Beyond Winning

by Robert H Mnookin

  As this example suggests, in legal negotiation the lawyer’s role as an advocate can lead him to have a perspective that is as distorted as his client’s, if not more so. A plaintiff’s lawyer may look for evidence that confirms his view that the plaintiff should recover and may only ask questions that uncover such information. His research may focus on finding cases that support his client’s position, and he may put little effort into research to develop the other side’s theory of the case. He may write memoranda arguing for his position and repeat his argument again and again with his colleagues and family members. Although all of this rehearsal is part of preparing for litigation, it also biases him because he internalizes information selectively. He may end up being no more able than his client to analyze dispassionately the strengths and weaknesses of his client’s position.

  In deal-making, lawyers also may exacerbate the problem of partisan perceptions. Experienced transactional lawyers are expected to protect their clients from risk, and they can become expert at imagining an extraordinary array of contingencies that might adversely affect their clients’ interests. An attorney may see as “realistic” the need to be highly suspicious of the lawyer and client on the other side—no matter who they are or what they do. The same lawyer may take offense at the partisan nature of the other side’s demands and expectations.

  It is of course possible that having a lawyer may dampen a client’s tendency to have partisan perceptions and may help the client reach a more informed and objective sense of his situation and of its legal opportunities and risks. In preparing for litigation, for example, lawyers are taught to anticipate what arguments the other side might make and to ask themselves how the other side might see a given situation. Similarly, a deal-making lawyer may be more able than her client to understand the other side’s interests and why a given term may matter. In both domains, a lawyer can act as a bridge by helping his client understand the perspective, interests, and arguments of the other side. But to do this effectively, a lawyer must be aware that he—like all other human beings—can have partisan perceptions and that his role as an advocate can lead him to fall prey to distorted thinking.

  Judgmental Overconfidence

  Psychologists have documented that people often place unwarranted confidence in their own predictions about future events. There are two aspects to this phenomenon: how accurate you are in your basic judgments; and how well calibrated you are in assessing the risk that you may be wrong. In our workshops, for example, we often ask participants a series of ten questions, such as “When did Attila the Hun fight the great battle of Chalons?” or “How many years was Oliver Wendell Holmes on the United States Supreme Court?” Participants must first make their best guess. Next, they are asked to set upper and lower bounds for each answer so that they are 95 percent confident that the correct answer will fall within their estimated range. For example, if a person thought that Attilla the Hun waged his campaign in 975 AD, and if he was very confident of the answer, his range would be narrow (say, from 875 to 1075 AD); but if he was less confident, he would have a broader range (575– 1375 AD). Obviously, getting the precise answer is difficult. Few people recall that the Hun fought the battle in 451 AD. But if a person is well calibrated about the risk of error, when asked for a 95% confidence interval, his range should be broad enough to include the right answer 19 out of 20 times. Most people, however, are overconfident—their range includes the right answer only about half the time.

  Consider the effect of this dynamic on dispute resolution. Overconfidence can lead litigants to overestimate their own chances of winning at trial—mediators often see cases in which the defendant and plaintiff each sincerely believes that his own side has a 75 percent chance of winning and only a 25 percent chance of losing. They both can’t be right, and indeed they both may be wrong. Max Bazerman and Margaret Neale demonstrated this in the context of an experiment involving final-offer arbitration in which the subjects were told that the arbitrator would be required to choose the final offer of either one party or the other. The disputants systematically overestimated the chance that their own offer would be chosen—by more than 15 percent on average.4

  Why may negotiators be overconfident? Researchers have proposed a variety of explanations, some of which are suggested by our discussion of partisan perceptions. The most important reason is that each party may have access to only some relevant information and may underestimate the importance of what they do not know.5

  One might hope that lawyers could help clients dampen or eliminate this bias by searching for disconfirming evidence and providing a more independent, expert, and experienced assessment. This sometimes happens. But comparing one’s own assessment with an advisor or with a group of peers has been shown in some instances to increase judgmental overconfidence.6 There is some research (not involving lawyers) suggesting that consulting with others does little to improve the objective accuracy of a person’s predictions, and instead makes that person more certain that his prediction is right.7

  Loss Aversion

  Daniel Kahneman and Amos Tversky have demonstrated in a brilliant series of studies that decision-makers tend to attach greater weight to prospective losses than to prospective gains, even when what is a gain or loss may depend on how the outcome is framed in relation to an arbitrary reference point.8

  People do not always behave rationally in the face of uncertainty and risk. Consider the following two sets of choices:

  (1) Please choose between:

  (a) A sure gain of $240;

  (b) A 25 percent chance to gain $1,000, and a 75 percent chance to gain nothing.

  (2) Please choose between:

  (a) A sure loss of $750;

  (b) A 75 percent chance to lose $1,000, and a 25 percent chance to lose nothing.

  If you chose (a) in the first problem and (b) in the second, you were not alone. As Tversky and Kahneman have shown, in the first choice, 80 percent of subjects pick the sure gain of $240, even though the expected value of choice (b) is greater ($250). People are generally risk-averse regarding gain and would rather have a bird in the hand than two in the bush. Facing the second choice, however, 80 percent of subjects generally gamble and choose (b). People are loss-averse—we dislike losing money—and therefore are risk-seeking concerning losses if there is some chance, however small, of evading any loss at all.

  Loss aversion may tend to make it more difficult to settle disputes. Some defendants may decide unwisely to litigate rather than settle out of court because they choose to risk a large loss rather than accept a smaller but certain one. Plaintiffs, on the other hand, may be more willing to accept a modest but certain settlement rather than gamble on the prospect of a potentially larger but uncertain gain through litigation.9

  Russell Korobkin and Chris Guthrie have run experiments that suggest that framing may affect a choice to settle rather than litigate. They contend that whether a settlement is viewed as a gain or as a loss can influence a client’s decision.10 Their experiments involved the following two situations. Would you accept settlement in one, the other, both, or neither of the following stories?

  STORY A: Your new $14,000 Toyota Corolla was totaled in an accident that was clearly the other driver’s fault. Your $14,000 medical bills were paid by your medical insurance, but you have no insurance to replace your car. The other driver has no money and is unemployed, but he does have automobile insurance. You have sued the driver’s insurance company for $28,000—the cost of the car plus medical bills. The defendant does not dispute your damages, but it claims that the policy has a maximum coverage value of $10,000 for accidents that occur while driving a rental car. Your lawyer tells you that if the case goes to trial and the judge decides there is such a limit, you will recover only $10,000. If the judge advises there is no such limit, you will recover the full $28,000. Your lawyer advises you that the language in the other driver’s insurance policy is extremely unclear, and he cannot predict whether you will win or lose. The defendant’s fin
al settlement offer is $21,000. You will have no legal fees. Do you accept their offer?

  STORY B: All the facts are the same as story A, except that your $28,000 of damages consists of the total loss of your new $24,000 BMW and $4,000 in medical bills that were paid by your medical insurance. Once again, there is a $21,000 final settlement offer.

  In their experiments, Korobkin and Guthrie found that undergraduate subjects were more likely to accept a settlement in story A than in story B, even though the opportunities and risks of the litigation were identical.11 Why might this be?

  In story A a $21,000 settlement allows the plaintiff to buy a new Toyota and still have $7,000 left over. This can be seen as a windfall. In the second case, a settlement for the same amount cannot replace the BMW. This can be framed as a loss of $3,000. However, the comparison of the settlement and the expected payoff of the litigation gamble is the same in both cases. Loss aversion provides a plausible explanation for the different experimental results.

  Are lawyers as prone to loss aversion as clients? One might plausibly believe that if a lawyer saw the opportunities and risks of litigation in these cases as identical, she would be no more likely to litigate in the second case than in the first. Interestingly, Korobkin and Guthrie repeated their experiments with lawyers in the San Francisco area, who were asked what they would recommend to their client when they were told that “after conducting a legal and factual investigation . . . you cannot predict whether your client is more likely to win or lose if the case goes to trial.” Unlike their other subjects, the lawyers’ settlement recommendations did not vary—the framing effects of story A and story B made virtually no difference. That result suggests that lawyers may not be as prone to loss aversion as clients. Obviously, this is only a single study, and a great deal more research would have to be done before one could have confidence that this is indeed true. Nevertheless, it is an intriguing result.

  Loss aversion does suggest that how a lawyer frames a settlement—as a loss or a gain—may make a difference either to his own client or to the other side. Mediators have long noted that they are more successful in settling a case when they point out what a party has to gain—disposing of the matter, avoiding additional costs, or achieving certainty—than when they emphasize what settlement will cost. This is the strategy the insurance industry uses to induce potential customers to buy insurance: it emphasizes the protection gained against a large but uncertain future loss, rather than emphasizing the sure but smaller loss we will suffer today in the form of insurance premiums.

  Loss aversion also may have an effect on deals. Consider the following imaginary experiment in which an owner of a piece of property needs to choose between two offers to buy. Buyer A offers $1 million, with $500,000 to be paid in cash at the closing and the remainder payable one year after the sale. Buyer A, however, faces a 50 percent chance of going broke during the year. If buyer A never pays the second installment, the seller will have no further recourse. The second buyer, buyer B, offers $775,000 all cash. Half the subjects would be told that they had inherited the property. The other half would be told that they had bought the property two years earlier for $1 million and now needed to sell. They would have to choose between the same two offers. Would those who bought the property for $1 million be more likely to gamble by selling to buyer A, in order to avoid the sure loss? More broadly, in deal-making if you were a buyer and you knew the seller faced a loss on the transaction, could you take advantage of loss aversion? Perhaps you could, by creating a higher nominal offer with a portion of the price structured as a contingent pay-out where it would be unlikely that you would ever have to pay the full price.

  Endowment Effects

  Another bias, a first cousin of loss aversion, is the endowment effect. When something belongs to you, you may attach a greater-than-market value to it because it is yours—you own it. Having to give it up voluntarily may represent a special kind of loss to you—the loss of an endowment—for which you will want extra monetary compensation.12

  Kahneman, Knetsch, and Thaler conducted an intriguing experiment that illustrates this phenomenon with coffee mugs.13 They gave each individual in one randomly selected group of subjects—the sellers—a coffee mug and told them that they could either take the mug home or sell it. The experimenters would buy the mug back for an unspecified but predetermined price. The sellers had to write down what price they would be willing to accept for their mugs. If their asking price was lower than the predetermined amount (which turned out to be $5), they got the cash. Otherwise, they kept the mug. For a second group of subjects, the experimenters didn’t hand out mugs. Instead, they asked how much the subjects would spend to buy a mug.

  The median asking price for sellers was $7.12, whereas the median offer price for buyers was $2.88. Although those with and without mugs faced the same choice—go home with a mug or with cash—those who already owned a mug demanded more to be compensated for losing it.

  Consider how the endowment effect may inhibit deals, especially where it is not easy to determine the market value of the goods or services at the heart of the transaction. A seller may have an exaggerated notion of the value of what he is selling, and the buyer’s and seller’s perceptions of the value of the transaction may vary greatly as a result. Of course, a buyer might suggest the use of a neutral appraiser to counter this effect. And because a party’s lawyer does not own the business or property that his client is selling, presumably she should not be tainted by this bias. Nevertheless, endowment effects may preclude some transactions that would otherwise make both sides better off than no agreement.

  Reactive Devaluation

  Lee Ross and his students have done experimental work suggesting that the evaluation of a concession, a deal term, or a proposed compromise may be different, and may change, depending on its source. In particular, a party may devalue a proposal received from someone perceived as an adversary, even if the identical offer would have been acceptable when suggested by a neutral or an ally. They have also demonstrated that a concession that is actually offered is valued less than a concession that is withheld, and that a compromise is rated less highly after it has been put on the table than beforehand.14

  One study of this phenomenon examined students’ attitudes toward Stanford University’s divestment from companies invested in South Africa in the 1980s.15 At the time, divestment was a very controversial issue on campus. The researchers asked subjects to read a brochure detailing the controversy and explaining two potential compromise proposals—the specific divestment plan, which entailed immediate divestment from corporations doing business with the South African military or police, and the deadline plan, which proposed creating a committee of students and trustees to monitor investment responsibility with the guarantee of total divestment within two years. There were three different versions of the brochure. For one group of students, the brochure said that the university supported the specific divestment plan. For another group it said that the university backed the deadline plan. And a third brochure gave no suggestion that the university supported either plan.

  As predicted, students devalued the plan that they believed the university supported. When Stanford was said to support the deadline plan, 85 percent of students thought that the specific divestment plan was a bigger university concession than the deadline plan. When the university was said to support specific divestment, only 40 percent thought it was a larger concession—most saw the deadline plan as a better deal for the students. And when the university was left out of the brochure altogether, approximately 69 percent thought that specific divestment was a bigger concession by the university—somewhere in the middle.

  Reactive devaluation operates in many legal negotiations. Lawyers frequently report that clients who initially are enthusiastic to settle for a given amount are disappointed when the other side actually offers that amount. If a neutral or third party offered clients the same settlement, they’d take it. But when it comes from the other side, they
back away.16

  Reactive devaluation occurs in part because a negotiator may assume that any proposal coming from the other side must benefit that side and that “anything good for my adversary can’t be good for me.” It is true that an offer must benefit the other side—otherwise they wouldn’t have offered it. It is also true that authorship is relevant to its evaluation. And it may be that by rejecting the offer a lawyer’s client could secure an even greater concession. But it certainly isn’t the case that a proposal made by the other side—for that reason alone—cannot serve your client’s interests as well. By helping a client evaluate a proposal in terms of the client’s own interests and alternatives, a lawyer may be able to counteract the tendency toward reactive devaluation.

  Emotion

  Emotions can run high in both legal disputes and deal-making. In some circumstances, passions can serve negotiators. It may be beneficial to be able to threaten credibly to act counter to your self-interest—and emotions may help you do that. Showing real anger at an unfair deal—even if both sides know the offer on the table is better than your best alternative—may persuade the other side to better its terms. And evincing a genuine disposition to behave honestly even in the face of temptation to cheat—something that research shows is not easily faked—may be the most important asset a negotiator can have.17