Bliss is it in this dawn to be alive, but to be a banker is very heaven--with apologies to Wordsworth. The Federal Reserve Board's monetary policy gurus are making cash available to banks at almost no cost, it can be re-lent to desperate borrowers at mouthwatering margins, and if anything goes really wrong, the government stands ready to bail you out. Free cash, or almost; high and rising charges to borrowers and consumers; bailouts if assets become toxic--what more can a bank president and his board want in this best of all possible worlds?

Freedom to set compensation, that's what. But that is not to be: The government has decided that it shall be the final arbiter of just how much banks that depend on it for their existence can pay their 5 senior executive officers and 20 "most highly compensated employees" in bonuses, and the compensation of the "100 most highly paid employees that are not subject to the bonus restrictions." Exempted are employees whose total annual compensation is not more than $500,000, and to whom any additional compensation is paid in the form of stock that can be sold only in the "long term." In its broadest terms, compensation packages must be "performance-based." Indeed, even those firms that have paid back bailout loans might well find themselves caught in this new regulatory net, since Kenneth Feinberg, the designated "pay czar," has taken the position that he has the discretion to claw back any compensation received by employees not only in firms that still owe the government money, "but in any company that received federal assistance," even those that have paid all their loans in full. Once a debtor, always in debt.

Feinberg contends that his discretion is complete and not subject to review--by anyone. "The [bank] officials can't run to the secretary of the Treasury. The officials can't run to the courthouse or a local court. My decision is final."

The pay czar does not have the field of compensation-setting all to himself. The ever-active Barney Frank has persuaded his House colleagues to approve legislation that is designed to control executive compensation--sort of. But not directly. Instead, it seeks to introduce pay restraint by sprucing up corporate governance. Shareholders are to get the power to have a "say on pay" by casting advisory votes on proposed pay packages, and compensation committees are to be given somewhat greater power to do their job properly. Unless the very thought of Barney Frank trying to curb management's ability to set its own pay without regard to shareholders' interests offends you, nothing much here to worry about. After all, it is over 60 years since Adolf A. Berle and Gardiner Means, in their classic The Modern Corporation and Private Property, wrestled with the problem created by the inability of scattered shareholders to make certain that the managers of their businesses pursued the owners' interests, rather than those of the management group.

Since the current financial upheaval spread to the real economy, we probably have thought harder and learned more about the problem of compensation in the banking sector than at any time since Berle and Means put pen to paper. The first and easiest lesson is that when the government is a major shareholder, or the source of the financing that provides life-support for any company, it will want to have a say in how much of the dependency's revenues go into executive pockets. It's hard to quarrel with the modern version of the old British notion that he who takes the king's shilling is obliged to render service to the king. No self-respecting private equity guy would allow the executives of a company he has just rescued with a major investment to set their compensation without giving him something to say about it. Neither will, or even should, the government. So pay czar Feinberg is busily reviewing the detailed executive compensation plans plunked on his desk by the not-so-magnificent-seven companies still beholden to the government: AIG, Bank of America, Citigroup, Chrysler, Chrysler Financial, General Motors, and GMAC. Feinberg has 60 days from receipt of the plans (submitted to him on August 14) to decide whether to approve or demand modifications of plans to compensate the employees of the seven firms. The entire financial services industry, even those that never heard of TARP or accepted any government aid, knows that what Feinberg decides will affect it. J. Mark Poerio, co-chair of the executive compensation group at consultants Paul Hastings, advisers to hundreds of financial firms, tells the Washington Post, "What Feinberg eventually approves will essentially serve as an outer boundary for what the law allows."

The rationale for government intervention on the widest possible scale is simple and, to some, persuasive. Compensation systems that encourage excessive risk-taking by firms that are too big to fail are not in the public interest--they contain the seeds of systemic collapse, as almost occurred at the start of the present crisis. So the government must have a say in the structure of bonuses of all banks to prevent excessive risk-taking.

Note that this argument applies to the risk-inducing nature of bonuses, hence the government's relaxed attitude towards bonuses in the form of restricted shares, which cannot be sold for several years, giving employees a stake in the long-term health and performance of their firms. Note, too, that the level of compensation is not supposed to affect whether the government approves any particular scheme, a distinction that might get lost in the hurly-burly of congressional hearings and White House press conferences.

Nor is the government certain to be deterred, in the end, by the presence of contracts setting compensation. After all, contracts that contravene public policy--think of agreements to fix prices, or a mob contract to "hit" a member no longer considered trustworthy--have long been held to be unenforceable. It is not a great stretch for the government to argue that contracts embodying compensation plans that encourage behavior so risky as to create systemic risk are null and void. After all, the government had no compunction about negating the contractual right of Chrysler's creditors to preferential treatment when it went bankrupt, in effect transferring such preference to vote-delivering trade unions. All in the public interest, of course.

The reasonableness of the government's claim for a seat at the compensation committee table doesn't mean that the government will do a good job of setting executive compensation. Populist and egalitarian pressures emanating from the Hill and the Oval Office will undoubtedly mount when Congress returns, and considers--if that is the right word--its reaction to the multimillion-dollar payouts planned not only by highly profitable firms such as Goldman Sachs, but by money-losers such as Citigroup, a bank that has received $45 billion in government aid in exchange for a 34 percent share in the company. Feinberg already has approved an $8.5 million pay packet, largely in the form of stock grants, to AIG's new president. He also is likely to approve plans such as those recently adopted by Wells Fargo--increase base pay and cut bonuses that might provide incentives to reckless lending and trading. And he must consider the impact of any decision on the ability of the firm involved to retain its top guns--although if his writ runs as widely as observers believe it will, unhappy bankers might have nowhere else to go--except perhaps to the new boutique advisory firms that are already wooing away bankers eager to find a niche below the government's radar.

Because Feinberg has refused to acknowledge that the provision protecting contracts signed before February 11 of this year puts them beyond his reach--the rules provide for a claw back of "any bonus based on materially inaccurate performance criteria"--he now has to deal with Citigroup's One Hundred Million Dollar Man.

Actually, Andrew Hall, the trader with Citigroup subsidiary Phibro, is entitled under his pre-February 11 contract only to $98 million, but a mere $2 million shouldn't be allowed to come between a fact and a good headline. Hall earned many times $100 million for Phibro and hence its parent, Citigroup, so his payout--Citigroup refers to it as a share of profits--is clearly performance-related, one of the stated goals of pay structure reformers. Absent Feinberg's ability to find some way around that contract, or a willingness by Hall to adjust his bonus (there is a rumor that he might be prepared to spread receipt over several years), Hall should whistle a happy tune en route to the bank while Treasury Secretary Tim Geithner hopes the congressional hearing at which he will be called upon to explain this provision of the law--something about constitutional protection against ex post facto legislation--does not last too long.

The sums involved aside, it is not unusual for loss-making banks such as Citi to feel it necessary to pay substantial bonuses to employees working in profitable divisions such as Phibro. None of this all-for-one-and-one-for-all stuff for traders, who seem to have learned their lesson from the town beggar in that great musical Fiddler on the Roof. When offered one kopek by the town butcher, the beggar demanded his usual two kopeks. On being advised by the butcher that he had a bad week, the beggar responded, "So you had a bad week--why should I suffer?"

We have learned, too, that most Americans are unappalled by big pay packets, so long as they are at least arguably related to performance. No one is calling for the lynching of Warren Buffett, who made a bundle on his save-the-day investment in Goldman Sachs. Or Bill Gates, whose billions result from his having invented a company that changed the way business is done and daily life is lived. Even Barack Obama, who capped pay and perks for executives at "firms receiving extraordinary help from U.S., taxpayers," and who thinks that upper-income Americans had it so good in the Bush years that they are obliged to be enthusiastic supporters of his plan to raise their taxes, has said, "This is America. We don't disparage wealth. . . . We believe that success should be rewarded." Either he means it, or feels it necessary for political reasons to say it. No matter. The point is that if this president sensed a wave of revulsion at high incomes, he would surely ride that wave. He doesn't, so he hasn't.

The most important and troubling lesson we have learned is that it is not how much executives in the financial sector are paid, but how that pay is structured. "Incentives matter" has long been a mantra of conservatives eager to allow the invisible hand to work its magic, rather than rely on government to direct economic activities. It was that belief in the ability of proper incentives to produce socially desirable behavior that underpinned conservative plans for welfare reform. With incentives and the public interest properly aligned, markets, not men, should decide on the allocation of a nation's resources, and on the division of the rewards for economic effort. Unless . . .

As Stanford professor Roger Noll put it in a communication he has generously allowed me to quote:

The financial whizzes did nothing illegal and were responding to the incentives they faced. The system of large cash bonuses for gains coupled with no penalty for losses leads them to play games in which the short-term probability of gain is high but the long-term probability of loss also is high. This is the basic underlying fact behind every financial crisis in the last 25 years. If we persist in a system in which a company makes X a year every year for ten years but then loses 25X in the eleventh, and we give Y in bonuses in the good years and zero in the bad, the whizzes will still prefer boom and bust.

When the pursuit of such incentives harms innocent bystanders, it is difficult to argue that there is no role for government to play in correcting what economists call market failure resulting from externalities, even at the risk of introducing government failure. We don't allow 8-year-old children to spend their days digging coal only because we are humane, but also because such an assault on the health and educational opportunities of these children imposes costs on society that are not borne by mine owners. We don't allow manufacturers to pollute if that damages the health of innocent bystanders, imposing costs on society. And we now know that the structure of financial incentives can lead to risk-taking that has serious consequences for society--for Main Street as well as Wall Street. If compensation is structured so that the rewards of risk-taking go to bankers and their shareholders, but the costs of failure are borne by a wider group, the bankers will take more risks than are economically efficient. And that is without giving weight to Adam Smith's shrewd observation that men tend to be excessive risk-takers even without a skewed reward system: "The overweening conceit which the greater part of men have of their own abilities, is an ancient evil. . . . The chance of gain is by every man more or less over-valued, and the chance of loss is by most men under-valued."

The argument that incentives and inclinations exist that lead to excessive risk-taking is not a moral argument, or a political one, or an argument in favor of a more equal distribution of income and wealth. It is solely an economic argument: Compensation structured as it has been in the financial sector results in an uneconomically excessive amount of risk-taking, just as a failure to make a polluter internalize the costs of pollution provides an incentive for him to produce more than if he had to pay all the costs he imposes on society.

Mortgage brokers have an incentive to write NINJA mortgages--no income, no job or assets--because they are paid up front, and then pass the risk on to banks. Banks immediately wrap these risky mortgages into securities, and sell them to investors who are reassured by the triple A ratings conferred by the rating agencies--who earn a fee only if the deal gets done. Everyone has an incentive to do the wrong thing--no surprise that they do just that. Not because they are law-breakers, not because they are antisocial, not because they can't wait to see the undeserving evicted from their homes. Simply because they are following the incentives embedded in the compensation packages that do not reflect the costs to society of their errors.

Until now, economists held that the fear of "reputational consequences" would deter such behavior. But most of these transactions that originate with a broker paid up-front are one-off--the same customer is unlikely to return, or learn soon enough the consequences of his brokers' behavior to warn others. Executives who bring down their institutions leave with golden goodbyes and access to talk shows on which they unashamedly--shame being in short supply these days--justify their actions en route to a game of golf at a country club, dues paid by the company from which they departed but at which an office and staff support are still available to make their transition to a new life friction-free.

I exaggerate: Not all cases fit that description. But almost all have one characteristic in common: The cost of the pursuit of the incentives contained in a compensation package, when that pursuit leads to major loss, has not been borne by the pursuer, but by thousands of people he has never met.

Since society bears the cost, society, as represented by its elected officials and their appointees, must have something to say about how to eliminate or at least mitigate incentives that are causing such woe to the innocent. Translating that general conclusion into effective regulation is no easy chore. The House hopes to accomplish the goal by improving corporate governance, making executive compensation subject to more effective review by shareholders and boards of directors. No harm there, but neither the shareholders nor the compensation committees of boards have any incentive to internalize the social costs of excessive risk-taking. The solution is made no easier by political posturing by representatives of both parties. New York attorney general Andrew Cuomo continued his undeclared drive for the Democratic gubernatorial nomination by revealing his horror at discovering that nine banks that received government aid paid out nearly $33 billion in bonuses last year. Which is not particularly relevant if the bonuses were paid pursuant to contracts or to employees whose divisions turned a handsome profit.

Equally unhelpful are the Republican opponents of any interference in the structure of pay packages. Reacting to the House bill, Representative Spencer Bachus of Alabama said that the government should not interfere in the operation of private-sector companies, adding, "Government bureaucrats don't know what's best for America," which is undoubtedly true (cf. health care reform), but not particularly relevant in a circumstance in which private-sector companies exist only because the government has seen fit to pour billions into them, and private-sector bureaucrats have an incentive to ignore the public interest.

So, like it or not, we are faced with a situation in which government regulators will have to try to figure out how to align private with public interests. Here the watchword must be, "Get the incentives right, and if huge bonuses flow to the skilled or the merely lucky, so be it." Step one is to make sure that every player has his own skin in the game. Lenders at all levels must be made to bear some of the cost of loans they make that go sour. Recipients of bonuses must have their entitlements based on longer rather than shorter time periods. And be subject to recapture under certain conditions. Or perhaps the system adopted by Credit Suisse might be used: Five billion dollars in bonuses were paid in January from a fund consisting of the bank's toxic assets--a plan called "eat your own cooking" by one banker. So far, the value of bonuses is up 17 percent since the plan was instituted, far less than the 75 percent increase in Credit Suisse shares, in which bonuses were paid in the past.

Rating agencies must also be required to have skin in the game, perhaps by being forced to take part of their compensation in the securities that they are blessing, and better still, by being subjected to the competition from which government rules and practice now shield them by making entry into the rating business unnecessarily difficult. And it wouldn't be a bad idea to have a portion of the pay of regulators held in escrow, to be returned to the taxpayers if their regulatory schemes prove inadequate.

Will even a perfect scheme of forcing financial institutions to internalize the costs their pay schemes now impose on society eliminate future financial crises? Of course not. The "animal spirits" that John Maynard Keynes said animate capitalists will always at times result in excessively exuberant behavior. Absent a taste for structural reform on the part of the administration, the banks that are too big to fail will remain too big to fail, creating moral hazard that even the most wisely crafted compensation scheme cannot offset. But some progress is better than none, some reduction of systemic risk better than none, some correction of incentives that leave society holding the bag while risk-taking executives escape unscathed better than none. And, most important, ardent capitalists should agree that some reform, however imperfect, that restores faith in their preferred economic system is better than none.

Irwin M. Stelzer, a contributing editor to THE WEEKLY STANDARD, is director of economic policy studies at the Hudson Institute and a columnist for the Sunday Times (London).

Next Page