. (2011). Are All CEOs Above Average? An Empirical Analysis of Compensation Peer Groups and Pay Design. Journal of Financial Economics, 100(3), 538–555.
  10. Harvard Law School Forum on Corporate Governance. (2024). 2024 Proxy Season Review: Compensation-Related Matters. https://corpgov.law.harvard.edu/2024/09/19/2024-proxy-season-review-compensation-related-matters