How Not to Combat Corporate Corruption
Coping with the Principal-Agent Problem
OCTOBER 01, 2002 by ANTHONY DE JASAY
The recent and unusually rich spate of corporate fraud by creative accounting has set off two streams of verdicts. One condemns capitalism as a rule system that has no moral and cultural foundations of its own. When the foundations it has inherited from earlier rule systems get eroded by wear and tear, capitalism becomes a source of iniquity and corruption. It is intrinsically evil, cannot be reformed, and must be done away with.
This view is an evergreen. Besides its other weaknesses, it ignores a clear lesson of history. Every rule system ever tried by mankind that was not based, as capitalism is, on finders keepers, the protection of ownership, and the freedom of contract has been found to be at least as corrupt and iniquitous as capitalism, and also much less able to relieve poverty and deliver the material prosperity all save saints crave. Anti-capitalists seem never aware of tangling themselves up in this self-contradiction.
The other, reformist stream accepts the ground rules of the system, albeit with reservations about the freedom of contract. However, it believes the system generates abuse by virtue of its increasing complexity. To stamp out abuse, the ground rules must be supplemented by regulations that, too, must become increasingly complex and demand vastly larger resources for their operation and enforcement.
However, going down the regulatory path may not turn out to be an advance. The more elaborate the regulations, the more elaborate become the abuses that evade them. The cost of enforcement keeps going up, and the operation of the regulated economy grows progressively more cumbersome, lawyer-ridden, and sluggish. At some point, the game is no longer worth the candle. Where this point lies is a matter of subjective judgment rather than cost-benefit measurement. Maybe we are still short of it, maybe already past it. The answer is not as important as is the constant awareness that such a point exists.
Instead of flailing blindly at the elusive white-collar crook, and writing reams of new laws and regulations, we should first seek a clearer understanding of what we are dealing with. Corporate fraud is a contingent but probable byproduct of the fundamental principal-agent relation that results when ownership and management are separated. (More generally, principal-agent problems arise when a function is carried out by an agent whose participation in the costs and benefits generated by the function differs from those of the principal. The most important principal-agent problem seems to be the one between society and its government.) The principal-agent problem separating the shareholders from the managers of a corporation is the price the former must pay for the specialized skill, attention, and effort the latter bring to the enterprise. It cannot be avoided altogether without abolishing the agents and requiring the principals to step in their place, that is, without reverting to owner management. Doing so would bar multiple ownership and throw away the advantages of the division of labor. Short of that, the problem can be mitigated by intelligent use of the freedom of contract.
The early forms of getting round the problem were to motivate the salaried manager by a pay raise or some ad-hoc reward for success, a bonus. More systematically, he would be entitled to profit-sharing according to some formula. This had obvious drawbacks, encouraging short-termism, a bias to seek profits even by risking unduly high losses, as well as a weakening of the enterprise through the cash outflow going to the manager. An improved solution seemed to be offered by the stock option, two incendiary words that act as a red rag to set off much current fury, some of it misplaced, some not. The stock option does not encompass the whole corporate principal-agent problem, but is certainly at the heart of it.
Like profit-sharing, the stock option rewards profits and fails adequately to punish losses and recklessness, particularly while the option is “out of the money.” However, unlike profit-sharing, it creates no bias against the long term, and it does not conflict with the interests of creditors since it causes no cash outflow. At least in principle, it comes closest to the ideal contract that minimizes the conflict of interest between shareholders and managers. However, the conflict can never be wholly resolved.
Product of a Bargain
To begin with, the terms of the stock option are the product of a bargain between the shareholders and the management. The shareholder interest is to fix terms so they are good enough to secure and motivate the managerial talent that will realize the full profit potential of the enterprise over its life span, but no better than would be needed to attract alternative management of comparable talent. The manager interest, of course, is to negotiate the highest possible terms the shareholders will concede, and never mind the terms an alternative management team of equal worth would be prepared to accept. However, with ownership widely dispersed, and non-executive directors in effect chosen by the management, the bargain over the terms of stock options is grossly asymmetric. Shareholders are, as it were, half-asleep, and it is often as easy for management to get outrageously favorable terms approved as it is to steal coins from a blind beggar’s bowl.
Next, there is an asymmetry of means between principals and agents even where their interests are identical. Both shareholders and managers tend to prefer a high stock valuation. There is not much shareholders can do about the price of their stock except hope for the best. Management has a range of means for “massaging” the valuation, some acceptable, others not. Smoothing earnings, and gaining the confidence of security analysts by never lying to them, are in order; creative accounting, hype, and downright lies are not. Most managements are intelligent enough to value their own reputation higher than the risky chance of getting away with fraudulent practices, but the evidence tells us that some will succumb to obvious temptations. The incentives of auditors are likewise mixed; honesty is probably their best policy, but sometimes they will risk a reputation rather than antagonize management.
All in all, stock options are capable of generating a panoply of more or less subtle incentives, with the morally dubious and corrupting among them being the most conspicuous. With public hostility aroused by corporate abuses, there is a danger that regulatory fervor will smother the less conspicuous but surely far more important good that options do for economic efficiency.
In this regard, a strange proposal has been floated. It is strongly suggested that accounting standards should be modified so as to require corporate accountants to treat stock options as a cost. The proposal is a product of muddled thought. If a firm makes a pretax profit of $1,000 and then pays $100 in profit-sharing to its managers, its pretax profit is reduced to $900; the managers’ share is a cost that, other things being equal, reduces the net worth of the enterprise. If instead of cash the managers get options on, say, 1 percent of the company’s stock at an exercise price that is “in the money,” pretax profits stay at $1,000, of which $990 belongs to the old owners and $10 to the new manager-shareholders both in the year in question and forever after. The stock options do not reduce profit or corporate wealth, but they dilute earnings per share and equity per share. The dilution is a side transaction that transfers value from old to new shareholders, rather than a cost to the corporation.
It is not clear how the capital transaction diluting the interest of old shareholders in favor of the new ones who exercise their options is meant to be translated into a cost that is supposed to reduce the corporation’s total earnings. What seems to be clear, though, is that the proposed accounting treatment of stock options, whether at the point of grant or of exercise, as a cost is intended as a deterrent against their free-and-easy use, which sometimes comes close to brazen managerial self-dealing. But distorting corporate accounts by a logically faulty regulation, however well-meaning, is not the proper way to protect shareholders from abusive practice. Even if it did protect them, it would do so by sacrificing a good deal of the efficiency that the stock-option method of coping with the principal-agent problem has given us.