Against high CEO pay

Imagine we lived in a feudal society in which lords exploited peasants. A defender of the system might argue that wealthy lords perform a useful service; they protect their peasants from invasion and theft thus giving them security and a little prosperity. And competition between lords should improve these services; bad lords will find their lands and peasants seized by better lords who become wealthier as a result.  

Such an argument would, however, miss the point. The case against feudalism is that the system as a whole is unjust and inefficient. The fact that good lords provide services and are richer than bad ones is quite compatible with this.

This analogy came to mind whilst reading Ben Ramanauskas’s defence of high CEO pay. The fact that good CEOs make a positive difference to a company does not in itself justify a system in which CEOs in aggregate – many of whom are far from good – get fortunes.

Two big facts suggest that such a system might well be inefficient.

One comes from Rene Stulz and Kathleen Kahle. They point out that there are now fewer stock market-listed companies than there were years ago, and most are less profitable than they were. This alerts us to the possibility that agency failures – the inability of shareholders to oversee managers – are damaging.

The second, bigger, fact is that as CEO pay has risen, productivity growth has fallen. This is consistent with the possibility that increased inequality between bosses and workers is inefficient (pdf). We have many theories as to why this might be. Inequality reduces trust, which is necessary for growth. High pay crowds out intrinsic motivations (pdf) and causes managers to focus (pdf) upon things that are monitorable (such as short-term earnings) rather than ones that are not, such as longer-term investments and innovations. Hierarchical systems lead to bad decisions because bosses lack complete information, and can demotivate junior staff. And high CEO pay encourages juniors to engage in office politics to seek promotion, and incentivize managers to strengthen their position by discouraging disruptive innovation and entrenching monopoly power.

But what of the argument that competition eventually eliminates bad bosses? True, it does sometimes; the relatively egalitarian John Lewis Partnership has done better than department stores such as House of Fraser or Debenhams, for example. But market forces are weak (and perhaps getting weaker).

One reason for this is simply state intervention; without this, almost all shareholder-owned banks with their high-paid CEOs would have vanished.

Another reason is that there isn’t really a properly-functioning market for CEOs. I own tens of thousands of pounds of shares, but I’ve never been asked to vote on a CEO’s pay. Instead, the vote is exercised by fund managers many of whom are more bothered by the relative performance of their funds than absolute performance. We have widespread agency failures in which mates pay each other. And as Milton Friedman said:

If I spend somebody else’s money on somebody else, I’m not concerned about how much it is, and I’m not concerned about what I get.

Unsurprisingly, then, there is little clear link between CEO pay (pdf) and corporate performance. It might be that high pay is better explained by rent-seeking than as a reward for maximizing shareholder value - though as it's almost impossible to know in most cases what constitutes maximal value, we might never know for sure.  

In fact, my analogy with feudal lords is too generous to CEOs. Whereas (arguendo) a bad lord would pay for his incompetence perhaps with his life, bad CEOs walk away with fat pay cheques.

When so-called free marketeers try to defend bosses’ pay, they do the cause of free markets a huge dis-service by encouraging people to equate free markets with what is in effect a rigged system whereby bosses enrich themselves with no obvious benefit to the rest of us.

Comments

[Mark]

As with many failures of capitalism, the solution is to use market forces against those who profess to support them, but in fact seek to protect their rents.

For a start, give all shareholders binding votes on appointments (including dismissals) and remuneration. This includes the underlying owners of shares held collectively in pension schemes and investment and unit trusts.

Then require candidates for executive positions to pitch for them in an open, advertised contest. Each should set out his or her track record (good and ill), strategy and approach to the position and proposed remuneration. Post the presentations on a portal accessible by shareholders and follow up with a question and answer session.

My guess is that such an approach would lead to more diverse candidates being appointed and to remuneration being more closely aligned to performance. In particular, candidates who offer to buy meaningful amounts of stock upfront, thus exposing them to the risk of capital loss should their leadership underperform, would likely receive support, since their economic interests would truly be aligned with shareholders'. Those who lack the means to purchase such equity initially might offer to do so by salary sacrifice, or might borrow the necessary funds against future income.

[MJW]

The defence of high CEO pay based on share price is facile, huge movements based on CEO ins/outs is a sign of markets irrationally overweighting loud noises in the absence of meaningful information.

It also extends to some pretty stupid conclusions; it suggests all you need to add millions of pounds in shareholder value (or to reverse millions of pounds in lost value) is replace your CEO regularly with another hyped up superstar, rather than investing in more fundamental capabilities. The magical thinking that a superstar CEO can resolve long-term structural issues with a sweep of a wand is rather convenient justification for mates rates rent extraction.

It doesn't work forever, because sooner or later it's obvious the Emperors New Clothes aren't all they're hyped up to be. But by the time that happens the money is gone.

[Dave Timoney]

High CEO pay is a symptom of the end of feudal relations, in the sense that it is a feature of modern, financialised firms rather than more traditional family-owned companies (or paternalistic trusts like John Lewis). Overpaid CEOs are rarely organic products of the business - either family heirs or churls who've worked their way up. They're closer to condottieri: mercenaries brought in from outside, often as part of a leveraged buyout or as prep for a flotation.

I appreciate this is stretching the analogy, but it highlights that high CEO pay is a product of surplus corporate cash, just as the prevalence of mercenaries in Northern Italy in the late Middle Ages reflected that region's liquid wealth (the emergence of banking and merchant capital markets in Rennaisance Italy is not a coincidence). It also suggests that the process of financialistion has inflated remuneration by creating an artifical market for CEOs in which pay norms are divorced from the productive base (so greater shareholder power would not necessarily arrest this).

The simplest way to restrain high CEO pay may be to increase the cost of capital.

[Aaron Headly]

Fat Cats/Bad Company: The Strange Cult of the CEO by Gideon Haigh, a guy I normally read for cricket information, also covered a lot of this in 2004. And pretty well, I thought.

[Blissex]

«The defence of high CEO pay based on share price is facile,»

Isn't it curious that CEO etc. bonuses are paid on the absolute change in the company share price and not on the difference between the change in the company share price and the change of the sector share price index?
It may be surprising that a lot of awesomely well paid compensation consultants rarely if ever recommend that... :-)

[Justus]

> For a start, give all shareholders binding votes on appointments (including dismissals) and remuneration. This includes the underlying owners of shares held collectively in pension schemes and investment and unit trusts.

This is pretty ridiculous in practice. I have shares in probably 10,000 companies (between the US, Developed Europe, and Emerging Markets). We already know that voters generally don't bother to learn about the half-dozen ballot measures they are faced with every 2-4 years. Now you expect them to make thousands of conscientious votes every year?

All of those votes would be along party lines (CEO X said something bad/good about a politician I like/hate) or signalling or other irrational reasons. And that's if they don't just always take the incumbent.

Anyone who owns a mutual fund gets to vote on proxy statements on a regular basis and it doesn't seem to have done anything to fix the many, many problems with boards of directors. Usually the incumbent wins with 99% of the vote. Those are numbers even dictators in 3rd world countries would like to have in rigged elections.

[Mark]

@ Justus, were my proposal to gain the force of law, you would of course have every right not to participate in votes in which you feel you have insufficient information. Were you to do so, the outcomes would be decided for you by those who are sufficiently informed and motivated to vote.

Currently, neither they nor you have any ability to influence the appointment or remuneration of the directors of the firms they own. I know which option I consider less imperfect.

[Luis Enrique]

my ageing brain is so addled, I thought I'd made that metaphor up myself.

yes and if each dukedom was listed on the stockmarket, then if a new more capable Lord takes the helm you'd probably see the stock price rise and then (some) economists take that as evidence that their remuneration reflects marginal product.

[DavidM]

However interesting this is, it's not too important. The pay of the CEO is insignificant in terms of corporate expenditure and inequality.

I would consider this important if there is a correlation between the CEO pay and the pay of the other directors and employees down to middle management. I mean, if companies with unequal CEOs also have unequal middle managers who make a lot more than their subordinates.

I suspect there is a strong correlation, mainly for ideological reasons (leadership cult), so this is probably important. Any data?

[Staberinde]

Perhaps there's also something else going on. If organic growth is ever-more difficult to come by, productivity growth bumps along the bottom and market disruption is a constant threat - perhaps companies imagine that having a highly-paid CEO is a form of Pascal's Wager?

Sure, the overpaid CEO probably won't make enough of an impact to justify an outsized package. But who wants to look over the rubble of their failed company, shareholder value destroyed, sold for a song, disrupted like a Kodak and think to themselves "If only we'd paid a few million more, we might have got ourselves a visionary CEO who could have avoided this mess..."?

I wonder if there isn't some perverse logic similar to this in remuneration committees' heads. "At leasrt if it all goes to shit, Bob, nobody can say we didn't pay top dollar for leadership."

[Tom Powdril]

@Mark

"For a start, give all shareholders binding votes on appointments (including dismissals) and remuneration."

They already have binding votes on directors at UK companies (annual in most cases), a triennial binding vote on rem policy and binding votes on new incentive plans. The large majority of shareholders vote for the large majority of these resolutions by a large majority.

I've been working in this field for 15+ years now and as a result have no faith in the capacity of shareholders to oversee let alone challenge corporate excess. Even pushing votes back down the system is unlikely to achieve much. Pension funds have been de-democratised in many countries, removing ordinary member reps and putting professionals (who think like asset managers) in place instead.

Relying on shareholders to do anything serious has been tried and failed. They are too weak and uninterested in most cases. Something else is required.

[Lindsay Berge]

I find the quote "If I spend somebody else’s money on somebody else, I’m not concerned about how much it is, and I’m not concerned about what I get." more revealing of Milton Friedman's ethics and morality than that of people in general. My experience is that people who make spending decisions on behalf of other people are often deeply concerned about both price and value for money. If this were not true, how could there be professional investment advisers, stockbrokers or even buying departments in large businesses? Spending other people's money on someone else is their reason for existence. Even economists in reality seem very concerned that public expenditure be efficient and put a great deal of effort into arguing over details when this was Milton Friedman's specific example of 'spending someone else's money on someone else'. Once the taxes have been collected, it should be a matter of indifference to them what the money is spent on if they cannot persuade the government to spend it on themselves.
It all seems rhetoric and justification for selfishness rather than genuine insight.

[Blissex]

«I find the quote "If I spend somebody else’s money on somebody else, I’m not concerned about how much it is, and I’m not concerned about what I get." more revealing of Milton Friedman's ethics and morality»

You have been miltoned by that old fraudster!
M Friedman writes here apparently as if the assumption is that the majority of agents is unethical, and you fall for that trap.

The real problem and the fraud in his argument is that he is assuming *unaccountable* agents, not that he is assuming that agents are selfish. The problem with self-dealing is not when people are selfish, it is when they are unaccountable.

Taken at face value his argument is an argument against capitalism based on the joint stock company, because joint stock company executives “spend somebody else’s money on somebody else”, it is their entire job. Indeed IIRC A Smith criticized joint stock companies because of the potential for agent-principal conflict due to weak accountability, a problem that has become colossal.

The argument for the role of government as a group-purchasing arrangement is not that the group-purchase agents (secretaries of state and permanent undersecretaries) are selfless philosopher-kings, but that even if they are not there are checks and balances and voting to make them accountable, that is democracy and civil society ensure good outcomes even if public spending is managed by selfish people.

Of course "culture" and "social capital" both matter a great deal, and countries where spending other people's money is considered both a honor and a responsibility tend to work better than others because accountability cannot be 100%.

«Once the taxes have been collected, it should be a matter of indifference to them what the money is spent on if they cannot persuade the government to spend it on themselves.»

That seems crazy talk to me: democratic accountability means that they have a say in how it is spend and that it is spent that way, regardless of whether it is spent on themselves or not.

As an aside I mention here one of my usual points: absolute selfishness cannot be assumed in any sensible model of the political economy in which economic actors have limited lifetimes and reproduce, because people will maximize *also* the welfare of their children, lest their genes become extinct. And their genes have a "say" in that choice.

[Blissex]

«Relying on shareholders to do anything serious has been tried and failed. They are too weak and uninterested in most cases. Something else is required.»

It is has been done! Here I shall repeat Blissex's First Principle: the effective agent of economic progress is bankruptcy, that is the key.

The main reason is the effects: it eliminates the long tail of not very good business that manage to barely survive, as our blogger mentions now and then, thus raising efficiency. It was not state ownership or crony capitalism per se that makes business in corporatist ("soviet" or "fascist") inefficient, it is the effective abolition of bankruptcy. It is not markets or state planning that drives businesses to be efficient or inefficient: it is the threat of bankruptcy if the business cannot at least repay creditors.

That is the effectiveness of bankruptcy is because it is triggered by *creditors*, not shareholders (who always hope to find an even "greater fool" and often collude with management in self-dealing and misreporting of accounts), and puts management of the bankrupt in the hands of the creditors. Usually creditor interests are far more concentrated and far better funded than shareholders and the liquidation of the company to repay (part of) debt is directly in their immediate interest.

This observation perverted to an extreme has become "jensenism", from business guru M Jensen who has recommended running joint stock companies on high leverage (low joint stock and lots of debt) because creditors are much better at disciplining management than shareholders; the perversion is that this has led the financial sector (and not just...) to seek as ultimate creditor the central bank, as "greater fool of first resort", because the executives and politicians controlling it can be "captured" (nice euphemism!).

[Blissex]

«countries where spending other people's money is considered both a honor and a responsibility tend to work better than others because accountability cannot be 100%.»

Just to complete the argument and show that accountability and not selflessness is by far the greater issue, accountability is needed even if the agents are not at all inclined to self-dealing; even if they were selfless philosopher kings, because they might be incompetent or they might just misunderstand or reject the preferences of the people contributing to the group purchase scheme.

[Blissex]

«the effectiveness of bankruptcy is because it is triggered by *creditors*, not shareholders (who always hope to find an even "greater fool" and often collude with management in self-dealing and misreporting of accounts),»

The reason for that is that both shareholders and management are "sell-side" (even if with divergent interests in the details) and creditors are "buy-side", or that shareholders and management have a stake in the upside of the company, and creditors don't.
Thus the usual huge difference between the opinions of stock analysts (sell-side) and bond analysts (buy-side) on the same company.

[D]

Not sure if this is a for or against, perhaps just a facts of life.. but are we seeing 'cost disease' applied to CEOs?

The productivity if some CEOs increases - for example due to globalisation and companies getting bigger - and this raises the wages of all CEOs, even CEOs where productivity has fallen, because they need to compete with other sectors where productivity has risen.

If this is the case, managerialism may also have a role to play too if it's made CEOs more interchangeable between sectors.