Why are CEO salaries rising?

Here is one good bit from the newly published Kaplan and Rauh paper “It’s the Market: The Broad-Based Rise in the Return to Top Talent“:

The evidence is not supportive of the arguments that the top incomes have been driven by managerial power or poor corporate governance in public companies.  Public company executives, who should be more subject to problems of managerial power problems, saw their pay and relative standing increase less over this period than executives of closely-held company businesses that are, by definition, controlled by large shareholders or the executives themselves and are subject to more limited agency problems.  Furthermore, the Bakija, Cole, and Heim (2012) findings are not consistent with loosening social norms being an important factor in the increase in incomes at the top, as it is the pay of closely-held businesses — where executive pay is private and undisclosed — that increased the most.

The paper is interesting throughout.


I don't know about you, but whenever something happens which I do not like, the cause is obvious:



1) It appears that powerful executives at all companies are getting richer and when compared across public and private corporation structures, the private governance structures increased the executive pay more.

2) At no point in this paper is there any exploration of the relationship of executive pay vs. performance of the company they manage.


Powerful executives in organizations reward themselves and others in their inner circle as generously as possible given the restrictions of their corporate organizations. When not exposed to external inspection or significant ridicule, they will be able to increase their compensation significantly beyond their public counterparts. Overall governance structures to restrict this behavior are universally ineffective, more so in the private sector.

Direct comparisons of business executives vs. athletes in an attempt to justify executive pay increases by claiming those executives are “superstars” and therefore should be compensated like individual athletic performers ignores fundamental differences between the job descriptions and appears only to be used as an illogical attempt to justify corporate executive compensation.

> Powerful executives in organizations reward themselves and others in their inner circle as generously as possible given the restrictions of their corporate organizations. When not exposed to external inspection or significant ridicule, they will be able to increase their compensation significantly beyond their public counterparts. Overall governance structures to restrict this behavior are universally ineffective, more so in the private sector.

Chalking it up to "Powerful People Reward Themselves" misses the question. What has changed lately? Why is inequality increasing now? Are they able to reward themselves more effectively? How come?

How effective can any review of compensation be when the review is done by other like minded business executives? Why would I not increase the pay of my fellow irreplaceable hard working executives when they are going to do the same for me? After all, the pay may be outrageous compared to a worker, but the actual cost to the company is not that bad and is spread out over a lot of stakeholders. There is no need for explicit collusion in any of these processes, only a single shared vision of entitlement.

This was not needed in the past because the executives of the past were not isolated from the realities of the regular people, they lived and worked around them to a greater extent and they were not bombarded with how great they are and how much every other executive is getting rich when they are not. This introduces expectations that can only be fulfilled by more money. Compare the attitudes of George Romney to his son for a perfect example of entitlement.

The way to stop it is to break the shared executive vision. If you have board members representing non business interests these would not be easily rubber stamped. Stick worker representation on the board, add in members from the clergy, anything that dilutes the vision with the reality on the street and in the actual company.

Not going to happen in the real world though...

A failed study, and one which you can pick another control group to find just the opposite conclusion.

I have represented Fortune 500 companies, and seen exeuctive salaries go through the roof just following enactment of poison pill provisions. I also know colleagues who have become inside general counsel, making over a million a year, while they were in private practice and making far less than that the year before.

Now, as to the comparable: I have also represented cooperatives and non-profits. For the cooperatives, in particular, the members are far, far more active and involved than traditional shareholders in exec comp. They have kept a cap on executive salaries, and executive salaries are even sometimes discussed at membership meetings.

A variant of a cooperative is a mutual insurance company. If you look at mutual insurance companies you will also find lower comp levels, and, more interestingly, if the mutual insurance company is demutualized (i e , converted to a stock company), exec salaries again go through the roof.

(By the way, yearly comp isn't the only pay...look at retirement packages and deferred comp contributions, memberships, uses of facilities, and consulting contracts).

If someone is interested in doing a paper and research on this, start looking at coop exec salaries, mutual insurance salaries, and even non-profits (although some non-profits are also getting out of line). The comparables are interesting, and vary with the amount of membership involvement.

If you read past the introductory paragraphs, it says that the authors go for the "superstar" or winner-take-all explanation for high CEO salaries. Well, if they really are superstars, then the high CEO pay must correlate with high return on investment, yes? Except it doesn't.

It seems to me that the authors don't bother distinguishing between the ideal of publicly traded companies (i.e., many individual shareholders able to exercise control over CEO remuneration) and the reality (i.e. most shares held by large corporations whose CEOs are themselves highly remunerated, and for that reason happy to accept high CEO wage in others). If you take that into account, and the fact that societal norms are rather different amongst the CEOs of Fortune 400 companies than the rest of us, then it's simply the case that CEO remuneration is what they can convince a captive remuneration board to approve. You only have to read a bit of "Inside Job" which, in passing, tells the story of a well-remunerated CEO more interested in golf and blackjack than running the company to realize that CEO pay is unrelated to their worth.

This paper is just another example of economists trying to justify what happens rather than seriously analyzing why it happens. If you take a broader view, you can see that those at the top have simply succeeded in capturing a larger fraction of GDP without in any way increasing growth - perhaps, even reducing it. It's the opposite of the 60's and 70's where it was union power which captured the fraction of GDP for workers.

> If you take a broader view, you can see that those at the top have simply succeeded in capturing a larger fraction of GDP without in any way increasing growth – perhaps, even reducing it.

This doesn't really contradict the superstar theory. Baseball players today could be objectively worse than those in the past, but command much higher salaries because of increasing media reach, merchandising, etc

Then why has compensation for managers of privately held businesses increased faster than publicly held ones?

Haven't read it, but I wonder how they define private. After all, I worked for a company that went public->private->public within a year because it got bought out by a private equity chop shop that loaded it with debt at torpedoed the business but made a lot of money in the process.

Yes, it's just another sign of the corruption of our profession that we have another paper concluding that CEO's earn their pay because of their productivity when the analysis actually shows no such thing.

First, I think it's useful to distinguish between the winner-take-all earnings that go to entrepreneurs in, say, technology (which seems to be much more justifiable) vs. the very high pay of hired executives and managers.

Second, let's say executive pay is justified by corporate and stock price performance (however defined). Isn't there still room to argue that no matter how financially justified a CEO is making $20 or $50 million a year, there could be (but no longer is) a social or cultural norm that restrains corporate boards and executives from paying themselves such high salaries and benefits?

Charles Murray in his book Coming Apart notes that in the 1950s and early 1960s many executives living outside of New York and Los Angeles thought it was too flashy to drive Cadillacs (let alone foreign makes such as Mercedes) so they drove Buicks instead.

In 1950, GM CEO Charlie Wilson was paid about $650,000, or about $5 million in today's money. The current GM CEO, Dan Akerson, made over $11 million in 2012. I'm not picking on Mr. Akerson as GM is making some of the best cars in its history and there are many CEOs who make much more than him.

Still, it's worth thinking about since GM in 1950s America was the Apple of its day. Yes, Steve Jobs was more an entrepreneur than a manager, but it seems norms about pay were simply more restrained and modest back then.

And the US had a booming stock market from the late 1940s to mid-1960s when stock returns nearly matched that of the 1982-2000 stock boom.

Wait, what am I missing? Am I just a worse reader than the rest of the commentators?

If the CEOs of public companies have so much power, why is their pay rising *less* than the CEOs of private companies? I hear a lot of anecdotes in the comments, saying it's a flawed study. What is flawed? Where is it flawed? By definition, if the shareholders of a private company don't like a CEO in the morning, he's out in the afternoon. Period. Only, that's not what's happening. They're paying the CEO's of private companies more. Or, rather, they're increasing private-CEO pay faster than public CEO pay.

So, the "public co. CEOs reward themselves and their insiders" meme doesn't stand up. Could we please come up with an alternative rather than stamping our feet and shouting?

Good point. Is this a case of the "mood-affiliation fallacy"? Bias is a powerful thing ...

" if the shareholders of a private company don’t like a CEO in the morning, he’s out in the afternoon"

But that's seldom what actually happens. Look at, for example, Carly Fiorina during the collapse of HP. She eventually got removed, but not during the period when HP was losing half its value.

I suspect that privately held companies are subject to much more detailed oversight than publicly held ones, although it's difficult to assign a numerical metric to this.

A more interesting question: why does the CEO never get outsourced, or replaced with a much cheaper H1B? You could save millions by only sacking one employee.

CEO's must have been really stupid: Their pay didn't appear large until some decades ago. I guess they had been slow learners in exploiting agency problems.

Social science invents social problems.

Ha! I always remember what an economist (Luke Froeb) once said: "If there is no solution, there is no problem."

I have a hard time believing that there actually is weaker corporate governance, and less oversight of CEO pay, than in the days before 1980 (1990, even?) when one had to call the investor relations department (and no Googling the phone number) and request company filings in the mail.

It does not follow that because it was harder to monitor investments back in the days before 1980, corporate governance would necessarily be weaker back then. It may be that, because it was harder to monitor investments and harder to invest in general, there was a smaller class of investors overall and that this smaller class was more knowledgeable and more demanding of management than is the current crop of investors. Management's typical answer to unhappy investors today is "sell your stock." This may fly in the investment climate of today where most ordinary investors are encouraged to buy mutual funds or ETFs, but it probably did not work as well back in the day.

It's interesting how TC fails to mention the article that follows Kaplan, Rauh in this month's Journal of Economic Perspectives.

Here's the title, abstract, and link:

The Pay of Corporate Executives and Financial Professionals as Evidence of Rents in Top 1 Percent Incomes

Josh Bivens and Lawrence Mishel

The debate over the extent and causes of rising inequality of American incomes and wages has now raged for at least two decades. In this paper, we will make four arguments. First, the increase in the incomes and wages of the top 1 percent over the last three decades should be interpreted as driven largely by the creation and/or redistribution of economic rents, and not simply as the outcome of well-functioning competitive markets rewarding skills or productivity based on marginal differences. This rise in rents accruing to the top 1 percent could be the result of increased opportunities for rentshifting, increased incentives for rent-shifting, or a combination of both. Second, this rise in incomes at the very top has been the primary impediment to having growth in living standards for low- and moderate-income households approach the growth rate of economy-wide productivity. Third, because this rise in top incomes is largely driven by rents, there is the potential for checking (or even reversing) this rise through policy measures with little to no adverse impact on overall economic growth. Lastly, this analysis suggests two complementary approaches for policymakers wishing to reverse the rise in the top 1 percent's share of income: dismantling the institutional sources of their increased ability to channel rents their way and/or reducing the return to this rent-seeking by significantly increasing marginal rates of taxation on high incomes.


Indeed, winner take all outcomes are viewed as being a form of rents and thus not tied to efficiency. That these high salaries are not tied to better corporate performance, whether they are privately or publicly owned, is simply not surprising.

Another notable point simply ignored by everybody so far is the fact that these CEO salaries have soared far more in the US than in most other countries, although, again, it is far from obvious that US corporate performance is all that much better, or is better than it used to be when CEO salaries were much lower.

My personal guess is that we are just dealing with ignorance. Back in the day supply lines were short, personal connections were obvious, prices were clear, and you had a dozen other firms to compare CEOs and see what happened if you went with the bargain rate CEO option.

Today, supply chains are deep and nigh unto incomprehensible. Personal connections are vastly more important, but you need to know the right wonks and staffers as well as the right congressmen, CEOs, and labor leaders. Prices are ever less clear as more and more of the economy swaps over customized and service pricing. And, of course, we've had mass consolidation and niche targeting so there are endless internet arguments about who is actually comparable - if anyone. You have ever smaller safety margins everywhere (no warehousing, fewer slack man hours, and fewer facilities).

So say you just bought up a major of a generic major corporation worth 10 billion. How will you rate potential CEO inputs? How much will it *cost* to accurately measure them? Even assuming you can for a reasonable price, how much risk are you taking with the bargain option? If you could quantify senior executives, how will you price marginal risk? If there is only a 5% chance that the bargain option will do worse than the gold plated option, just how bad does the potential downside have to be before you say, screw it I'm paying the extra few million per annum as insurance?

People and investors tend to be risk averse particularly in ambiguous situations. CEO pay may be objectively too high, but that is hard to measure and harder still to justify with the possibility of major downside risk. In practice, this should mean that executive pay should be the most obscene for the firms that have the most ambiguity measuring a CEO's utility and at the firms where low likelihood/high impact downsides are the most likely.

"Furthermore, the Bakija, Cole, and Heim (2012) findings are not consistent with loosening social norms being an important factor in the increase in incomes at the top, as it is the pay of closely-held businesses — where executive pay is private and undisclosed — that increased the most."

This statement likely contradicts itself. Either the pay levels of execs at privately-held companies cannot be reliably obtained, in which case rising norms are not driving pay rates up but the authors of the study have no idea what's going on at closely held companies and thus cannot comment on them whatever.

OR other execs can get reliable pay info at closely held companies via the very same methods the researchers used and thus rising norms could be a real factor in driving their pay rates up.

I'd like to see a detailed empirical study of whether economists suck up to the rich more today than in the past.

Hate to say this.. nut Steve wins the thread...

Many of the comments above are founded on nothing more than envy. It is probably hard wired in humans to look for equality in desserts, on account when we were living in bands of 40, all production was joint production and getting fat was evidence of cheating. Possibly all human intelligence is founded on detecting such cheating. [Which would explain why that intelligence is so wanting.]

Friends, equality may have been efficient 40,000 years ago, but it ain't today.

And your comment is founded on nothing more than stupidity.

It's not "hard wired in humans to look for equality in desserts". Those prehistoric "bands of 40" didn't have "equality in desserts". Different talents and skills were recognized, the bravest and greatest hunters or warriors were glorified and received high status and material rewards, etc.

What's hard wired is a sense of fairness. Since this rise in CEO pay is just an increase in rents and not tied to efficiency, it appears our basic intuitions about fairness are quite accurate.

Touchy, touchy. "What's hardwired is a sense of fairness." And we can tell that it's a sense of fairness and not envy that motivates you, because of course.

Top executive talent is difficult to come by not so much because of necessary brainpower, but because of experience, insight and communication skills required. Those take time to develope and can be acquired only by serving in other executive positions. True, most of highly paid CEO don't outperform their peers, but NBA players don't win more than 50% of the games on average either.

Yes, it's a sense of fairness, not envy, that motivates me. I don't have a problem with people being paid more than others.

The rise in CEO pay is just an increase in rents and not tied to efficiency. I'm sure these top executives are fantastic and all, but it's not showing up in greater productivity or efficiency.

“equality in desserts”

I think we have found the title of Cowen's next food book.

"it is the pay of closely-held businesses — where executive pay is private and undisclosed — that increased the most"

While I don't doubt the authors words, I just wonder how they know this if the pay is "private and undisclosed?"

For those being consumed with class rage and envy, I do not understand why they do not target their ire and their envy to large stock holders.
Let's say the CEO of a global multinational that works in 10 countries with billions of dollars in revenue (e.g. GM CEO from above) does not deserve the 20 million dollars. Why should the idle rich (e.g. an upper east side layabout), with enough GM stock to get an appreciation of 50 million dollars, deserve it more?

Many people defend high executive compensation on the following grounds: if a company's market capitalization increased from $1 billion to $15 billion during a CEO's tenure, what wrong with paying the CEO, say, 1% of that gain, or $140 million.

But how much of the increase in a company's value should be attributed to a manager's skill?

The U.S. enjoyed a tremendous stock market boom from 1982 to 2000. Many companies' market values skyrocketed as sales and earnings rose. Sales and earnings increased in part surely from genuine managerial skill but also in part because the overall economy recovered from the stagflation of the 1970s.

Also, a significant part of the increase in companies' stock prices was the result of the increase in earnings multiples, or the amount investors were willing to pay for every dollar of earnings (the P/E ratio). Again, some of the increase in the P/E ratio reflected greater investor optimism in the company's management and prospects, but some of it surely reflected euphoria in the latter stages of a stock boom about to go bust.

A lot of people look like geniuses when stock prices go up and up.

Companies over x dollars market cap or y dollars profit or z dollars total exec pay should be required to have a dividends representative for shareholders, where someone is paid real money to try to pry exec compensation out of exec hands and place it into dividends, so the people who actually own the company can have their interests more directly represented.

I have absolutely no idea how to second guess executive pay. How in the world are people coming up with "the right number" such that they are comfortable saying what all the stakeholders are approving is out of bounds somehow. I'm never going to defend any particular cases, but man unless someone can point me to rents (which may be the case in some industries but is difficult to defend broadly) or regulatory capture or whatever, I don't understand how so many people can feel so confident looking at levels and drawing the conclusion they know the price of executive labor better than all of the stake holders.

If the tax rates on income were more appropriate (say the P & S estimates of optimal), this would be less of a problem. Agree with SS that idle wealth should be taxed away at higher rate though.

Comments for this post are closed