LSE - Small Logo
LSE - Small Logo

Alexander 'Sandy' Pepper

November 15th, 2022

The market failure approach to executive pay

0 comments | 18 shares

Estimated reading time: 3 minutes

Alexander 'Sandy' Pepper

November 15th, 2022

The market failure approach to executive pay

0 comments | 18 shares

Estimated reading time: 3 minutes

Thirty-five years of empirical evidence have failed to establish a strong link between executive compensation and firms’ financial performance. Sandy Pepper writes that it’s time to change the prevailing paradigm on executive pay. He says inflation in top pay is a “mad, bad system” and an important ethical problem.


 

Join Sandy Pepper, LSE’s  Elisabeth Stheeman, Eva Micheler, and Alex Voorhoeve, plus Bloomberg’s Katherine Griffiths in a public lecture at LSE on 21 November 2022.

Ever since Michael Jensen and William Meckling explicitly linked agency theory with executive compensation in their famous article Theory of the firm in 1976, this way of conceptualising top pay, now known as ‘optimal contracting theory’, has been the dominant way of thinking in academic circles.

I think it’s time for a new paradigm.

The main problem for optimal contracting is that empirical evidence gathered over 35 years has failed to establish a strong link between executive pay and firms’ financial performance. Most economists now appear to accept that the most significant empirical correlation is between executive pay and firm size, not between executive pay and financial performance, as predicted by agency theory.

Optimal contracting theorists argue that a strong correlation between CEO pay and firm size is not necessarily inefficient. Big companies are presumably more complex to run than smaller companies. They must attract the best management talent in order to operate efficiently. Thus it makes sense for them to provide the largest pay packets.

But there is something wrong with this argument. To accept ex post that a correlation between CEO pay and company size is an acceptable outcome, when aiming ex ante for a causal connection between CEO pay and firm performance, is a flawed attempt to rescue a theory in the face of falsifying evidence. I call it the ‘But I’m Still in Scotland’ fallacy.

An absent-minded professor sets off to drive to Edinburgh where he is speaking at a conference. On the way he takes a wrong turning and ends up in Glasgow by mistake.  On the telephone to the chair of the conference to explain his error, he argues that this is nevertheless an acceptable outcome, as he has still arrived in Scotland. Really?

The market failure approach to executive pay

Economists have known for a long time that labour markets are different from other commodity markets. This is particularly true of the market for the people that the French economist Thomas Piketty describes as ‘super-managers’. An efficient market requires many buyers and sellers, homogenous products or at least good substitutes, free market entry and exit, plentiful information, and little economic friction. The problem with the market for super-managers is that practically none of these conditions hold good.

Isomorphism

Because executive labour markets fail to provide effective price signals, the non-executive directors whose job is to determine the remuneration of top managers seek alternative ways of resolving the uncertainty which they face in deciding how to determine top pay. In 1983 economic sociologists Paul DiMaggio and Walter Powell described how three ‘isomorphic processes’ come to operate in response to such uncertainty, in their article The iron cage revisited. ‘Isomorphism’ describes a process whereby social practices develop similar forms over time. Remuneration committees copy the pay strategies of other comparable organisations (known as ‘mimetic isomorphism’). Companies are constrained by laws and codes of practice established by government and regulators (‘coercive isomorphism’). They seek advice from remuneration consultants who benchmark pay data and recommend standard solutions (‘normative isomorphism’).

The remuneration committee’s dilemma

Remuneration committees also face a ‘prisoners’ dilemma’ as they seek alternative ways of rationally determining top pay. As a result, they pay over the odds, in the vain hope that they might attract one of the better super-managers and avoid the worst. Offering higher pay becomes the dominant strategy, even though by doing so, companies will generally be no better off than if they all paid more moderate salaries. In more typical labour markets, with greater numbers of participants and more ready substitutes, the same pressure to pay over the odds doesn’t arise.

The investors’ collective action problem

While it is in the interests of minority shareholders in public companies to monitor the activities of managers, they will want to do so at minimal cost. They will certainly wish to avoid incurring monitoring costs that materially eat into their income and gains. While a £3.25 million bonus paid to the CEO of a FTSE100 company might seem a lot of money, to a large investment management firm with £50 billion of assets under management, holding, say, one per cent of the company’s shares, the amount involved is relatively trivial, especially if the question is about whether the CEO’s bonus is ten or twenty or even thirty per cent higher than it should be.

Investors have historically been prepared to accept rent-seeking behaviour by managers as long as their reasonable expectations of dividends and capital gains are met, because they expect the cost of intervention to exceed any individual benefit.  It is only in recent years that a small number of institutional investors have started to take companies to task on executive pay.

The LTIP valuation issue

A further reason why CEO pay in the UK and US has increased so much more rapidly than average earnings is due to the delivery mechanism. If you were to offer an executive £100,000 in cash or a share-based performance-related long-term incentive (an ‘LTIP’) with an economic value of £300,000 don’t be surprised if he or she would prefer to take the cash. By the time they have applied subjective probability-based discounts for uncertainty of around seventeen per cent, and time discounts of thirty three per cent per annum, the subjective value which the executive attaches to the LTIP may be as little as one third of its economic value. A consequence of providing people with an asset they do not fully value is that they want more of it to compensate for their subjective ‘loss’ in comparison with less risky, more certain, and more immediate forms of reward.

Isomorphism, the remuneration committee’s dilemma, the investors’ collective action problem, and the LTIP valuation issue operate independently and may be additive—all four effects could be at work at the same time. From a public policy perspective, none are easy to solve through conventional means. It is why I believe that government, companies, investors and executives all need to recognise inflation in top pay for what it is—an important ethical problem. When it comes to senior executive reward, for too long companies have behaved as if they are in the equivalent of an arm’s race. It is a mad, bad system, and it needs to change if inflation in executive pay is to be brought under control.

♣♣♣

Notes:

About the author

Alexander 'Sandy' Pepper

Alexander 'Sandy' Pepper is Emeritus Professor of Management Practice at LSE. 

Posted In: LSE Authors | Management

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.