Beyond Winning

by Robert H Mnookin

  Uncertainty about the housing market will further complicate Sam and Betty’s task. Neither of them knows what will happen if Sam turns down an offer of $250,000 and Betty puts in additional effort in the hope of receiving $275,000 or $300,000. Most likely, however, Betty will have more information on this point than Sam. Can he trust her to reveal this information candidly, when it might be in her interest for him to accept the lower offer?

  Consider the homeowner’s dilemma at an even earlier stage of the transaction, before the house goes on the market. After thinking about these problems, Sam might realize that Betty has an incentive to set a low selling price for his home so that it could be sold quickly and with little effort. Reaching for the stars isn’t in Betty’s interest. It might not be in Sam’s interest either, but he wants to be sure that Betty is giving him information candidly. He might thus decide to ask a number of agents for competing estimate recommendations. Although this could provide him with some reassurance, competition of this sort is not a complete solution. Instead, such competition may encourage agents to make unrealistically high estimates in the hopes of securing an exclusive listing. After the listing is secured, an agent might put the house on the market for the high price but then expend little effort trying to market the house. After some period of time, the agent might then approach the owner and indicate the necessity of lowering the price to increase the chances of a sale. In the end, the homeowner may end up worse off for having initially set an unrealistically high price, particularly if a record of large unilateral price concessions is taken by prospective buyers to indicate that the house is of questionable value. Again, information disparities make it difficult for the principal to align the agent’s incentives with his own. The homeowner may be unable to monitor the agent’s efforts or the accuracy of a single agent’s estimates.

  Why doesn’t Sam just pay Betty by the hour? Many professionals—including lawyers and accountants—have traditionally been compensated in this way. At first glance, this may seem a straightforward way to guarantee that the agent expends the needed effort to get a good price. In reality, however, compensation by the hour creates an incentive for an agent to put in more time than may be necessary to get a good price. To earn a large commission on the sale of Sam’s house, Betty will necessarily have to invest a great deal of time. A quick sale with little effort will be less profitable for her than a sale that takes longer. Other things being equal, of course, Sam would prefer a sale sooner rather than later. Betty’s incentive to put in extra time doesn’t necessarily meet Sam’s needs.

  An hourly fee also creates monitoring problems. How does Sam know the number of hours Betty is actually putting in? And how does he know whether those hours are being spent efficiently, in a way that most benefits Sam? Is she diligently pursuing buyers, contacting other agents, and creating attractive brochures and ads to market the property? Or is she just holding open houses over and over again so that she can bill Sam for the set-up and break-down time? Sam might have reason to fear that Betty will not use her time most productively under an hourly fee arrangement.

  Sam could also offer to pay Betty a fixed fee for her work. Assume that Sam expects to list the house for $250,000. He and Betty know that if the house sells for this amount she’ll earn a commission of $15,000. But neither knows what the actual sale price will be. The market is hot. Maybe Sam will receive offers above his asking price—it’s been known to happen in his neighborhood. Or maybe no buyer will come along and he’ll have to drop the price to $230,000, or even lower. If Sam believes that the hot market will work to his advantage, he might offer to pay Betty $15,000, regardless of the sale price. He would thus insure against the possibility of a greater fee, at the risk that he would overcompensate Betty in the event the market failed him and the price had to be lowered.

  Fixed fees have certain advantages. They encourage the agent to get the job done within the cost parameters set by the fixed fee. However, fixed fees can create perverse incentives of their own. If Betty will receive $15,000 regardless of her effort or the sale price, why should she put in the time required to sell the house at $250,000, as long as she sells it at some price?

  What about a percentage fee with a clause to reduce the percentage if the house sells very quickly? Sam has already expressed concern that the house might sell in just a few days with minimal effort on Betty’s part. If that’s the case, why should Betty get her full 6 percent commission? Sam might propose that if the house sells within seven days of listing, Betty’s commission will be reduced to 4 percent. Even if Betty agrees to this fee structure, however, it creates a new set of incentive problems. Now Betty has an incentive to delay. Why sell the house on day five if on day eight she’ll make an additional 2 percent?

  What about some hybrid of a percentage fee and an hourly fee? After all, Sam’s real concern is that Betty will slack off if the house doesn’t sell quickly. It’s on day fifty that he needs Betty to work at selling the house, not on days one and two. Thus, Sam might suggest a lower percentage fee—perhaps 5 percent—plus an hourly bonus for work performed after day fourteen. In this way, he might hope to inspire Betty to put effort in when he needs it most. But from Betty’s perspective, this arrangement forces her to put effort into trying to sell a house that’s not priced right for the market. Why should she bear the burden in such a situation? Why shouldn’t Sam lower the price and thus generate more sales interest? And why should she work toward an early sale—which Sam, too, would prefer—if it just means that she’ll get a lower percentage fee?

  MONITORING SYSTEMS

  If incentive contracts don’t completely solve the problem, why can’t a principal just watch over his agent and ensure that the agent performs satisfactorily? This is the second management mechanism: monitoring. If Sam knows which marketing activities are most likely to result in the sale of his house, he can simply follow Betty around and see whether she engages in those activities. This mechanism is often used by employers, who monitor their employees and compensate them based, in part, on how well they perform.

  The problem with monitoring, however, is that it is expensive and it doesn’t always tell the principal what he needs to know. In order to determine whether an agent has performed appropriately, the principal must be able both to observe the agent’s behavior, which is often impossible, and to distinguish desirable from undesirable behavior, which is often beyond the principal’s expertise. Sam, for example, can’t watch Betty’s every move. To do so would waste the time he is saving by hiring her in the first place. In addition, even if he did watch her closely, he might not be able to distinguish between high-quality and low-quality work. If only three people attend his first open house, should he blame Betty? Were her marketing efforts substandard compared to what other agents would have done? Sam is unlikely to know.

  Perhaps Sam could employ another specialist or expert to monitor Betty. This approach is not uncommon. For example, a corporation’s in-house lawyers often monitor the efforts of the corporation’s outside lawyers, who work for private firms. Similarly, outside corporate directors often monitor the efforts of management. It should be obvious, however, that this is hardly an ideal solution. Hiring yet another professional to provide services is expensive—and the compensation arrangement for this other professional may in itself create distorting incentives. Moreover, a conspiracy of sorts may develop between the agents. In the corporate world, management is often responsible for selecting their monitors—the outside or “independent” directors. This inevitably raises concerns about informal collusion. In a general sense, such collusion results from the fact that similarly situated agents have more frequent contact with each other than principals and agents do. To the extent that agents expect to have repeat dealings with one another, this may well affect their behavior—sometimes in ways that may benefit the principal, but other times in ways that do not.

  BONDING

  Principal-agent differences can also be dampened by requiring the agent to post
a bond, usually in the form of money, at the start of the agency relationship, which he must forfeit if he acts in a way that conflicts with the principal’s interests. In the construction industry, a contractor may post a bond underwritten by an insurance company that can be used to complete the job for the owner if the contractor goes broke during the project. Pensions are sometimes considered such a bond: throughout their careers employees are induced to act in their employers’ best interests for fear of losing their pension’s large financial rewards. Similarly, compensation that is above market rates can be considered a form of principal-agent bond: if an employee is found acting contrary to the employer’s interests and is fired, he forfeits the market surplus that he has enjoyed up to that point.

  An agent’s concern for her reputation can also serve as a bond to protect her principal.4 Even if Betty has an economic incentive not to spend extra time working for a sale price above $250,000, and even if she knows that Sam cannot effectively monitor her shirking, Betty might still work diligently in order to keep her professional reputation intact. Real estate agents often acquire clients through word of mouth. Without recommendations from previous clients like Sam, Betty is unlikely to succeed in her business.

  While in some circumstances the principal may be able to affect the agent’s reputation, this is generally an imperfect solution to agency problems.5 It may be difficult to observe or verify that a particular outcome—success or failure—is attributable to the agent’s actions.

  In addition, principals can exploit agents as well as the other way around. For example, a homeowner might use an agent to acquire valuable information about the home’s expected value, and even to begin testing the house on the market, but then exclude some friend or acquaintance from the agency contract and subsequently sell the house directly to this third party. By doing so, the buyer and seller could share in the savings of the agent’s fee, while the agent would be left uncompensated for her efforts.

  For our purposes, one major lesson emerges: although these management mechanisms can reduce principal-agent differences, none of them eliminates the tension completely, alone or in combination. Our third tension is inescapable: there are always agency costs. In a particular context, some mechanisms will obviously be better than others. But reputational markets are never perfect. Monitoring is always costly. And any compensation scheme creates incentives that can be perverse in some circumstances. In a relatively simple transaction such as a real estate sale, the parties may not find it worthwhile to expend resources writing elaborate agency contracts. To do so would just further shrink the pie. In addition, trying to exert control over an agent can have paradoxically negative consequences on the agency relationship: in part, agents are value-creating for their principals because they are independent decision-makers, not puppets.

  Principal-Agent Problems in the Legal Context

  The principal-agent relationship of most interest to us here is the relationship between a lawyer and a client who are involved in a legal negotiation. Like all other agency relationships, this one poses problems for both parties, owing to differences in preferences, information, and incentives. Here, we briefly outline some management mechanisms that can dampen the principal-agent tension when it arises in the context of a legal negotiation.

  INCENTIVES

  To tackle incentive problems, lawyers and clients have developed an array of fee structures—all inevitably flawed. The most common of these are:

  • Contingency fee

  • Hourly fee

  • Fixed fee

  • Mixed fee

  • Salary

  Contingency Fee: In this arrangement, the lawyer earns a percentage of the recovery, if any, that he wins for the client. This structure is most often used by plaintiffs’ attorneys in tort litigation, and it has the same advantages and disadvantages as the percentage fee in our real estate example. Its chief advantage is that the plaintiff pays nothing unless there is a recovery. A contingent fee also enables a plaintiff to engage in a lawsuit that she otherwise might not be able to afford. In essence, the client is selling the lawyer a third of her lawsuit in exchange for the lawyer’s services. It is a reasonably effective way of aligning the parties’ interests, in that the lawyer has an incentive to win a large recovery for the client. The incentives are not perfectly aligned, however, because the lawyer is putting in all the effort and only receiving a fraction of the benefit. The contingent-fee lawyer may be better off with a quick settlement that takes little effort rather than a higher recovery that requires substantially more work. A contingency fee can also allow a client to exploit her attorney. Plaintiffs’ lawyers typically screen their cases carefully because they are bearing part of the risk of failure.6

  Hourly Fee: Under this arrangement, the lawyer is paid by the hour. This fee structure is most often used by defense counsel in litigation and by deal-making attorneys. Its advantage is that it motivates the lawyer to devote the time needed to achieve the best result for the client—particularly when it is not clear from the outset how much time the matter will consume. The disadvantage for the client is that it removes any necessary link between the benefit the lawyer’s work confers on the client and the amount the client pays. The lawyer may be tempted to do more work in order to earn more, even if the work is unnecessary. On the other hand, hourly billing may disadvantage the lawyer in some circumstances. For example, lawyers may be reluctant to charge on the basis of normal hourly fees when the lawyer’s special expertise and experience can produce very substantial economic benefits for a client in a short period of time.

  Fixed Fee: Here, the lawyer earns a specified amount for handling a particular legal matter. This arrangement gives the lawyer an incentive to get the work done in as short a time as possible, and it caps costs for the client. On the other hand, the client may have an incentive to try to expand the scope of the work covered by the fixed fee.

  Mixed Fee: Hybrid fee arrangements are becoming increasingly common. For example, a client may pay her lawyer a diminished hourly rate plus a bonus if the lawyer achieves good results. Although a hybrid of this sort may align incentives reasonably well, it is often difficult to implement. A precise formula for computing the bonus may be hard to establish in advance, especially where there is no single, easily measurable benchmark for a good outcome. The parties may simply agree to negotiate the amount of the bonus after the fact, but at that point lawyer and client may have different notions of what more, if any, the lawyer deserves.

  Salary: A salaried lawyer works for a single client, whether a government agency or a private corporation. Bringing counsel in-house does not eliminate the principal-agent tension, however. The lawyer still has interests of her own. The incentive effects will depend on the details of the salary arrangement and career paths within the organization. Compensation may share the characteristics of either a fixed fee arrangement or even hourly fees, depending on how the lawyer’s pay is computed. In-house counsel are sometimes thought to be more risk-adverse and less willing to provide independent legal advice that the client may not want to hear, because their career depends on keeping the favor of a single client.

  MONITORING SYSTEMS

  The principal-agent tension may be dampened by monitoring the agent’s activities. This is difficult and expensive in the legal context, with respect to both inputs and outputs. To know whether a lawyer is acting solely in his client’s interest, the client must possess enough knowledge to evaluate the lawyer’s decisions and must be able to observe the lawyer’s behavior. There may be no easy way for a client to verify information about a lawyer’s true work habits, diligence, or timekeeping practices. Similarly, it may be quite expensive for a client to monitor the quality of her lawyer’s work, unless the client is herself an attorney. Often in-house corporate counsel can monitor the activities of outside counsel, but this is hardly a cost-free solution.

  REPUTATIONAL BONDING

  To the extent that potential clients have access to accurate infor
mation about an attorney’s reputation, the attorney will have more incentive to build and maintain a reputation for trustworthiness and hard work. If an unsatisfied client can go elsewhere in the market for legal services, a lawyer is more likely to act loyally and diligently to keep that client.

  But this constraint is an imperfect one. Once a lawyer-client relationship has been established, it is often very costly for the client to leave one lawyer and start a new relationship with another. In the middle of a lawsuit or a complicated transaction, for example, a new lawyer would have to invest a great deal of time to learn what the old lawyer already knows about the matter. Because the client will typically have to pay to educate his new lawyer, these extra costs of switching lawyers midstream mean the market cannot completely constrain opportunism. For this market constraint to operate most effectively, moreover, clients must be able to evaluate the performance of their lawyers, which, as noted above, is no simple matter.

  To the extent that a lawyer and client expect their relationship to extend over time, each is less likely to act opportunistically in the present. If the shadow of the future is long, the risk of losing future business may deter present disloyalty. In the corporate world today, however, steady long-term relationships with outside counsel are becoming the exception, not the rule. Rather than long-term retainers, clients increasingly hire lawyers for a single transaction or for a particular lawsuit.7 In such short-term one-shot relationships, each side may be more tempted to try to exploit the other.