Anshu Jain, CEO of Deutsche Bank, recently requested a pay cut.

Anshu Jain, CEO of Deutsche Bank, recently requested a pay cut.

In an ideal free market, people are supposed to get paid what they're worth. By ideal, I mean a free market with enough competition such that companies who overpaid or underpaid their employees would be punished by the success of companies who properly paid their employees.

If you can only make $2 rivets, I shouldn't pay you more than $2 per rivet. And if you can make $10 rivets, I should pay you more than a $2 rivet maker. Indeed, if I don't, someone should come along and pay you more and you should work for them.

That's the promise of a free market, simple as it is.

Some notable economists have carefully articulated how things like mininum wages break this simple rule of the market. If the government forces companies to pay 3 dollars $3 for rivets, then $2 rivet makers join the unemployed. That is, unless they can work for the government. In which case, tax payers pick up the difference.

But what about the opposite: Should there be a maximum wage?

For the same reason that minimum wages might do harm in some situations, it's easy to see how maximum wages might also do harm. If doctors got paid the same amount as lifeguards, then why should anyone spend upwards of 15 or more years of their adult life learning to be a doctor? It's hard work and you see people in the worst possible shape. Wouldn't you rather skip the school, sit at the beach, and see the more beautiful side of life? Thankfully, where increased wages signal a demand, supply follows.

But what about CEOs of companies? Forbes 500 CEOs in 2008 were paid almost 200 times more than the average worker. Meanwhile, polls in the U.S. indicate that the majority of Americans agree that CEOs should have their pay limited. Should they?

A recent article by Jacquart and Armstrong looks at the evidence. Their evidence consists of a review of numerous experimental studies and records of performance of thousands of firms and CEOs.

The conclusion is as simple as a day old bird: CEO incentives do not buy better performance. If anything, incentives make performance worse. In an effort to make a long story short, I've listed their central points and findings below.

1. Above a certain level of complexity, it's extremely difficult to predict performance. Numerous studies in the last decade have shown that, especially in financial markets, economists and self-proclaimed forecasters do not have priveledged access to the future. A quote from one of the authors puts things in perspective—it's called the seer-sucker theory: "No matter how much evidence exists that seers do not exist, suckers will pay for the existence of seers."

This isn't to downplay the knowledge of economists—they know a lot of useful stuff and we could all learn a thing or two from paying more attention to them. But what they don't know is what the financial market is going to do or how best to predict the future. They know really well, however, how best to invest given that the future looks like the past. This works, of course, until it doesn't.

2. Hiring practices for CEOs are, well, silly. Surveys of large numbers of recruitment consultants find that CEOs are often hired based on unstructured interviews and character references. Meaningful character references come from people you know. In other words, we're talking about good-ol-boy networks. Just to be clear, good-ol-boy networks are poor markets because they eliminate high quality competition.

3. CEOs are often compensated for performance increases that are beyond their control. For example, CEOs in the oil industry benefit when the price of oil increases. In other words, as the authors put it, "CEOs are paid for being lucky."  And yes, sometimes luck is just luck, not a well deserved gift from the universe.

4. CEO pay raises impair performance. Compared with similar CEOs who did not get pay raises, stocks in the companies of CEOs who did get pay raises underperformed by between 15 and 26 percent. Now we're really talking. If you pay CEOs more, you hurt your company.

5. Excessive incentives motivate fraud. Financial incentives weaken people's moral self-identity. This probably has the additional unfortunate consequence that, in the eyes of the too-incentivized CEO, the moral status of everybody else goes to zero. If you don't already know some of the evidence for this one, you haven't been paying attention. There are so many examples it's hard to choose. Lehman Brothers. The Atlanta public school cheating scandal. Bernie Madoff. Sumo wrestlers. Soccer Teams...It should be a song.

All of this is written about in more depth and with more psychological detail in Jacquart and Armstrong's article, which you can download and read for yourself here.

However, they might be wrong.

It could be, as I would say if I were the CEO of anything bigger than my office, that CEOs get paid what the market will bear. I agree with this to a degree. Remember the $10 rivet maker. If a company can't tell the difference between rivet makers, then if the company pays too much, it's the company's loss and the stockholders if they choose to keep those stocks.

But what about when the loss isn't a market loss? If stockholders lose money from fraudulent CEOs, that isn't a market loss. That's theft. If governments have to pay to police over-incentivized CEOs, that's a cost paid for by taxpayers, kids in the future, and all that. And if companies that are 'too-big-to-fail' fail, which they occasionally will because that's how life is, then CEO bonuses from these companies are basically paid for, again, by taxpayers, kids in the future, and all that.

In sum, a maximum wage may not be good for free markets. And where the market isn't free, neither should the salaries be. In any case, the evidence suggests that prudence in CEO payments would probably make things better for just about everybody—including the CEOs.


Jacquart, P., & Armstrong, J. S. 2013. The Ombudsman: Are Top Executives Paid Enough? An Evidence-Based Review. Interfaces, 43(6): 580-589.  

You can follow Thomas on twitter here.

Statistical Life

From the evolution of intelligence to memory and information overload.

Most Recent Posts from Statistical Life

Waiting for the Haiku in Mindfulness

Have we taken the bones out of Dogen so that we can pick our teeth?

Data Science and the History of Subjective Wellbeing

To put wellbeing on par with GDP, we need to understand the history of emotions.

The Future of Crowdsourcing in the Classroom

If we let students create content, they make connections we couldn't.