August 31, 2020
This simple and important question does not get anywhere near the attention it deserves. And, just to be clear, I don’t mean are they worth $20 million in any moral sense. I am asking a simple economics question; does the typical CEO of a major company add $20 million of value to the company that employs them or could they hire someone at, say one-tenth of this price ($2 million a year) who would do just as much for the company’s bottom line?
This matters not only because a thousand or so top executives of major corporations might be grossly overpaid. The excessive pay of CEOs has a huge impact on pay structures throughout the economy. If the CEO is getting $20 million it is likely the chief financial officer (CFO) and other top tier executives are getting in the neighborhood of $8-12 million. The third echelon may then be getting paid in the neighborhood of $2 million.
And these pay structures carry over into other sectors. It is common for university presidents to get paid in the neighborhood of $1 million a year. The same applies to the heads of major charities and foundations, including those that have combatting inequality as part of their mission.
This story would look very different if CEOs got paid 20 to 30 as much as a typical worker, as would have been the case in the 1960s or 1970s. In that case, CEOs would be getting $2-3 million a year. The CFOS and other top-level executives would presumably make $1.5 to $2 million, with the third tier likely getting high six figures. In that world, the presidents of universities and top executives of major foundations would likely earn $400 -$500 thousand a year.
More for the Top Means Less for Everyone Else
If it is not obvious, more for those at the top means less for everyone else. If their pay is not actually justified by their productivity, they are taking a larger chunk of the same pie. This means smaller slices for everyone else.
As I like to point out, if the minimum wage had kept pace with productivity growth since 1968, as it did from 1938 to 1968, it would be $24 an hour today. In that world, a full-time minimum wage worker would be earning $48,000 a year. A two minimum wage earner couple would be earning $96,000 a year.
While that may sound pretty good to many of us, it is not possible in an economy where the CEOs are getting $20 million, when they only add $2 million to the company’s bottom line, and the next in line get $8-$12 million, and university and foundation presidents pocket more than $1 million a year. This is a story where we would see excessive demand in the economy, leading to serious problems of inflation.
If we want to raise pay for the bottom in a big way, we have to drive down pay at the top. This would be a problem if we actually had to pay the CEOs $20 million to get them to perform well, from the standpoint of producing profits for the company or returns to shareholders, but the evidence is that we don’t.
Evidence of Whether CEOs Earn Their Keep
The best place to start on the evidence is the great book by Lucian Bebchuk and Jesse Fried, Pay Without Performance. While this book is now somewhat dated (it was published in 2004) it compiles much of the literature available at the time on the relationship of CEO pay to returns to shareholders. It includes many studies that show CEOs pay often bear little resemblance to what they do for shareholders. For example, the pay of oil executives skyrockets when the world price of oil rises, an event for which they presumably are not responsible. Another study found that CEOs tend to get big pay increases when they appear on the cover of a major business magazine, even though returns to shareholders generally lag the overall market.
A more recent study found that corporate boards seem to have stock option illusion. (This is a variation of the concept of “money illusion” which economists often argue is a problem for workers not recognizing the impact of inflation on their wages.) Corporate boards did not reduce the number of options issued to CEOs and top executives in the 1990s, even as the rising stock market caused their value to soar. Another study of a large number of companies over a ten-year period found a negative relationship between CEO pay and shareholder returns, implying that the high pay of CEOs and other top executives was coming at the expense of returns to shareholders.
A couple of years ago, Jessica Schieder and I did a study where we looked at the impact of the Affordable Care Act’s (ACA) limit on the deductibility of CEO pay for health insurance executives on their pay. Since the corporate tax rate at the time was 35 percent, the ACA’s cap on deductibility effectively raised the cost of CEO pay to the company by more than 50 percent. (Before the ACA, a dollar of additional CEO pay cost the company 65 cents. When it was no longer deductible, it cost the company $1.00.)
We tried every plausible specification for regressions, controlling for profit growth, revenue growth, stock appreciation, and anything else that might reasonably be expected to affect CEO pay. We could not find any evidence that the loss of the deduction had any negative impact at all. This means that each dollar of CEO pay cost the company 50 percent more, the companies were still paying CEOs just as much as before.
If we accept that CEOs are not producing returns to shareholders consistent with their $20 million paychecks, the question is why? After all, would we expect to see companies paying cashiers or dishwashers $200,000 a year when the going rate in other companies is one-tenth as much?
Why Is CEO Pay Out of Whack?
The problem is with the governance structure of the corporation. CEO pay is most immediately determined by corporate boards, who largely owe their jobs to top management. Furthermore, keeping their jobs depends almost entirely on keeping other board members happy. Board members who are nominated for re-election win well over 99 percent of the time. Since these jobs typically pay several hundred thousand dollars a year for a few hundred hours of work, board members generally want to keep their jobs.
One sure way of pissing off other board members is asking questions like, “can we get another CEO, who is just as good, for half the pay?” It is a safe bet that this sort of question is almost never asked in corporate board rooms, even though this is supposed to be precisely the question that they should be asking all the time.
In principle, the board’s primary responsibility is to ensure top management is not ripping off shareholders. In reality, their allegiance is instead overwhelmingly to top management, as detailed in Steven Clifford’s great book, the CEO Pay Machine. Clifford was the CEO at a mid-sized company who later sat on several corporate boards, so he has much first-hand experience of the process.
In this story, the problem is that we have the CEO’s pay, and that of other top executives, being determined by a board that is loyal to them rather than shareholders. The most obvious remedy is to find a mechanism that will give more control of the CEO’s pay to shareholders, so that it can be brought back down to earth.
To be clear, I am not talking about fundamentally changing control of corporations. Many people, including Senator Elizabeth Warren, have proposed giving workers a substantial say on corporate boards. This is a reasonable proposal. Germany has had a system in place for decades that gives workers half the seats on corporate boards, with no obvious ill-effects for its economy.
But this is not the issue I am raising here. I am simply saying that shareholders should have more control over what the CEOs and other top executives get paid. My personal favorite is a modest change to the “Say on Pay” referendums that were put in place in the Dodd-Frank financial reform legislation.
Under this provision, the shareholders get to vote on the compensation package for their CEO every three years. This is a simple yes or no proposition. There is no direct consequence for the CEO, the board, or the pay package of losing a say on pay vote. It is non-binding. As it stands, more than 97 percent of pay packages are approved.
The rules for Say on Pay could be changed so that there are explicit consequences. Suppose the directors sacrificed their own pay if a CEO’s pay package was voted down. With the vast majority of pay packages being approved, most directors would likely think they have nothing to fear. However, when a few packages did get voted down, it is likely that directors would start to ask whether they could get away with paying their CEO less. This could put some serious downward pressure on CEO pay.
This may not be sufficient to get CEO pay back to earth, but it seems like a good place to start. From a political perspective, it seems hard to argue with the idea that shareholders should be able to determine what they pay their CEO and top management.
If this was successful in bringing down CEO pay, we would be in a very different world. As noted earlier, we would see pay scales at the top reduced across the board. Not only would the top echelons of the corporate hierarchy get paid less, but we would see lower pay at the top of other institutions as well.
It is sort of striking that conservative economists can get completely bent out of shape over the idea that some workers may get paid a dollar or two above the market wage, because of the minimum wage. But, few progressive economists seem especially troubled by CEOs getting paid many millions above the market wage because of a corrupt corporate governance structure. Paying a bit more attention to the market here might be a good idea.
In response to the mass protests following the police killing of George Floyd, there has been a renewed interest in opening doors to Blacks and other disadvantaged groups so that they have more opportunities to get higher paying and higher prestige jobs. While these efforts are important, seeing the little progress we have made in the last half-century, it is hard to believe that longstanding barriers will be eliminated any time soon.
As much as we need to eliminate racism and other forms of discrimination, the consequences will be much less in a world where the university president is getting $400,000 a year and the custodian is getting $48,000 a year than the world we have today. There is certainly no excuse for rigging the market in ways that redistribute money from everyone else to those at the top. This disproportionately hurts Blacks and other victims of discrimination now, but even if we managed to somehow eradicate racism there is no reason to rig the system so that it needlessly forces so many people to live miserable lives.