The Market Rate

If excessive pay were chiefly the result of poor governance, such improvements might have been expected to bring about a large and permanent downward adjustment in executives' remuneration. After the stockmarket bubble burst, median pay did indeed fall a little, and the arithmetic mean much more. Some, if not most, of the fall can be put down to the decline in the stockmarket. After all, executives are paid partly in shares, a currency that had just suffered a big devaluation. But the effect was relatively small and temporary. By 2004 pay in the largest companies was again growing much faster than average workers' wages. By 2005, according to Mercer, the total pay of the chief executives of 350 large companies had easily surpassed its level in 2002 (see chart 1 in "In the Money"). Better governance may have lowered pay, but not by much.

In fact, the argument that powerful incumbent chief executives use their influence to wring extra money out of weak boards is hard to square with what has happened in the market for outside executives. In the 1990s, when incumbent executives were extracting more money than ever before from their particular firm, a record number of managers voluntarily gave up this advantage by leaving for new companies, where they would typically start off with less influence. Moreover, executives were able to keep their salary high when they changed firms. The fat pay packet which according to the theory depended on managers' power over their old board turned out to be transferable to other jobs where the managers had to negotiate with strangers on the new board. That looks like a market at work, not a rigged game.

The third main complaint about governance in the 1990s was that executives were granted too many options on easy terms. Yet, on closer inspection, this accusation also needs qualifying. It is clear that a sort of mania took hold in the 1990s and that boards were extravagant with options. This makes options a failure of governance, but not one that fits the common view that top managers were exploiting investors. Brian Hall of Harvard University and Kevin Murphy of the University of Southern California have established that top executives received under 10% of the options handed out at that time. The rest went to other employees — in some companies to all employees. Such generosity was not restricted to Silicon Valley. In 2000, for instance, "old-economy" companies gave employees below the top five an average of $2,559 each in options.

Dishing out options to the office cleaners makes no sense. Options are an expensive way to pay employees: they are worth less to workers than they cost the company to issue. This is because options cannot be freely traded, and workers may feel that if they already depend on their firm for their job, investing in the business as well means having too many eggs in one basket. For such people, cash is much more valuable. For companies, one reason for issuing them was to inspire loyalty and provide an incentive to stay on. But option grants outside management have fallen steeply since 2004 and will, consultants predict, eventually disappear altogether.

Governance critics also saw something sinister in the easy terms of options packages. In the 1990s these were almost always "plain vanilla", granted at the current share price, with a vesting period of three years and without any performance hurdles. Executives got them just for turning up to work ("pay for pulse") as they waited for the bull market to lift the share price.

That design was hardly demanding, but it was encouraged by America's accounting rules rather than by greedy executives. Contrary to both economic logic and the principles of accounting, options granted at or above the day's share price did not count as an expense in the accounts and eventually delivered a tax credit to the company. On the other hand, options "in the money", ie, below that day's share price, or with performance hurdles attached, were charged as a cost. Advocates of governance reform argued, reasonably enough, that "expensing" options should not affect a company's stockmarket value, as it would not affect its cashflow. The failure to attach hurdles to options was a failure of governance.

However, such advocates underestimated people's enslavement to accounting conventions. In the past two years there has been a flowering of long-term incentive plans of a different kind. According to Scott Olsen, a pay expert at PricewaterhouseCoopers, performance hurdles are now common. There are longer vesting periods and holding periods after the options have been exercised. Worried about dilution, investors have successfully pressed companies to issue "restricted" shares which their owners cannot sell freely. The shift from standard options was prompted by an accounting change that came into force in America in 2005, requiring options of all sorts to be expensed. The performance hurdles are mostly based on internal measurements, because market-based tests still receive less favourable accounting treatment.

Everybody's Doing It
International pay patterns add to the evidence that American pay was determined chiefly by the market. National pay cultures vary, with long-term incentives playing a far more important part in America than anywhere else (see chart 5). Options for executives were first adopted in continental Europe about ten years ago and grew rapidly almost everywhere. According to Harm van Esch of Russell Reynolds, an executive-search firm, industries that are particularly exposed to international competition, such as the technology sector, have changed faster than more local ones such as construction.

• « previous
• 1
• 2
• 3
• next »