That CEOs are overpaid is something, as Leonard Cohen would say, “everybody knows”; including the directors and shareholders who ultimately decide their pay. Yet firms are unwilling to do anything about it, because to do so would damage internal relations, undermine status and run against the norms of the system.
In Britain, the head of the Institute of Directors said the “current rate of executive pay is unsustainable”. Several global business leaders have criticised “excessive compensation”. Paul Anderson, then retiring chief executive officer (CEO) of BHP Billiton, saw “no way to justify the incredible compensation” of CEOs.
An Australian survey showed a majority of directors considered CEOs were overpaid – yet boards of directors set CEO pay. In-depth interviews with non-executive directors brought out comments like “I don’t think any individual is worth that much”.
The retiring global CEO of Royal Dutch Shell conceded:
“if I had been paid 50% more, I would not have done it better. If I had been paid 50% less, then I would not have done it worse”.
This disaffection is founded in reality. From the mid 1980s, real executive salaries grew substantially faster than average real wages.
Yet that difference in growth rates has not always existed. In America, Britain and Australia, series on executive pay and average earnings tracked each other fairly closely through the 1970s and early 1980s; but from the mid-1980s they diverged. CEO pay helps explain the [rise in top income earners’ share](http://andrewleigh.org/pdf/TopIncomesAustralia%20(ER.pdf) of national income, which in turn is only evident since the early 1980s.
Yet evidence shows the link between CEO pay and performance is often either negligible or negative - though there is some evidence of a weak, positive link, but this may be only in particular periods, such as boom times.
The issue that bothers the public most, though, is not CEOs being out of line with shareholders (interested in performance) but CEOs being out of line with society. Even if a link existed between CEO pay and performance, this should not in itself create the recent pay divergence. There has not been the substantial improvement in economic performance necessary to explain it.
What then drives executive pay? This recent study shows several influences.
Size and asymmetry
The first is size. Large firms have greater power. Executives of larger corporations command more resources and have greater opportunities for “value skimming”, with “proximity to large flows of revenue and fees”; so high executive pay reflects “the advantages of position”.
The second key element is the unbalanced (“asymmetric”) nature of “bargaining”. Rather than having opposing identities to executives, the members of boards or committees setting CEO pay are from the same social milieu with broadly similar interests, precluding “arms length” bargaining.
As in all markets, supply and demand are relevant. But in labour markets for executives and directors, social norms are a critical institution. These norms can be summarised by seven behavioural rules.
First, the status of a corporation depends in part on the status, and hence pay, of its CEO. As Warren Buffett argues, “No company wants to be in the bottom quartile as far as CEO pay goes”. Firms like to have “celebrity” CEOs, and “celebrity” status may boost CEO pay, even though shareholder returns typically decline after a CEO reaches “celebrity” status.
The ability of CEOs to extract rents - “the extra returns that firms or individuals obtain due to their positional advantages”, is influenced by their social capital or networks. So CEO compensation includes a “social circle premium”, boosted by “golfing in the same exclusive club, sharing directors who understand the local pay norm and displaying luxury mansions”.
Being a CEO in a company well connected to other company boards increases CEO pay.
Women typically are less integrated into ruling class old boys networks, and this helps explain the substantial gender pay gap amongst executive directors and CEOs.
CEO pay is heavily influenced by comparisons. This is known as relative pay deprivation (or, as Buffett says, “envy is bigger than greed”). Executives typically believe they are above average and hence deserve to be paid above average. Many firms try to pay above average, almost no-one wants to pay below average.
Observers talk of the domino effect, the Wobegon effect (after a mythical place where every child is above average) or even the “Wal King effect”. When business lobbyists claimed that pay disclosure rules caused a surge in pay growth, they conceded the relevance of relative pay deprivation. There is statistical evidence of it in Australia.
Institutions emerge to help this asymmetric pattern bargaining. Leapfrogging happens with or without disclosure rules. Private institutions take advantage of relative pay deprivation concerns. Pay surveys, a recent development, create reference points for CEO pay. Remuneration consultants have gained prominence in the last three decades. They “are only seen to be doing their jobs if remuneration rises”. Remuneration committees on boards provide the appearance of independent review, setting “undemanding targets” while “ratcheting” pay.
The incentive structure of executive pay adjusts over time. It does this to minimise the risk that pay falls to match performance, to justify high growth and to deflect shareholder concerns.
There has been a major change in composition of CEO pay towards greater use of incentive pay. If this was just about aligning pay and performance, base pay would reduce as incentive pay increased, and total CEO pay on average would change little. Instead the growth of incentives has been central to the growth of total CEO pay.
The inflation of CEO pay, and its high level in the USA, has principally occurred through expansion of incentives. Incentives give the appearance of tying pay to performance and can mask its level as “only relatively few individuals with technical insight are able to understand what an executive is being paid”.
However, CEOs resist pay reductions. If incentive formulas would cut remuneration, then a restructuring of CEO packages (eg to increase the base or revise incentives) might prevent this. So a study of restructures of CEO pay packages found many amendments leading to higher pay, but no amendments if packages paid above expectations.
Different norms shape pay in different segments, though references will be made to other segments to justify increases. The general factors inflating US executive pay tend to be at work in Australia, but that does not mean Australian and US CEO pay are determined in a single global labour market. While beneficiaries have long used international comparisons as a justification for high growth, recruitment of Australian CEOs from overseas was not common.
Manufacturing workers’ pay varies less between industrialised countries than does CEO compensation. The 20 highest paid US CEOs received three times the average remuneration of their European equivalents, but on average ran smaller businesses by turnover. Cross-national differences in CEO pay are “an expression of deeper social values”, including differences in tolerance of inequality (including amongst economic elites) and acceptance of “market” outcomes (though everywhere, inequality is worse than people want or perceive).
The ability of CEOs to ratchet pay upwards is contested. CEO power to shape pay is constrained by several factors. These include: shareholder activists (including superannuation funds); actual or threatened legislation to impose limits on CEO excess; and popular resistance. These wax and wane.
After world war two, an “outrage constraint” would “limit executive paychecks”. But from the 1980s, financialisation and the move towards market liberalism shifted, increased CEOs’ capacity to extract rents and the size and visibility of finance executive rewards, setting new benchmarks for others to follow. The growing “gap between the rich and the super-rich” gave CEOs motivation to aim even higher.
A few years ago corporate Australia was engaged in “a race against time to persuade politicians not to intervene”. The subsequent “two strikes” rule forced them to “adopt a more cautious and thoughtful approach” that was seen as “a genuine tightening”.
So it’s not quite the case that “nothing hapens”. But it only “happens” for a while, until the fuss dies down.
A more extensive version of this, with full references, is published in the Journal of Industrial Relations.