ADR DECISIONAL ERRORS Why We Make Them and How to Address Them

mpiring errors were a bigger story in the recently completed than the U ’ record-setting 27th pennant. Decision errors reached comic propor- tion in game four of the playoffs in Anaheim. In what ESPN dubbed “the worst umpiring performance at an Angels game since Leslie Nielsen in The Naked Gun,”1 third-base Tim McClelland called Yankee

Umpiring errors reached comic proportions in the recently completed World Series, highlighting the fact that people reg- ularly make decisional errors. There are many reasons for this and a variety of dispute resolution processes that can be tailored to fit the specific fuss in a cost-effective way.

BY DONALD R. PHILBIN JR. Don Philbin is an attorney-mediator, negotiation consultant and trainer, and arbitrator. He has resolved disputes and crafted deals for more than 20 years as a commercial litigator, general counsel and president of communications and technology-related companies. Don has mediated hundreds of matters in a wide variety of substantive areas and serves as an arbitrator on several panels. He is an adjunct professor at the Straus Institute for Dispute Resolution at Pepperdine Law School, Chair of the ABA Dispute Resolution Section’s Negotiation Committee, and a member of the ADR Section Council of the State Bar of Texas. Don is listed in The Best Lawyers in America (Dispute Resolution), The Best Lawyers in San Antonio, and the Bar Register of Preeminent Lawyers.

Reprinted with permission from the Dispute Resolution Journal, vol. 64, no. 4 (Nov. 2009-Jan. 2010), a publication of the American Arbitration Association, 1633 Broadway, New York, NY 10019-6708, 212.716.5800, www.adr.org. ADR

safe when he was out, out when he was safe, and think that if humans were completely rational and “blew an obvious call on what should have been a shared all the available information, they would double play at third base.” McClelland could not reach similar decisions. But you would be wrong. even explain why he missed what seemed obvious This was confirmed by a study showing that peo- with the benefit of instant replay. ple with exactly the same information reach differ- Fortunately, McClelland’s mistakes were not ent conclusions. For example, one study showed outcome determinative, but they fueled the case for that buyers rarely want to pay as much as sellers expanding Baseball Commissioner Bud Selig’s lim- demand. Subjects asked to price a generic coffee ited replay system beyond disputed calls. cup for sale assigned it a value of $7.12. Their But those of us who work in the dispute reso- would-be buyers initially offered $2.88 for the lution field were probably not surprised by the same cup—2.5 times less.8 While the different val- umpiring errors, since we know uations may largely be a matter of that errors are inevitable in sports, assigned position, the spread our personal and business lives, Decision-making between the offers is attributable and certainly in the crucible of in large part to the assigned posi- litigation. As Yankee Derek Jeter errors have tion of the parties making or put it post-game: “Umpires are many causes. receiving those offers. human. They make mistakes Interestingly, people with less sometimes.”2 The negotiator for Overconfidence information about an issue are the umpire’s union also noted more definitive in their position that players are just as susceptible can certainly than are those with more com- to making mistakes: “Part of the plete information. They tend to game,” he said, “is the potential push parties to undervalue aspects of a situation for human error, not just for about which they are ignorant. In umpires, but for players, that’s the margins of one study, subjects given only part of the game.”3 half of the evidence in a case pre- Decision-making errors have a negotiating dicted the jury’s decision with many causes. Overconfidence can range. It’s human greater confidence than those certainly push parties to the mar- who were given all of the infor- gins of a negotiating range. It’s nature to place mation.9 Not only were they human nature to place more more confident than those who emphasis on facts supporting a more emphasis were better informed, they were desired outcome than those that not able to adequately compen- contest that position. Over- on facts sup- sate for their overconfidence confidence leads people to dis- when told that their evidence was count small probabilities and porting a desired lopsided. Not only are we predis- luck, and overestimate their own posed to camp near our positional situation. It’s sometimes referred outcome than interests, it’s human nature not to to as the “Lake Wobegon” effect. those contesting fully appreciate another perspec- Researchers have begun to tive. A noted 19th Century quantify the magnitude of this that position. Boston educator observed that, cognitive error. In one study, “To be ignorant of one’s igno- more than 80% of interviewed rance is the malady of the ig- entrepreneurs described their chances of success norant.”10 as 70% or better, and 33% described them as Decision-making errors could also involve the “certain.” 4 That compares with a 33% five-year phenomenon called “anchoring” (an over- survival rate for new firms. In another study, cou- reliance on a particular fact or piece of informa- ples about to be married estimated their chances tion, especially information dropped early to of later divorcing at zero, even though most knew anchor negotiations). The problem is that the that the divorce rate is between 40%-50%.5 anchor may be unreasonable and increase the Yet another study found that negotiators in odds of impasse and other unintended conse- final-offer arbitration6 overestimated by 15% the quences. There is evidence that even knowledge- chance that their offer would be chosen.7 able professionals are susceptible to manipulation Optimistic overconfidence can make it very dif- of anchors. By varying real estate asking prices, ficult for parties to see the same problem similarly. researchers were able to throw professional real It can be involved whether we have equal infor- estate agents off track in negotiations.11 mation or asymmetric information. You might Negotiation training, or obtaining a legal

2 NOVEMBER 2009/JANUARY 2010 opinion or an economic analysis, can help assess that “there are highly skilled doctors who get the reasonableness of an anchor. It can also place sued a lot and doctors who make lots of mistakes us in a better position to decide whether to make and never get sued.” He says the differentiator is a first offer or wait, or disregard an offer or not shoddy medical care but rather a shoddy atti- demand by the other party that is based on an tude: “[P]atients say that they were rushed or unreasonable anchor. ignored or treated poorly” and it made them Then there is the phenomenon of reactive mad. Gladwell quotes medical malpractice lawyer devaluation, in which people tend to reject or Alice Burkin: “People just don’t sue doctors they devalue whatever the other side offers, even if it’s like ... [A good] bedside manner and a willingness favorable (“They wouldn’t have offered those to answer patient questions are effective ways to terms if those terms strengthened our position reduce the odds of facing a malpractice suit,” she relative to theirs.”). We also tend to reject things said. that are free and want things that are expensive— Judges also makes mistakes—or at least many an example of the adage that the “grass is always litigants think so after they force a judge to greener on the other side of the fence.” decide a case that they could not resolve among Why do people reactively devaluate offers themselves. While we do not know that much from the other party? The simple fact is that about judicial decision-making, efforts have been there are things we just do not want to hear from made to quantify decisional errors among judicial adversaries. Our perception of the source of the populations. Chris Guthrie, Jeff Rachlinski and offer colors our perception of the offer. This was Andrew J. Wistrich, using Gladwell’s “rapid cog- demonstrated in an experiment in which different nition” concept (i.e., that people make decisions groups of students were told a different story based on a snap answer that comes to mind first; about the Cold War. One group was told that a Gladwell calls it a “blink”) attempted to deter- group of unknown strategists made a proposal to mine whether judges make snap decisions.14 reduce nuclear warheads by one-half, followed by They asked over 250 Florida trial judges who further reductions over time.12 Another group were attending an annual judicial conference to was told that President Reagan made the propos- answer a survey questionnaire containing three al, and a third group was told that Soviet leader questions (called the Cognitive Reflection Test): Gorbachev was the source of the proposal (he 1. If it takes 5 machines 5 minutes to make 5 was the real author). The surprise was not that widgets, how long would it take 100 the groups reacted differently to the proposal machines to make 100 widgets? depending on its source, but the wide range of 2. A bat and a ball cost $1.10 in total. The bat difference between two of the groups. When costs $1.00 more than the ball. How much attributed to the U.S. President, 90% reacted does the ball cost? favorably. That dropped marginally (to 80%) 3. In a lake, there is a patch of lily pads. Every when attributed to unknown negotiators, and day, the patch doubles in size. If it takes 48 then to 44% when attributed to the Soviet leader. days for the patch to cover the entire lake, Thus the arms control proposal from “unknown how long would it take for the patch to strategists” and President Reagan were viewed cover half the lake? almost equally and twice as favorably as when the The questions were designed to distinguish “in- proposal came from the opponent. tuitive” from “deliberative” thinking since each Strong emotions, like anger, can be a powerful had an intuitive wrong answer and a deliberative trigger of judgment errors. When people are right answer. The intuitive snap answer to ques- angry they can make “attribution errors”—i.e., tion 1 was 100 minutes, but the right answer was they take credit for positive outcomes and attrib- five minutes. The snap answer to question 2 was ute negative outcomes to external factors, no 10 cents but the right answer was five cents. matter what actual caused them.13 When making Finally, the snap answer to question 3 was 24 decisions, it is important to contemplate the days, while the right answer was 47 days. attributions we are making to another person, How did the judges do? Just 28.2% picked the and the ones they are surely making to us. right answer to question 1; 44% answered ques- Our attitude toward others impacts their atti- tion 2 correctly; and 50.4% gave the correct tude toward us. Not only does that attitude affect answer to question 3. This represented an aver- the negotiation dance, it shows up in claim esca- age correct score of 1.23 out of 3.00. lation. For example, doctors who treat patients But more interesting is the fact that 31% of with respect tend to be sued less often. In his the responding judges did not get any of the bestselling book Blink: The Power of Thinking questions right. True, there could be inherent Without Thinking, Malcolm Gladwell observes problems with the questions, which were remi-

DISPUTE RESOLUTION JOURNAL 3 ADR

niscent of intelligence tests and SATs taken in Alternatives to Litigation high school, not at all like the “application of law Since most disputes don’t have to be resolved in to fact” questions that judges are used to answer- litigation, and most disputants do not want the res- ing on the bench. In addition, the judges who olution process to take years and cost a mint, we responded to the questions during a packed con- have alternatives to going to court. At the Pound ference may not have put much effort into Conference in 1976, Frank Sander, a pioneer in the answering them. But it nevertheless suggests that field of alternative dispute resolution, recognized judges are susceptible to the same decisional that there are multiple ways of processing disputes. errors that plague the rest of us, even other It was at that conference that he coined the “multi- umpires.15 That should trouble disputants who door courthouse” metaphor.18 Sander also authored choose to litigate over making a deal themselves the concept that the dispute resolution process during negotiations, in mediation or reaching a should “fit the fuss.”19 settlement on the courthouse steps.16 That goal is even more important today as the Besides judicial errors, there are other struc- discovery conundrum is magnified with electron- tural features of our judicial system that foster ic discovery. Retrieving and producing a usable decisional error. For example, it is a rigidly rule- email can be debilitating—for whichever side has bound and adversarial process, with parties repre- to bear the expense. No matter whether the dis- sented by warrior advocates, who test the veracity putes involve sports figures, a family in conflict, of adversary witnesses at trial through live and an employee-employer problem, or a business often acerbic cross-examination. Our uniquely deal gone south, the process can be and should be Judges also make mistakes—or at least many litigants think so after they force a judge to decide a case that they could not resolve among themselves.

American rule allowing liberal pre-trial discovery tailored to the dispute. of information—including e-mail and other Cost and time are critical factors in designing forms of electronic information—in the posses- processes to avoid or reduce decisional errors, sion of the adversary and third parties often leads even when made by the disputing parties them- to delays, procedural disputes, and pre-trial selves. What works in one area does not always motions. It could result in the discovery of rele- work in another. For example, a World Series vant information, but whether it does or not, the pennant game broadcasted nationally can afford cost is high. Litigants may need a war chest as to have 12 cameras to record different plays for large as the New York Yankee’s payroll to fully instant replay purposes. But that would be too participate in a process that may try their pa- costly an approach in little league baseball. The tience while taxing their treasure. result is that we sometimes read about parents It’s also difficult to tailor litigation to specific who become violent when an unfavorable call is disputes. We follow one-size fits all procedural and made involving their child. evidentiary rules regardless of the scope of the Arbitration and mediation are both well- issues involved or the magnitude of the damage. known ADR systems. Arbitration results in a Of course, many criminal matters17 and civil binding and final decision by a private neutral disputes in which a legal precedent needs to be decision maker, called an arbitrator. The parties forged must go through the court system. And can select the arbitrator, which allows them to some disputes are clearly unsuited for other select an expert in the field of the dispute. Parties means of dispute resolution. For example, it is can agree to use arbitration’s efficient procedures likely that even the most capable mediator would and set time limits for the issuance of an award. have had trouble mediating Brown v. Board of To address the reality that arbitrators can also Education. The parties’ distant positions were on make decisional errors too, there is a system for bright display in that case, as were the underlying limited review of awards by trial courts,20 or if the interests and biases that animated their dispute. parties agree, an appellate arbitrator or panel. Indeed, it took the power of the federal courts Mediation has been wildly popular in the and the artillery of the National Guard to get United States, probably because the parties retain some of those parties to yes. ultimate control over the outcome while leverag- ing the talents of a neutral mediator to reduce

4 NOVEMBER 2009/JANUARY 2010 human bias. The Reagan-Gorbachev study quanti- facing a gain (they prefer a sure gain over a less fies what we intuitively know—the same idea will certain larger gain) and risk-seeking choices be received more favorably from a neutral party when facing a loss (they prefer riskier outcomes than an opponent—by a factor of two. Since we all to sure losses). So one group of people facing a like decisions we make better than those that are gain would prefer receiving $240 in hand to a foisted upon us, mediated agreements tend to be 25% chance of receiving $1,000 (worth on aver- more self-enforcing than judgments. age $250). The same group, when facing a loss, The psychological causes of decisional errors would prefer a 75% chance of losing $1,000 can come into play in these processes as well. (worth $750) to a sure loss of $750.23 There are, however, techniques that can address But these risk profiles can change. For exam- them. For example, counsel can use decision trees ple, many game show contestants are risk takers. and animated scenario modeling to display and You probably won’t see many negotiation theo- assign probabilities to various outcomes, includ- rists risking humiliation on national television, ing whether to negotiate, litigate, arbitrate or even if the payoff is potentially high. Having said mediate, or use some other process (see Figure that, a Harvard-educated entertainment lawyer A),21 or how much to pay to close an information recently went for broke on “Who Wants to Be a gap that has been framed by an offer. The Millionaire?” He astutely answered a long series amount has everything to do with the spread of questions, accumulating a purse of $475,000. between the decision points (litigate v. settle).22 That series of wins no doubt inflated his confi- A technique employed to curb reactive devalu- dence in his ability to play this game. When ation in mediation is to reframe the discussion in asked whether he wanted to risk his hard-won more positive terms and work through various purse on a single question that could double his possible outcomes. payoff to $1 million, he said yes. The problem But clients want their lawyer to be their cham- was that he didn’t know the answer to the half-a- pion and agree with their optimistic view of the million dollar question and it was not self-evident case. Most do not like their champion poking or logically deducible. He opted to ask the audi- holes in their case. This is why counsel will often ence for its opinion and was ultimately influenced suggest getting an opinion from an expert, or if by its incorrect answer—a stark example of the the case is in mediation, will urge the mediator to phenomenon of anchoring. provide a reality check. Mediators, not counsel, What happened here is that a very smart, ini- are the ones who generally use reframing to tially successful player turned into a prisoner of move the mediation along. psychological biases. He must have realized the The bottom line is that if the parties’ positions ebbing of his overconfidence when during a seem- are aggressively opposite and neither is tempered ingly long period of introspection upon hearing by a reality check, an impasse is likely to occur. the problematic question, he generalized that risk This scenario tends to lead the parties to embark seekers usually go to business school while law on a dispute resolution process involving a decision schools attract the risk averse. by a third party, such as arbitration or litigation. Risk profiles have an impact on deciding which A key psychological factor that tends to affect dispute resolution process to use. In the classic most decision making is the person’s tolerance article, “Fitting the Forum to the Fuss: A User- for risk and aversion to loss. Some people are Friendly Guide to Selecting an ADR Procedure,”24 risk-takers and others are risk-averse. Attitudes Sander and Goldberg focused on: (1) achieving the toward risk are dependent, in part, on whether disputants’ goals in making a forum choice, and (2) the party believes they face a gain or risk a loss. the obstacles that their choice might overcome.25 Plaintiffs generally seek recoveries that defen- Leonard Riskin focused on mediation as the forum dants resist paying. Unless they have sunk a great of choice and described different mediator styles deal of money into the transaction or lawsuit, or using a two-dimensional “grid.”26 Those styles their agreement contains a fee-shifting provision range from facilitating dispute-focused conversa- (i.e., a loser pays clause), plaintiffs tend to believe tions to offering case evaluations. A neutral media- that they will have a gain from settling and per- tor may elicit “best” and “worst” outcomes by haps a bigger gain from going to trial. In the leading disputants through a process of “scenario absence of strong counterclaims or an offer of assessments” that takes into account psychological judgment that may change the cost dynamic, biases on both sides. defendants are in the opposite position: they face Some scholars oppose mediators making eval- a sure loss by settling and perhaps a bigger loss uations27 and many disputants are sorry they ask- by going to trial. ed for an evaluative answer after they get it. But People tend to make risk-averse choices when the fact is that disputants will not agree to resolve

DISPUTE RESOLUTION JOURNAL 5 ADR

the disputed issues unless they believe the settle- direct party-to-party negotiations. Direct negoti- ment terms are better than they would get at ation is the most frequently used dispute resolu- trial, taking into account the cost of litigating, tion technique. The negotiators retain complete including appeals, the time value of money, and control over the process and the solution. how long it might take to resolve the case in An alternative to doing nothing is choosing to court.28 use cooperative negotiations or collaborative law. If disputants choose not to mediate, they could These are related processes in which the parties simply ignore the dispute and then see what hap- usually commit in a “participation agreement” not pens. It could get worse or go away. If it doesn’t to file a lawsuit until they have participated in a go away, disputants could decide to resolve it in number of problem-solving meetings and exhaust- the same way they got into their deal—through ed negotiations. Finally, they could go to a more

Figure A Dispute Resolution Options Ignore X Direct Negotiation X Cooperative Negotiators X Collaborative Law X Facilitation X Fact Finding X Transformative X Facilitative X Collaborative X Counsel Ellicitative X Mediation Evaluative X Evaluative-Directive X Wisely Directive X Directive X Med/Arb X Conciliation X Mini-trial (Mock Bench Trial) Dispute Arises X Summary Jury Trial (Mock Jury) X Early Neutral Evaluation X Private Judging X Settlement Conference X Arb/Med X Appellate Panel X Class Actions X Complex/Large Adjudicative Arbitration X Expedited Proceedings X Final Offer (Baseball) X Bounded X Foreign Direct Investment X Settlement Counsel X Bench Trial Litigate X Jury Trial Legislation X X

6 NOVEMBER 2009/JANUARY 2010 adversarial procedure—litigation or arbitration. different fusses. Baseball can decide how broadly Brain scientists are decoding what psychologists it will use instant replay to sort out major league have observed in the human condition for decades: mistakes. Disputants increasingly recognize the Everyone makes mistakes. Being a major league advantages of using win-win models of dispute umpire, a seemingly rational and objective lawyer, resolution over win-lose (or win at any cost) or even a judge does not remove the human sus- models, and are implementing systems to manage ceptibility to making cognitive errors. So we turn conflict upstream. our attention to reducing their frequency in nego- In addition, the role of counselor has necessar- tiation. And, like most endeavors, that involves ily broadened to full-spectrum representation cost and balance, particularly when it comes to and market pressures will continue that trend. deciding on a conflict resolution system. Conflict is inevitable. But the ways in which we Fortunately there are now many forums for manage it is evolving rapidly. n ENDNOTES 1 http://sports.espn.go.com/mlb/ 13 Dan Lovallo & Daniel Kahneman, tained in an appendix to Smith & playoffs/2009/columns/story?colum- “Delusions of Success: How Optimism Martinez’s article, supra n. 6, at 165. nist=caple_jim&id=4581598;www. Undermines Executives’ Decisions,” 22 For a more complete discussion, youtube.com/watch?v=x-S-eeInJVk& Harv. Bus. Rev. 56, 59 (July 2003). see Donald R. Philbin Jr., “The One NR=1; see also an early and abbreviated 14 “Blinking on the Bench: How Minute Manager Prepares for Me- version of this article in the online ver- Judges Decide Cases,” 93 Cornell L. Rev. diation: A Multidisciplinary Approach to sion of the Harvard Negotiation Law 1, 2007; see generally, Richard A. Posner, Negotiation Preparation,” 13 Harv. Review, available at http://www.hnlr. How Judges Think (Harvard Univ. Press Negotiation L. Rev. 249 (2008). org/?p=471. 2008). 23 Korobkin, supra n. 5, at 309. 2 http://sports.espn.go.com/mlb/ 15 U.S. Supreme Court Chief Justice 24 Sander’s work continues to ani- playoffs/2009/columns/story?colum- John G. Roberts Jr. said, “Judges are like mate modern dispute resolution. nist=caple_jim&id=4581598. umpires ... Umpires don’t make the 25 Sander & Goldberg, supra n. 19, at 3 Sports Business Daily 10/22/2009. rules; they apply them.” Bruce Weber, 66; Peter Robinson, “Contending with 4 Daniel Kahneman & Amos Tversky, “The Deciders: Umpires v. Judges,” Wolves in Sheep’s Clothing: A Cau- “Conflict Resolution: A Cognitive Per- N.Y.T. (July 11, 2009)(available at www. tiously Cooperative Approach to Media- spective,” in Barriers to Conflict Resolution nytimes.com/2009/07/12/weekinre- tion Advocacy,” 50 Baylor L. Rev. 963, 45, 48 (Kenneth Arrow et al., eds. 1995). view/12weber.html?_r=1). 964 (1998) (“Mediation is facilitated ne- 5 Russell Korobkin, “Psychological 16 Randall L. Kiser, et al., “Let’s Not gotiation.”). Impediments to Mediation Success: Make A Deal: An Empirical Study of 26 Leonard L. Riskin, “Understand- Theory and Practice,” 21 Ohio St. J. on Decision-Making in Unsuccessful Settle- ing Mediators’ Orientations, Strategies, Disp. Resol. 281, 284-85 (2006). ment Negotiations,” 5 J, Empirical Leg. and Techniques: A Grid for the Per- 6 Final Offer Arbitration or Baseball Stud. 551-91 (Sept. 2008), (for empirical plexed,” 1 Harv. Negotiation L. Rev. 7 Arbitration is “a form of arbitration in research on negotiating errors in litigat- (1996). which each party submits a proposed ed cases). 27 Mediator evaluations come in monetary award to the arbitrator. At the 17 The multi-door courthouse in some three primary forms: (1) “gestalt evalua- conclusion of the hearing, the arbitrator jurisdictions allows some criminal mat- tion” (overall reaction without detailed is required to select one of the parties’ ters to be mediated. feedback); (2) detailed feedback with or proposed awards, without modification.” 18 This conference commemorated without “gestalt”; and (3) decision ana- Stephanie Smith & Janet Martinez, “An the 70th anniversary of Roscoe Pound’s lytic approach. Marjorie C. Aaron, “The Analytic Framework for Dispute System famous lecture, “The Causes of Popular Value of Decision Analysis in Mediation Design,” 14 Harv. Negot. L. Rev. 123, Dissatisfaction with the Administration Practice,” 11 Negot. J. 123, 123 (1995); 165 (2009). of Justice.” see also Laurence D. Connor, “How to 7 Kahneman & Tversky, supra n. 4, at 19 The quoted phrase is derived from Combine Facilitation with Evaluation,” 47. Frank E.A. Sander & Stephen B. Gold- 14 Alternatives to High Cost of Litig. 15 8 Id. at 55. berg, “Fitting the Forum to the Fuss: (1996); Dwight Golann, “Benefits and 9 Id. at 46. Guide to Selecting an ADR Procedure,” Dangers of Mediation Evaluation,” and 10 Amos Bronson Alcott (1799-1888). 10 Negotiation J. 49, 67 n. 7 (1994). “Planning for Mediation Evaluation,” 15 11 Dan Orr & Chris Guthrie, “An- 20 Only certain arbitrator errors can Alternatives to High Cost of Litig. (1997). choring, Information, Expertise, and be challenged in court. The alleged 28 Dwight Golann said, “Absent an Negotiation: New Insights from Meta- errors must be authorized by state or analytical structure for understanding a Analysis,” 21 Ohio St. J. on Disp. Resol. federal law or common law. Mediators complex case, the parties have no mech- 597, 607 n. 21 (2006). are facilitators, and unlike arbitrators, do anism with which to consider how the 12 Lee Ross, “Reactive Devaluation in not decide the parties’ disputes. So they mediator’s feedback on individual issues, Negotiation and Conflict Resolution,” are rarely the subject of litigation. if accepted, will affect their case’s value.” in Barriers to Conflict Resolution, supra n. 21 Some common definitions of these Id. at 125. 4, at 26, 30. dispute resolution procedures are con-

DISPUTE RESOLUTION JOURNAL 7