The Morality of Bargaining: Identity versus Interests in Negotiations with Evil Open Access

In his latest and most engaging book to date, Robert Mnookin, the Samuel Williston Professor of Law at the Harvard Law School and Chair of the Steering Committee of the Program on Negotiation, takes up two compelling questions. The first and most interesting is this: when is it morally acceptable to refuse to negotiate with “devils,” that is, certifiably “evil” people such as Adolf Hitler, terrorists, racists, and the like? Second, in those cases in which such negotiations are morally permitted — or even required — what are the best practices for conducting them?

Regarding the first of these two questions, the book breaks provocative new ground in the negotiation literature. (Editor's note: For a different perspective on similar questions, see Carrie Menkel‐Meadow's review of Avishai Margalit's book On Compromise and Rotten Compromise, elsewhere in this issue.) Mnookin's rich, real‐world examples provide an excellent foundation for his discussion, and I hope this essay will provide a helpful summary, useful critique, and theoretical analysis of what the book has to offer. The question of negotiating with evil may be far from most people's everyday experience, but it is a pressing, important issue for those who serve as diplomats; political leaders; prosecutors in cases of terrorism, serial killing, or child abuse; hostage negotiators; and others fighting on the front lines against those who mean to do us harm. Moreover, we have a duty as citizens to understand the moral choices that such negotiators are making in these tough cases when they claim to be acting in our name.

Mnookin answers his second question regarding best practices by referring us to familiar principles of interest‐based bargaining and techniques for overcoming well‐established emotional biases. While this aspect of the book is useful for the general reader, it is old news to the readers of Negotiation Journal. Hence, my review will focus mainly on the issue of refusing, on moral grounds, to negotiate with those we consider evil.

In this introductory section, I will briefly summarize Bargaining with the Devil: When to Negotiate, When to Fight and identify what I admired most in the book. In the next section, I will lay out Mnookin's moral framework for deciding when it is “wise,” as he puts it, to negotiate with evil people and when we are permitted (or required) to walk away from such encounters. Third, I will summarize four historical case studies Mnookin gives us to examine how his decision framework fits these cases. In the final section, I will argue that Mnookin's approach, while a useful start on this project, is, as he readily acknowledges, incomplete. Utilizing a basic decision‐making taxonomy provided by James G. March, the Jack Steele Parker Professor of International Management Emeritus at Stanford Business School and the author of numerous works on decision theory, I will advance a complementary framework for thinking about these difficult decisions — a framework I will call “identity‐based bargaining.”

Before getting started, however, I want to alert readers who teach negotiation to the opportunity this book presents. The historical examples that Mnookin recounts are tailor‐made to spark genuine, passionate classroom debates that will force students at every level — from undergraduates to executives — to confront their assumptions about the deeper purposes of negotiation. Most negotiation courses currently have a session or two on bargaining ethics — with the usual emphasis on lying, bluffing, falsifying priorities, and other, more subtle forms of deception. But Mnookin's new book allows teachers to examine the moral assumptions underlying even the most creative forms of negotiation.

Indeed, the very word “collaboration,” which we normally define as a positive, problem‐solving approach to bargaining, takes on a sinister meaning when we use it in the context of bargaining with “devils” such as the Gestapo and the Soviet KGB. One of Mnookin's case studies involves Rudolf Kasztner, a Hungarian Jew who tirelessly negotiated with Nazis to save Jewish lives near the end of World War II. His “cash for lives” deal‐ making put him on morally ambiguous ground, however, and he spent much of his life after the war trying, without complete success, to clear his name from charges by fellow Jews that he was a “collaborator” rather than a hero.

The examples Mnookin presents offer a welcome and long‐overdue opportunity to enrich the standard negotiation curriculum. For too long, negotiation pedagogy has implicitly assumed that the decision to negotiate is a function only of interests and alternatives. But that decision is about more than just consequences — it is a form of self‐expression that says something about who we are, who we are willing to be seen as, and, ultimately, who we are willing to become. Mnookin's Bargaining with the Devil opens the door to that discussion.

As I hinted above, the basic organization of Bargaining with the Devil reveals that it is actually two quite different books that have been bound together by a single metaphor. The table of contents tells us that the book has four parts — Understanding the Challenge, Global Devils, Business Devils, and Family Devils. But there are really just three functional sections. Part one (the introduction, chapters one, two, and part of chapter three) introduces and frames both the moral and psychological issues the book will explore. Part two (part of chapter three and chapters four, five, and six) presents a set of case studies from twentieth‐century history that allows Mnookin to normatively assess the wisdom of four famous decisions to negotiate (or not) with evil people. These are: Soviet dissident Anatoli (Natan) Sharansky's refusal to negotiate with the Soviet KGB Rudolf Kasztner's troubled negotiations with Nazi Adolf Eichmann that led to the escape of 1,684 Jews from Nazi‐occupied Hungary on the so‐called “Kasztner Train,” Winston Churchill's decision to forego negotiations with Hitler in 1940, and Nelson Mandela's famous decision in 1985 to initiate negotiations from his prison cell with the white South African regime — a decision that ended in the creation of a new political system for that country.

Part three of the book (chapters seven, eight, nine, and ten) presents a final set of four examples taken from Mnookin's own world of professional practice and experience. To his credit, he presents both successes and failures: his famous mid‐1980s arbitration/mediation with Jack Jones of a software dispute between IBM and Fujitsu, a challenging and ultimately unsuccessful interest‐based bargaining training program Mnookin delivered to the San Francisco Symphony Orchestra in the late 1990s, a failed divorce negotiation Mnookin witnessed at close quarters, and a family dispute over a Cape Cod summer home that Mnookin helped resolve through mediation.

The metaphor of “bargaining with the devil” threads its way throughout the well‐told stories in both parts two and three. But by the end of the book, when we are sitting next to Mnookin as he skillfully guides three adult siblings to a successful partition of their family's summer house, the concept of “the Devil” has lost its punch. I found myself muttering more than once, “Stop that metaphor!” as devilish labels that applied appropriately to KGB interrogators in part two were used to characterize unhappy siblings and labor negotiators in part three.

This may simply be the price of my knowing more about interest‐based bargaining than the average reader is likely to know. It is certainly true that people in disputes often demonize each other and that talented mediators and advisors can help them overcome their emotional and attribution biases to find mutually satisfactory solutions. But as good as the personal narratives are in part three of Bargaining with the Devil, there are other and better books on win–win negotiating, including the classic Getting to Yes (Fisher, Ury, and Patton 1991) and Mnookin's own Beyond Winning (Mnookin, Peppet, and Tulumello 2004). Mistrust (the subject of part three) is one thing; evil (the main subject of part two) is quite another. Morality has much more to do with the latter than the former — and it is to morality that we now turn.

Bargaining with the Devil had its genesis in an argument in 2001 between Mnookin and his distinguished colleague (and founder in 1979 of the Harvard Negotiation Project), Professor Roger Fisher. For the description below, I checked the Harvard Program on Negotiation's (PON) website (http://www.pon.harvard.edu), which has posted written versions of both men's remarks.

Two months after the attacks of September 11, 2001, these two leaders of the negotiation field conducted a public debate at Harvard on whether the United States should have accepted a September 19 offer made by the Taliban to negotiate on issues of mutual interest. Fisher said “yes” and Mnookin said “no.”Bargaining with the Devil is Mnookin's extended meditation for the position he took in that debate.

It is worth detailing both the context and the gist of this debate. In the immediate wake of the September 11, 2001 attacks on the World Trade Center and the Pentagon, the administration of President George W. Bush demanded that the Taliban rulers of Afghanistan hand over Osama bin Laden, the mastermind of the attacks, and close all al‐Qaeda terrorist training camps. According to an address given in Kabul by Taliban leader Mullah Mohammed Omar to clerics on September 19, 2001, Omar asked the United States to provide proof that bin Laden was implicated in the attacks, offered to try him in Afghanistan if such proof was forthcoming, and stated, “If the American government has some problems with the Islamic Emirate of Afghanistan, they should be solved through negotiations.”

On September 21, 2001, Bush responded by issuing a formal ultimatum in a speech before Congress, calling on the Taliban to turn over all terrorists residing in their country. “These demands are not open to negotiation or discussion,” Mr. Bush said. “The Taliban must act and act immediately. They will hand over the terrorists, or they will share in their fate. . . . From this day forward, any nation that continues to harbor or support terrorism will be regarded by the United States as a hostile regime.” The Taliban steadfastly refused to hand over bin Laden. War in Afghanistan ensued — though it was swiftly overtaken by events in Iraq — and continues, sadly, nearly a decade later.

Roughly two months after these events, on November 13, 2001, sponsored PON a debate in which Mnookin and Fisher argued whether the United States had been wrong to refuse to talk with the Taliban in September. Events were still moving fast. The Taliban had just withdrawn its forces from Kabul and Northern Alliance military units friendlier to the United States had occupied the Afghan capital. Bush's strategy was looking good. Nevertheless, Roger Fisher's position on November 13 was categorical: one should always talk if talk is possible. “The notion that negotiation is risky,” said Fisher, “comes from the idea that negotiation is about making concessions. This is wrong. Negotiation is talking and listening, understanding what the other side wants and having a chance to persuade them.” In effect, Fisher was saying that there is a moral duty to talk— if only through a back channel or a neutral intermediary — unless such talk is impossible. With the Taliban actually inviting dialogue in September 2001, Fisher was unwilling to say, based on his extensive international negotiating experience, that such talks would have produced nothing whatever. He therefore objected to Bush's refusal to negotiate.

Mnookin's response was a point‐by‐point, cost–benefit analysis of what the United States might have achieved through any such negotiations. Based on his analysis, Mnookin concluded that the costs and risks of negotiating far outweighed any possible benefits. Mnookin asked five questions (questions that, in his book, he places in the hands of a fanciful “Mr. Spock” of Star Trek fame, who, as a Vulcan, can be counted on to look at all situations in a strictly “rational” way):

1. What were the interests at stake?

2. What were the alternatives to negotiation?

3. Were there likely outcomes that would meet the interests of both parties and would any agreements to such outcomes actually be carried out?

4. What were the costs to the United States of choosing to negotiate?

5. Was the U.S. alternative of using military force morally justifiable?

The crucial parts of this analysis were Mnookin's conclusions to questions three, four, and five. He felt that, even if options were identified, the Taliban had proven entirely unreliable and would be neither willing nor able to implement them; that the United States would lose enormous credibility around the world by opening discussions with Taliban leaders — undermining our ability to assemble a coalition to fight terrorism; and, finally, that bin Laden's declaration of war on the United States, which Mnookin imputed to the Taliban as Osama bin Laden's allies, more than justified military action as an alternative to negotiations.

Although he did not do so at his November 2001 debate with Fisher, Mnookin states in Bargaining with the Devil that he considered the Taliban in September 2001 to be “evil” and that his “moral intuitions” as well as his pragmatic analysis told him that negotiations with them would be wrong. And it is at this point that the book begins breaking ground not covered in the earlier exchange with Fisher.

To advance his new thesis, Mnookin provides us with a definition of “evil.” That passage is worth quoting at length because it makes clear that Mnookin wishes to challenge those in the dispute resolution profession who see labels such as “good” and “evil” as mere perceptions without objective content:

He further states that not everyone who commits an evil act is categorically “evil.” Instead, he writes, “I would call someone an evil person only if by disposition he or she repeatedly commits evil acts” (p. 15). These definitions may be a bit lawyerly for some tastes. What, after all, is “grievous harm”? What is a “compelling justification”? Does “repeatedly” include just two acts? But they are more workable as decision aids than common dictionary definitions of the word evil, such as “morally reprehensible,” which have a tendency toward circularity.

In addition, Mnookin is surely correct that he is taking a controversial position that many who have dedicated their lives to peacemaking will vigorously oppose. Indeed, some would argue that the word “evil” often causes evil acts. Marshall Rosenberg, founder of the Center for Nonviolent Communication in Arizona, for example, would say that labeling others as “evil” allows us to switch off our natural feelings of empathy toward them and act in violent ways that we would otherwise find humanly impossible. Ironically, given that Mnookin uses several examples in which the Nazis are the evil ones, Rosenberg has cited the ability of the Nazi regime to mobilize ordinary Germans to destroy the Jews as the classic example of this phenomenon (Rosenberg 2003, 2006).¹

Mnookin holds otherwise and, with his definition of evil in hand, he concludes in Bargaining with the Devil that his decision to side with Bush regarding the Taliban “was a relatively easy one” because (1) the Taliban were implicated in the attacks of September 11 and were “evil”; (2) his strong moral impulse was against negotiating with such an enemy; and (3) his cost–benefit analysis, revisited from his November 2001 debate with Roger Fisher, affirmed that negotiations would have been a bad bet anyway (pp. 7–8).

The harder cases, Mnookin asserts, arise when these two decision factors — a pragmatic concern with interests and a principled abhorrence of evil — collide. How does one decide whether or not to negotiate when the moral heart says “no” to negotiation while the rational head says “yes”? His answer, which he gives near the end of the book (pp. 262–265), is to outline four decision‐making steps:

1. “Systematically compare the expected costs and benefits.” This involves running through the five questions that Mnookin used to examine the Taliban problem at the PON debate and that he places in the hands of the rational “Mr. Spock.”

2. “Get advice from others in evaluating the alternatives: don't do the analysis alone.” Because none of us really are like Mr. Spock, we need a variety of people involved in our analysis to bring different perceptions, biases, emotional attachments, and intuitive judgments into our decision process.

3. “Have a presumption in favor of negotiation, but make it rebuttable.” This is a form of decision rule to use once you have performed the cost–benefit analysis required by steps one and two. In case of a tie between pragmatism and principle, pragmatism wins. But the theoretical possibility exists that a principled stand against negotiation can prevail over that presumption in some cases.

4. “When deciding on behalf of others, don't allow your own moral intuitions to override a pragmatic assessment.” This second decision rule burdens agents or representatives — and this includes all political and business leaders — with a special duty to avoid acting on principle whenever the cost–benefit analysis comes out in favor of negotiation. An exception to this rule may exist if the representative has persuaded everyone who will share the costs of a refusal to negotiate to do so freely and willingly (p. 49). As we shall see later on, Mnookin does not explain how a political or business leader with a large, diffused constituency might qualify for this exception. So the practical application of it seems to be limited to cases of individuals serving as agents for small groups who are able to persuade these groups to go along with their purely principled stand against negotiation.

Underlying this set of decision rules is a set of three explicit assumptions about the way people process information in stressful conflict situations. In general, these assumptions prompt Mnookin to be deeply suspicious of the “moral intuitions” mentioned above in Rule Four. Many of these background understandings about decision processing are also the everyday working hypotheses of dispute resolution professionals, and they are therefore worth reciting before going on to the examples Mnookin explores.

The first assumption is that people have two “fundamentally different ways” (p. 16) of perceiving reality and making judgments. This is the familiar “System 1” (intuitive/emotion‐laden) and “System 2” (deliberative/reason‐based) dual‐processing model from cognitive psychology that is now the standard academic account of human mental and emotional functioning in making decisions. As Daniel Kahneman stated in his 2002 Nobel Lecture: “The operations of System 1 are fast, automatic, effortless, associative, and difficult to control or modify. The operations of System 2 are slower, serial, effortful, and deliberatively controlled; they are also relatively flexible and potentially rule‐governed” (2002: 2).² Mnookin notes that System 1 processing is “quirky and selective about the data it receives, which can lead to mistaken conclusions and poor decision‐making” (p. 17).

The second assumption is that the operation of intuition (System 1) is subject to an especially reliable pattern of biased and distorted responses in human conflict situations. People in conflict are quick to resort to tribalism, demonization, and dehumanization of their opponent, self‐righteousness, zero‐sum/win–lose thinking, fight/flight responses, and literal or figurative “calls to battle” (p. 18). Mnookin calls these “negative traps.” He also posits an associated set of “positive traps” that assume people have equally intuitive tendencies in conflict to make peace, appease others, forgive, and accept fault or blame. But these latter inclinations play virtually no role in the book other than to stand as opposite to the negative traps. If anything, they seem to be Mnookin's way of explaining how peacemakers such as Marshall Rosenberg can be so intuitively hostile to the idea that some people are evil.

The third assumption is that moral judgments, like “subjective impressions of all kinds . . . arise from the intuitive side of the brain: they are gut feelings that are instinctively reached and deeply felt” (p. 35). Mnookin cites recent magnetic resonance imaging studies that demonstrate the involvement of emotions in situations involving moral conflict and relies on University of Virginia psychologist Jonathan Haidt and philosopher David Hume for the view that morality has its origins in sentiment (p. 276, note 3). Thus, at least for the purposes of this book, Mnookin implies that a deliberative, purely reasoned approach to moral decision making in conflict situations is either impossible or so rare that it is not worth exploring as a special case. No doubt there are some moral philosophers who would dispute this claim (see Margalit 2010 and Menkel‐Meadow 2010), but Mnookin is not writing a philosophy book. Rather, as he freely admits, he is writing a “how to” book for making these tough decisions in practice. His model of intuitive moral reasoning need only be the “normal” case for this assumption to be justified — and I found it easy to agree with him.

These three assumptions go far in explaining the high hurdles that Mnookin places before any decision refusing an opportunity to negotiate with an evil person. Basically, he does not trust moral intuitions. More often than not, Mnookin suggests, they are little more than hot‐headed, unreliable impulses toward total war with a perceived enemy that are subject to immediate justification based on tribal self‐interest, inflated claims of moral self‐righteousness, and emotional appeals that call us to battle. We must be very, very careful before allowing such impulses to overrule the objective, cost–benefit analysis that our inner Mr. Spock can give us regarding our true interests.

With this background on Mnookin's decision framework in mind, we are now ready to examine the four historical examples of bargaining with evil he uses to explore his thesis. He gives us two examples of people refusing to negotiate with evil and two examples of people negotiating more or less successfully with evil. And after telling each of their stories, he comes to a judgment about whether the person in question exercised sound judgment and acted with sufficient justification (or decided “wisely,” as Mnookin puts it). It is not giving too much away to tell you that he decides in all four cases that the decision maker did so, and it is hard to argue with any of his conclusions. But I am less sure about the way Mnookin applies his framework consistently in reaching his results. And that uncertainty is what will prompt me to suggest a different, and complementary, framework for analyzing these cases at the end of this essay.

For clarity in the review below, I have reshuffled the order in which he presents these cases so that we can better examine the two “no” decisions and the two “yes” decisions as pairs. Hard choices can ruin your day — perhaps even your life. But they can make for both good reading and dynamic classroom discussions. We will start with the two “just say no” decisions.

Anatoli (Natan) Sharansky was born in 1948 in the Ukraine. He grew up as a nonobservant Jew, largely unaware of the culture, language, and history of his people. As one of the brightest students at the Moscow Institute of Physics and Technology, however, he came under the influence of Andrei Sakarov and the Committee for Human Rights as well as the Soviet Zionist movement. These groups “sought to pressure the Soviet regime to grant Jews the right to leave the USSR for Israel,” and it was through his affiliation with them that Sharansky came to see himself in a new light, developing, according to Mnookin, “a strong Jewish identity” (p. 38). Sharansky later wrote, “[F]or the first time in my life, I was no longer afraid to say what I really believed — about my fellow citizens, the country I lived in, and the values I adhered to. At the age of twenty‐five I finally learned what a joy it was to be free” (p. 38).

In 1973, he met Natasha (later Avital) Stieglitz, and a year later he joined her in applying for permission to immigrate to Israel. They were married on July 4, 1974, and the next day, in the hopes that Sharansky would soon follow, she left for Israel. His application to immigrate was denied, however, and by 1977, at the age of twenty‐nine, he was a leading spokesperson for the Soviet Zionist movement. The KGB (the Soviet national security agency) then seized him, charged him with the trumped‐up crime of selling state secrets to the U.S. Central Intelligence Agency, and threw him into prison.

Almost immediately after his arrest, the KGB offered Sharansky a deal: if he would confess and condemn the Soviet Zionist movement, then the KGB would allow Sharansky to join his wife in Israel after only a short term in prison. It was also understood that he could repudiate his confession as soon as he was out of the country. Sharansky refused to consider this and turned the various attempts to negotiate with him into a game (he was a chess master and game theorist) in which his goal was to cooperate with as few KGB demands as possible.

Why did Sharansky decide to stand on principle and remain in prison for nine miserable years? He later gave three reasons: (1) he did not want to betray his friends in the movement; (2) he thought that “collaborating” with the KGB would compromise his movement's “only asset”— its “strong moral position” that gave people in the world a reason to concern themselves with Soviet Jews; and (3) he knew that if he recanted, it would make it easier for the regime to “initiate new repressions and another round of arrests” (pp. 44–45).

Sharansky's case is by far the most psychologically complex example included in Bargaining with the Devil because he discussed his own emotional and cognitive processes at length in the wake of his release. The rational political considerations that led him to resist negotiations at the outset of his ordeal began to fade into the background as the KGB's relentless attempts to break him continued. In retelling this story, Mnookin spends a good deal of time recounting Sharansky's fascinating description of the mental tricks and emotional games he played to maintain his “no negotiation” position. Sharansky writes that “I felt like a chess player facing a much weaker opponent . . . I knew all their moves in advance: their threats and warnings, their attempts at blackmail, their flattery and their promises” (p. 40).

More fundamentally, Sharansky relied on an “intuitive, automatic feeling” and desire “to continue being free” and “to enjoy my inner freedom in prison” as “the basis of resistance” (p. 46). He noted, “As a religious, rational person, I was relying on my instincts, but as a scientist I had to rationalize these instincts. I had to explain to myself, rationally, why I should not cooperate with them.” He thus trained himself to objectify his captors (he treated them “like the weather”) so that “nothing [they did] could dehumanize me. I could only humiliate myself — by doing something I might later be ashamed of” (p. 47). In short, faced with KGB devils as daily interrogators, Sharansky utilized every “negative trap” on Mnookin's list to demonize his enemies. As a result, he successfully resisted all of their attempts to coerce him into a compromise.

His wife, meanwhile, spent these same years organizing a relentless international campaign to secure his release. Finally, in 1986, President Ronald Reagan responded to the movement to free Sharansky and offered to exchange two Soviet‐era spies for him. The Soviets agreed and Sharansky at last immigrated to Israel. Defiant to the end, Sharansky literally walked zigzag across the Soviet border after his KGB handler warned him to “go straight to it [the border] and don't make any turns” (p. 37).

Was Sharansky morally permitted under these conditions to turn his back on negotiation? Mnookin's answer is “yes.” As he explains it, “one could quibble with whether his [Sharansky's] decision was rational or not. Cost–benefit analysis would hardly require him to risk his life and liberty for the Soviet Jewish cause. His choice to resist was courageous, even heroic. And in my view, it was also wise. I say this for two reasons. First, he did not simply rely on his moral intuitions. He understood the risks. Second, he alone bore the costs of resistance” (an endnote mentions that Sharansky's wife fully supported his stand) (p. 49 and note 9).

Let's take a closer look, however, at how Sharansky's case holds up to the more elaborate analysis Mnookin hopes readers will use in their own disputes with devils. First, the KGB certainly qualifies as “evil” under Mnookin's definition. The KGB sought, repeatedly, to intentionally inflict grievous harm on Sharansky without compelling justification. This opens the moral possibility of a refusal to negotiate.

Next, we consider Mnookin's suggested four‐step decision process. On the first step — the need for a thorough cost–benefit analysis — the Sharansky case demonstrates the complicated, intertwined relationship between System 1 (intuitive) and System 2 (deliberative) mental processes. On the “benefit” side, Mnookin tells us that Mr. Spock could approve the three political reasons Sharansky gave for his initial refusal to collaborate with his enemy. Avoiding the betrayal of friends in a political movement, protecting that movement's only asset (its moral high ground), and fearing the possibility of unleashing further arrests are all solid, rational reasons to resist negotiation.

But Mnookin wonders why Sharansky “so completely ignore[d] his other interests” on the cost side. What about his desires to avoid death, regain his freedom, join his wife in Israel, and work from the outside to promote his movement (pp. 45–46), all of which could be furthered by negotiation? Mnookin speculates that Sharansky eliminated them from consideration by a sheer act of will, fearing they might undermine his principles (p. 47). Moreover, at a deeper level, Mnookin tells us that Sharansky was not thinking about his decision in cost–benefit terms at all, but rather acted on emotions related to “self‐respect, moral purpose, and identity” (p. 47).

As I shall argue later, I think the Sharansky case reveals an essential limitation in Mnookin's cost–benefit‐based framework. By relegating considerations such as “self‐respect, moral purpose, and identity” entirely to the nonrational, System 1 realm, Mnookin misses a chance to consider how people think — in deliberative ways — about such matters. But taking Mnookin's framework on its own terms, I agree with him that Sharansky was sufficiently aware of genuine cost–benefit considerations (“he understood the risks”) to pass muster under step one of the four‐step test.

Mnookin's second rule requires consultation with others. I am sure that, before his arrest, Sharansky took full advantage of chances to discuss with his fellow dissidents the question of how they should resist the KGB if they were arrested. Many in the movement had been harassed, some had been seized, and some had taken the KGB deal, after all. Of course, Sharansky was unable to continue these consultations after he was arrested, but I think we can safely say that Sharansky followed Rule Two.

Third, what about Mnookin's “rebuttable presumption” in favor of negotiation? On this point, Mnookin and I disagree. Mnookin suggests that Sharansky faced a conflict between System 2 pragmatism and System 1 principle and that he allowed his moral instincts to override a more rational cost–benefit analysis. As Mnookin puts it, Sharansky “used the intuitive, feeling part of his brain to decide what his goal should be,” and then “rigged the [cost–benefit] analysis,” manipulating it “so it would lead him to the ‘right’ conclusion” (p. 47). In the end, as noted above, Mnookin approves Sharansky's refusal to negotiate only because, having thought the matter through, he and his wife were the sole people burdened with the cost of his emotionally motivated decision.

I have several concerns with this conclusion. First, it is not clear to me that the sensible political considerations animating Sharansky's refusal to negotiate in the first place (based on the needs of his Soviet Zionist movement) lapsed sometime during his incarceration. Perhaps they did, but we are not given any evidence of this. All we are told is that Sharansky, locked in his isolated cell, stopped thinking in strictly political terms, and transformed the conflict into a personal test of wills against the Soviet regime. Moreover, in contrast to Mnookin, I am inclined to believe that Sharansky continued to weigh, at some level, both his personal and political interests in his freedom. And in that weighing, he could well have concluded that his political movement would be better served by his remaining defiant. Many political leaders, including Nelson Mandela, discussed by Mnookin later in this book, have made similar calculations, and we respect them for doing so without questioning their rationality.

In Sharansky's case, these underlying political considerations are somewhat obscured by the elaborate psychological defense system he erected to cope with the social isolation he suffered. But these mental tricks and games do not cause me to question his overall commitment to reason. I say this because the relationship between System 1 and System 2 thinking is too complex to make a clean, “either/or” distinction possible. Indeed, Sharansky himself exemplifies this complex relationship in the way he writes about it. Look again at one of Sharansky's quotes recounted above: “As a religious, rational person,” he tells us, “I was relying on my instincts, but as a scientist I had to rationalize these instincts.” Can humans have instincts that are rational and that manifest differently in matters of religion, science, and cost–benefit analysis? I think so, but Mnookin's association of System 2 rationality exclusively with pragmatism does not readily allow for such possibilities.

And what of Mnookin's fourth rule requiring a political leader to avoid letting his or her moral intuitions against negotiation overwhelm a more pragmatic judgment in favor of talking? As noted above, Mnookin treats Sharansky's refusal to negotiate as a personal one affecting only two people, so he does not reach this final question. If we consider Sharansky as a leader of the Soviet Zionist movement, however, I think it is safe to say that his steadfastness over nine hard years paid huge pragmatic dividends. It inspired a global political campaign for his release that greatly strengthened the Zionist movement and it contributed to the overall erosion of Soviet credibility during a crucial period of history. Perhaps, had Sharansky engaged his jailers in genuine dialogue four or five years into his ordeal (rather than relying exclusively on his wife's efforts to trade for his freedom), he might have found a way to negotiate his way out of the Soviet Union on acceptable terms. Did he miss some chances to find a workable solution? It is impossible to know.

In the end, I commend Mnookin for choosing the Sharansky case as his first example because it vividly highlights the difficulties of post hoc evaluation in morally complex situations. Cost–benefit analysis — even in the hands of a Mr. Spock — is based on each observer's perceptions and judgment. Where I see a balance favoring political benefits, Mnookin sees a balance tipping toward personal costs. Moreover, it is impossible to reconstruct the actual motivations of people under extraordinary stress. Even Sharansky himself seems unsure, in retrospect, just how reason, instinct, and emotion combined to sustain him against his KGB tormentors. All we know for sure is that he survived without compromising his principles.

Finally, the case illustrates how exceedingly narrow the doorway is that Mnookin says we must pass through if we want to walk away from a negotiation on nonutilitarian grounds. Either we must be acting almost entirely on our own behalf (which is how Mnookin views Sharansky's case) or, if we are in a leadership position, we must gain the consent of all the people who will share the cost of resistance with us. Mnookin gives the example of the CEO of a small private company who, if he wanted to walk away from a negotiation with a business partner he had come to see as evil, would have to convince the board of directors and all the shareholders to share his moral convictions (p. 49). As I noted earlier, this is a difficult standard for any leader of a large organization or movement to meet and thus places the moral burden in these decisions entirely on cost–benefit considerations.

On May 28, 1940, after three days of intense discussions with the two top leaders of his war cabinet — Edward F. L. Wood, known as Lord Halifax, and former Prime Minister Neville Chamberlain — British Prime Minister Winston Churchill stepped outside his office and addressed a large group of his cabinet ministers sitting around a very large conference table. “I have thought carefully in these last days whether it is part of my duty to consider entertaining negotiations with That Man [Hitler],” he declared. He then recounted what he expected from such talks: “The Germans would demand our fleet . . . our naval bases and much else. We should become a slave state . . .” In the face of this possibility, Churchill declared that no negotiations would take place. “We shall go on and we shall fight it out . . . and if at last the long story [of Great Britain] is to end, it were better that should end, not through surrender, but only when we are rolling senseless on the ground.” His ministers cheered and, as Churchill later described it, “Quite a few seemed to jump up from the table and came running to my chair, shouting and patting me on the back” (p. 101). Thus was it decided that the war would go on without negotiations — even as British troops in Europe scrambled toward the French coastal town of Dunkirk, where one of the great rescue efforts of the war would soon occur.

Of all the stories Mnookin tells in Bargaining with the Devil, this one has the most detailed texture and holds the greatest historical interest. The reason is simple: the true account of the deliberations within the war cabinet was suppressed for decades — with both Winston Churchill himself and the chief protagonist who pressed for negotiations, Lord Halifax, denying during their lifetimes that these discussions ever took place. The truth came out only in January 1990, long after both men had died, when the secret minutes of the war cabinet were released. The back‐and‐forth of this debate and the collisions between the articulate, urbane Lord Halifax, who favored on pragmatic grounds the opening of a back channel through Italy to explore negotiations, and the emotional, stubborn Churchill, who sought any and all reasons to reject this idea, are recounted in full dramatic detail by Mnookin.

Both Halifax and Chamberlain were “old hands” at dealing with Hitler and had done so repeatedly during the 1930s — with depressingly consistent results as Hitler extracted concession after concession and violated agreement after agreement. Nevertheless, with war well underway and France about to fall, Halifax held a key meeting on May 25 with the Italian ambassador in London. Halifax was under strict instructions to signal nothing to the Italians, but he somehow came out of the meeting with a possible negotiation opening: would England be willing to join Italy and other European powers (i.e., Hitler) to discuss a “just and enduring European settlement” (p. 91)? This was the question Halifax presented to Churchill and Chamberlain the next day and that, over the following three days, became the subject of intense debate.

Mnookin takes us step by step through these discussions, rating the debates much like a scorekeeper at a prize fight. On cost–benefit grounds, Mnookin tells us that Halifax had the stronger position for at least the first couple of days. Faced with the possible destruction of his homeland by bombing and with the almost certain destruction of the British army at Dunkirk, Churchill was hard pressed to offer reasons why negotiations should not at least be explored. Writes Mnookin, “Frankly, I think Halifax had Churchill on the ropes. If the analysis were to stop here, one might well conclude that wisdom would obligate Churchill to explore the possibility of negotiation by using the Italians as intermediaries” (p. 96).

What then changed the tide of the debate? First, events shifted on the ground. On May 27, the war cabinet learned that Italy's Fascist dictator Benito Mussolini would soon be declaring war on France. This undermined the idea that Italy could act as a neutral broker between Germany and Great Britain. But the French now pressed Churchill to consider giving Italy concrete territorial concessions to buy off Italy's declaration of war. This was an unappealing stance and Churchill rejected it outright.

Second, a new argument was advanced against negotiations by a new player at the meeting in which the French plea was considered. Reflecting political concerns very similar to those expressed by Sharansky early in his decision‐making process, Archibald Sinclair, the air minister and head of the Liberal Party, contended that “being in a tight corner, any weakness on our part [shown by putting territorial concessions on the table] would encourage the Germans and Italians, and would tend to undermine morale both in this country and in the [British] Dominions.” Other members of the war cabinet picked up on this theme, and Churchill ran with it. Negotiations under these conditions, Churchill flatly stated, “would ruin the integrity of our fighting position in this country” (p. 98). Besides, he went on, “if the worse came to worst, it would not be a bad thing for this country to go down fighting . . .” (p. 99).

Mnookin is critical of Churchill at this point. He describes Churchill's “go down fighting” attitude as “crazy” and argues that a leader has no right to “sacrifice his countrymen” on the basis of his own moral instincts (p. 99). Halifax was despondent the evening after this meeting, confiding to his diary that “it does drive me to despair when [Churchill] works himself up into a passion of emotion when he ought to make his brain think and reason” (p. 100).

In the end, after a morning meeting on May 28 in which Halifax and Churchill clashed once again, Churchill adjourned the meeting until that evening. He had made up his mind. And shortly thereafter, he spoke to twenty‐five members of the greater cabinet, giving the rousing speech that began this section. “It was idle to think,” said Churchill to them, “that if we try to make peace now we should get better terms from Germany than if we went on and fought it out” (p. 101).

Mnookin approves of Churchill's refusal to negotiate — but not as enthusiastically as one might expect given that he acknowledges Churchill as one of his “heroes” (p, 105). There are two basic problems. First (and I am reading between the lines here), the way Mnookin tells this story suggests that Churchill only went through the motions of a cost–benefit analysis, using the debates more to search for justifications that would bolster his instinctive view than to engage in genuinely open‐minded, rational thought. Mnookin says he was “powerfully struck” both by how much Churchill relied on his intuitive side and by the “respect” he (Mnookin) felt for Halifax and his arguments (pp. 103–104).

Second, and unlike the case of Sharansky, Churchill's position as a national political leader meant that the price of his stubbornness would be visited on his people, who could not possibly be polled to gain their consent on this question. While Mnookin acknowledges this problem in commenting that political leaders do not have the right to “sacrifice” their citizens on the basis of their personal moral intuitions, he does not allow this factor to play a role in his final assessment. Instead, he concludes that “the war cabinet's discussion persuades me that Churchill's refusal to enter into negotiations was wise” (p. 104). This seems to suggest that a cost–benefit debate alone — even if conducted by the decision maker more for form than substance — is sufficient justification for refusing to negotiate so long as adequate pragmatic factors are brought to light in the course of this discussion.

This final gloss on Mnookin's application of cost–benefit analysis left me somewhat puzzled. Mnookin's rule four states: “When deciding on behalf of others, don't allow your own moral intuitions to override a pragmatic judgment [favoring negotiation].” But whose pragmatic judgment does this rule refer to? Is it sufficient that a leader simply sit in a room and listen to various cost–benefit arguments before deciding to follow his instincts? It appears that Churchill may have done just that (Halifax certainly appeared to think so), but Mnookin nevertheless approves Churchill's decision on cost–benefit grounds. And I wonder if that is not more or less what President George W. Bush did in his decision to turn his back on the Taliban — without a Lord Halifax (or Roger Fisher) in the Oval Office to make the pragmatic case for talking.

Perhaps this example involving Churchill and the Nazis shows once again how exceedingly difficult it is to pass judgment on history when one knows that a given decision worked out very, very well in the end. In addition, it may be that the case for negotiating with Hitler is virtually impossible to argue without appearing to be, in some subtle historical sense, a morally suspect “collaborator.”

In this section I will briefly summarize the two cases that complete part two of Bargaining with the Devil and that offer examples of engaging — with different degrees of success — in negotiations with evil people. I can move a bit more quickly through these because the cost–benefit calculus Mnookin prefers is much easier to satisfy when the decision favors talking.

The Nazis marched into Budapest on March 19, 1944 and, before the end of World War II roughly a year later, killed more than 500,000 Hungarian Jews — some 437,000 of whom were shipped to the extermination camp at Auschwitz. Rudolf Kasztner was a Jewish leader in Nazi‐occupied Hungary who made a deal during this occupation with SS Colonel Adolf Eichmann that saved nearly 1,700 Jews from death in concentration camps by organizing a special train for them to Switzerland in exchange for a payment of millions of Swiss francs. Kasztner's story is complicated by postwar charges that he collaborated with the Nazis and neglected to alert fellow Jews that their lives were in danger. I cannot do the full tale justice here, but Mnookin does a masterful job.

Over a period of many months, Kasztner actually created a relationship with the Nazi leaders in Budapest that allowed him not only to strike the deal that led to the famous “Kasztner Train” but also to open talks for the possible saving of 1 million Jewish lives in exchange for 10,000 winterized trucks. This latter deal never came to pass, but it shows the scale on which Kasztner was willing to talk. Moreover, as the war came to an end in the spring of 1945, Kasztner teamed up with a Nazi officer named Kurt Becher, who had been part of the “Kasztner Train” deal, to save still more lives. Becher had been put in charge of a number of concentration camps by SS leader Heinrich Himmler, and he traveled with Kasztner from camp to camp persuading the officers in command to stop the killing and surrender the camps peacefully to the advancing Allies. In exchange, Kasztner promised to testify on Becher's behalf at any postwar trials — a promise he fulfilled and that resulted in Becher escaping any punishment for war crimes.

Was Kasztner justified in negotiating with the Nazis? Once again, Mnookin's answer is “yes.” In fact, this is a no‐brainer. It would be a foolish moral rule that forbid someone from trying to save lives through negotiation in such a desperate situation. But Mnookin hedges. The historical record is so cloudy on the actual facts of Kasztner's relationships with his Nazi counterparts that any resounding positive judgment is difficult. Meanwhile, there is evidence that Kasztner was so focused on his train that he may have ignored larger concerns for his community, become overconfident, and ended up being outmaneuvered by his Nazi counterparts. In other words, in his frantic efforts to save lives he revealed that he was not a great negotiator. But Mnookin applauds him in the end. “Did Kasztner sell his soul to the Devil?” Mnookin asks. “Absolutely not. But he may have been outsmarted.”

A simple cost–benefit analysis appears to be sufficient to take care of Kasztner's case. If he could do some good, then he was justified in trying, provided his efforts did not condemn even more people to death than he was seeking to save. A more interesting moral question in this case, however, might be to ask whether a person in Kasztner's shoes should have a positive “duty to rescue” others through the use of his negotiation skills. Suppose someone decided, on principle, to refuse any attempt by Eichmann and Becher to trade lives for money on the grounds that such deals would risk this person's reputation as a staunch opponent of the Nazi regime. Would the decision to avoid negotiations in such a case be morally defendable under Mnookin's framework?

I think not. If a negotiator has the skill to create deals that will save lives, a cost–benefit analysis would quickly reveal how much might be gained by opening up a channel for discussions. And, as a representative of others' interests, the negotiator would not be permitted under Mnookin's Rule Four to place his or her personal scruples above the pragmatic needs of the community — unless, perhaps, this moral conflict might compromise the person's negotiating skills and render him or her ineffective.

Nelson Mandela is Mnookin's second example of a political leader opting to negotiate with “evil” people — in this case the racist white South African government that had imprisoned him for decades. Once again, Mnookin's retelling of this story is inspiring, and I can give only a brief summary.

For my purposes, the story begins in 1953, when Mandela first publicly advocated armed resistance against South Africa's racist apartheid system as a leader in the African National Congress (ANC) and was censured by his own party for doing so. Matters took a dramatic turn in 1960, when the ANC's chief political rival, the Pan Africanist Congress organized a mass rally in the Sharpeville township to protest laws requiring blacks to carry identity “pass books” at all times. A tiny police force was overwhelmed by this protest, reinforcements were called, and by the end of the day, sixty‐nine blacks had been murdered, many shot in the back while they fled (p. 117).

As a result, the ANC began having similar rallies, chaos ensued, and the government declared a state of emergency while banning both parties. In 1961, Mandela established a military arm for the ANC and declared that it was time to “move on to the next stage: guerrilla warfare and terrorism.” He was arrested a year later, before he had ever led a single terror attack, and in 1964 he was sentenced to life in prison, with the first eighteen years of his incarceration spent on a legendary “hellhole” called Robben Island (p. 119). While in prison, Mandela received numerous offers from the government to release him if he renounced violence. But he refused — on principle — all such attempts to negotiate. “Only free men can negotiate,” he wrote in a statement that was smuggled out of prison. “I cannot and will not give any undertaking at a time when I and you, the people, are not free” (p. 107).

While Sharansky and Mandela shared a dedication to their causes and were willing to endure years of suffering to further these causes, the contrast between the ways they handled prison demonstrate how personality can affect the ultimate success or failure of a negotiation process. Sharansky resisted every attempt by the KGB to appeal to his basic humanity. He treated the KGB “like the weather” and made a game of defeating every attempt to engage with him. Mandela, on the other hand, reached out to get to know his jailers from the very beginning, learning their Afrikaans language, customs, and culture. Mandela believed without qualification in the transformative power of human interaction and communication — even with “evil” archenemies. As he would later write, “Sit down with a man [and] if you have prepared your case very well, that man . . . will never be the same” (p. 120). By the time Mandela was transferred off Robben Island, the prison had become known to many as “Mandela University” because of the wide influence he had over all the prisoners as well as many of the staff.

The turning point in Mandela's relationship with his oppressors came in 1985, some twenty‐three years after he was first jailed as he sat in an isolated cell in Pollsmore Prison. Sensing that the momentum had begun to shift in the political environment — and aware that he and some other ANC leaders had been moved to a new and more hospitable prison for reasons that might be related to the government's willingness to talk on new terms — he decided to open negotiations with his enemy. Crucially, he decided to do so without consulting any of his ANC allies or fellow leaders — some of whom were living just three floors above him in the same jail while others were in exile. Why? Because “my colleagues upstairs would condemn my proposal, and that would kill my initiative even before it was born” (p. 108). He made his first move by sending word to South Africa's minister of justice, Kobie Coetsee, and asking if they could discuss in secret how negotiations between the ANC and the government might be started. As he would later write about this decision, “There are times when a leader must move out ahead of his flock” (p. 108).

Mnookin beautifully characterizes this decision with a metaphor that is worth quoting at length:

The negotiations that followed this tentative opening were to last for five years and involved a tempestuous process marked by ebbs and flows of violence. The first breakthrough came in 1986 as black South Africans engaged in protests around the country and the government declared yet another state of emergency. Coetsee and Mandela met for three hours and began, for the first time, to lay concrete issues on the table. This led eventually to talks in 1988 between Mandela and a special committee of senior government officials, which included the most hated of them all, Neil Barnard, the head of the state's dreaded secret police.

By this time, Mandela had been moved to a small cottage in the Victor Verster Prison, where he had a swimming pool and a personal cook, and he was using his own reluctance to be set free as his main leverage to gain concessions in his negotiations. Finally, on February 2, 1990, President F. W. de Klerk (who would share the Nobel Peace Prize with Mandela) announced that he had agreed to most of Mandela's conditions for release. A few days later, on February 9, the two men went through the final steps of their elaborate “closing” dance. De Klerk stated his intention to release Mandela the next day by flying him to the capital for an elaborate media event. Mandela refused, stating that he wished to leave a week later by walking out the front door of the prison and thanking his guards. They compromised: the release was the next day but it was staged at the prison. It would be years before elections were held and the new South African government was formed. There was still more violence, more massacres, and more mistrust to be overcome. But in 1994, Mandela was elected as South Africa's first black president and he immediately signed a new constitution that guaranteed equal rights for all, regardless of race.

Mnookin considers this to be one of the best negotiations of all time — and awards Mandela “the title of the greatest negotiator of the twentieth century” (p. 135). He especially admires the way Mandela found ways to reframe the traditional structure of negotiations between adversaries. “He rejected the simple‐minded notion that one must either negotiate with the Devil or forcibly resist. He did both. He was willing to make concessions, but not about what was most important to him. With respect to his key political principles, he was immovable” (p. 135). Mnookin also has kind words for de Klerk. He “found a way to end white rule without surrendering white prosperity,” writes Mnookin (p. 134).

As for his moral assessment, Mnookin says simply that “the most important lesson [of Mandela's case] goes to the core of this book: We must reject as foolish the categorical claim that it is wrong to negotiate with an evil adversary” (p. 135). Moreover, Mandela passes Mnookin's systematic four‐step moral test with a 75 percent grade — he followed three out of four rules. Rule One requires a thorough cost–benefit analysis before deciding to initiate negotiations — something Mandela had plenty of time to do in prison. But Mandela violated Rule Two requiring consultation because political necessity forced him to make this risky decision entirely on his own. He definitely followed the “presumption in favor of negotiation” (Rule Three), and finally, he did not allow his impulses against negotiation — which had been part of his moral makeup for decades — to override his “pragmatic assessment” to talk (Rule Four). Indeed, by getting out in front of his followers, he taught both the ANC and the white government how to negotiate in spite of a complete lack of trust. How did he do it? By engaging the power of his persuasive gifts and demonstrating his commitment to a cause greater than either of the parties.

As the examples and the analytic framework in Bargaining with the Devil make obvious, Mnookin is seldom an advocate for refusing to negotiate. Rather, he wishes to make three relatively modest claims. First, evil exists. Labeling people who repeatedly seek to do us grievous harm without adequate justification as devils can be perfectly acceptable. Second, faced with evil counterparts, cost–benefit analysis may often counsel against bargaining with them. Bush's refusal to respond to the Taliban in 2001, Churchill's steadfast refusal to negotiate with Hitler in 1940, and Mandela's uncompromising attitude toward negotiation with the white South African government throughout the 1970s and early 1980s (until his dramatic change of heart in 1985) are all presented to us as solidly justified by rigorous cost–benefit analysis.

Third, there are a few, relatively rare occasions when principle alone is sufficient justification for refusing to talk to an evil counterpart. Mnookin points to Sharansky's remarkable nine‐year stand against the KGB to prove the existence of his hypothesis. Basically, these rare occasions arise when you and everyone who may be affected by your refusal to negotiate are willing to bear the full costs of the decision. And even in these cases, Mnookin imposes a duty to first engage on a full‐fledged cost–benefit analysis to test the wisdom of your moral intuitions. Implicit in his argument is his assumption (stated at the outset) that purely moral motivations are “intuitive” and “emotional” (System 1), whereas cost–benefit analysis is more fully “rational” and deliberative (System 2).

In this section, I wish to return to a point I raised in discussing the Sharansky case and dispute Mnookin's assumption about the nature of moral motivations. In doing so, I hope to demonstrate that the cases Mnookin has assembled can be seen in a new and in some ways clearer light. To do this, I will rely on the work of decision theorist and Stanford professor emeritus James G. March and his book A Primer on Decision Making: How Decisions Happen, written in 1994 with the assistance of Chip Heath.³ For those who may not be familiar with March's work, Daniel Kahneman hailed A Primer on Decision Making as “brilliant, wise . . . and deeply rooted in a large body of modern research.”

March writes that “standard contemporary discourse [regarding rational choice], particularly in the traditions of decision theory, tends to equate reason with the logic of consequences. The idea is that a reasoning decision maker will consider alternatives in terms of their consequences for [their] preferences” (1994: 100–101). Nowhere is this “standard discourse” more prevalent than in the negotiation and dispute resolution communities, as shown by Mnookin's decision to entrust the cost–benefit perspective to that most rational of all characters from fiction, the Vulcan Mr. Spock. The moral perspective on decision making, by contrast, is relegated to the realm of emotion and intuition. As Mnookin sums up the situation at the beginning of Bargaining with the Devil: “I believe moral judgments . . . arise from the intuitive side of the brain: they are gut feelings . . .” (p. 35). In short, there is no “inner rational voice” such as Mr. Spock — say, a fanciful counselor named after political philosopher John Rawls — presented in the book to personify nonconsequentialist rational thinking.

Nor is this surprising. Cost–benefit analysis is an imminently practical model for considering and weighing alternatives in 99 percent of all conflicts and negotiations, even if the practitioners of utilitarian, interest‐based bargaining are sometimes baffled and frustrated by people who stubbornly refuse to follow their well‐meaning, common‐sense advice to act rationally.

But there is a problem: Mnookin has written a book about the other 1 percent of cases — those relatively few situations that uniquely call on us to concern ourselves with moral principles that transcend consequences, costs, and benefits. I contend that in these important cases the decision to negotiate is not a contest between cost–benefit rationality and moral intuition, as Mnookin contends, but is (or ought to be) a contest between rival forms of rationality. And if that is the case, then Mnookin's decision calculus urgently needs an additional, complementary framework.

What might this alternative way of thinking rationally about the decision to negotiate look like? March teaches us that it is neither exceptional nor mysterious. But instead of being based on the logic of consequences, it is based on what March calls the logic of “appropriateness and rules.” Most of us recognize this as the mode of thought we use when we make decisions regarding such things as duty, obligation, and, most important, identity. We make these decisions by asking questions such as: “In this situation, what must I do as a father (or mother)?” or “What would a good Buddhist do here?” Our social roles and identities deeply affect our decision making. Soldiers wearing battle dress must sometimes decide to kill people intentionally, whereas doctors wearing white coats must never do so. Some Jews and Muslims must always refuse pork; some Hindus may not eat meat of any kind.

March holds that decision making based on fully articulated notions of duty, obligation, and identity is no less “rational” than decision making based on consequences, and I find his argument compelling. As a former lawyer who once argued cases that turned on logic, language, history, and custom as well as economics and social policy, I think March's identity‐based approach to rationality places Mnookin's analytic framework in an entirely new light, a light that helps us better understand the limits of all exclusively consequentialist approaches to negotiation and conflict resolution.

Of course, consequentialists know about duty, obligation, and identity, and have done their best to subsume these concerns within the wingspan of cost–benefit analysis. According to March, they make two standard moves to push identity‐based decision making to the side — and both are on display in Bargaining with the Devil. First, they express deep suspicion of the rationality of identity‐based decision factors, typically portraying them as “unthinking and automatic,”“arbitrary” or “imposed” (March 1994: 101). This comment permits us to revisit the basic, dual‐processing model of decision making on which Mnookin relies. Under his model, moral decision making is a System 1 affair governed by intuitions and emotions, whereas cost–benefit analysis is rational and deliberative (System 2).

But is this really the case? Where do the interests and preferences that trigger cost–benefit analysis come from if not from intuition and emotion? Is it really possible for people, even with the help of a Mr. Spock, to separate their understanding of their own interests from the materialism, greed, desire, and fear that drive these interests? Cost–benefit analysis requires us to engage in at least two close‐to‐impossible mental/emotional feats. First, we must anticipate the future when the future, as we all know rationally, is unknowable and our bets about it are biased by wishful thinking, risk aversion, and other unavoidable System 1 mechanisms. Second, we must correctly predict how we will feel in the future once we obtain something we wanted or experience something we feared. Dan Gilbert has recently shown in Stumbling on Happiness (2006) that even if the outcome of our decisions is exactly what we hoped or feared, we are frequently quite surprised by our resulting reactions. We are not as sad as we expected after bad things happen, and the thrill of victory wears off much more quickly than we thought it would, returning us to a general equilibrium state of moderately positive or negative emotion that seems to depend as much on personality as fate.

To his credit, Mnookin acknowledges this problem, telling us at one point that “reasonable people” can “analyze the same [negotiating] situation and reach different conclusions, based on different predictions and different assessments of the costs and benefits”(p. 33). But his analytic framework fails to take the obvious next step and recognize that this indeterminism is based, in part, on how System 1 affects our perceptions of interests — just as it affects our moral instincts.

The second move consequentialists make to deal with (and subsume) identity‐based decision making, according to March, is to absorb it within their cost–benefit framework. “Sure,” pragmatists say, “people's identities and duties are important. And we take these matters into account by listing them on the white board as being among their interests. In the end, they are free to trade off other things to achieve more on these important, identity‐based issues.”

But it is exactly this move that does not help much in the 1 percent of cases that Bargaining with the Devil asks us to consider. Why? Because we are not talking about trading one issue off against another in such cases. We are talking about whether to negotiate at all. This is an all‐or‐nothing decision, and it will often be based on decision factors that are not (ever) subject to compromise.

If we go back and look at the cases of Bush, Sharansky, and Churchill and their refusals to negotiate, what pattern do we see? All three were, first and last, motivated by overwhelming commitments to their senses of duty, obligation, and identity. Cost–benefit analysis had a role in framing the debates around them and may have even affected their decision making at the margin, but March's “identity‐based” decision model does a much cleaner job of explaining both their thinking and behavior.

Let's briefly review each of these cases. I can easily imagine George W. Bush asking himself as the cost–benefit debate raged around him in the wake of 9/11: “What should a United States president do when the nation's largest city is attacked by a radical group of Islamic terrorists harbored by a corrupt regime in Afghanistan? Negotiate with them? I don't think so. History would say I failed as a leader.” One can make a set of fully rational arguments about what the duty of a president might be in this circumstance — and one such duty might be to listen to a cost–benefit analysis. But others might look at what our best presidents in the past have done in similar situations (what did Lincoln do when the Confederacy attacked Fort Sumter?) and what America's cultural identity might be in the wake of a decision to negotiate with terrorists in these circumstances.

The same holds true for both Sharansky and Churchill, who were dominantly motivated, I think, by the former's identity as a Jew and the latter's identification with Britain's historical commitments to duty, honor, and country. (Churchill was a world‐class historian as well as a politician.) Both men were struggling in their respective decision‐making crises to “do the right thing” in terms of the demands of duty and identity, and not just (or even mainly) based on the anticipated, pragmatic consequences that might follow a decision not to negotiate. Interestingly, an identity perspective also goes far toward explaining the “irrational” behavior of some of the parties in the personal stories Mnookin recounts in part three — especially his examples of the stubborn San Francisco Symphony musicians, whose need for dignity and control motivated them to revert to hard bargaining a few years after Mnookin had provided training in collaboration; the bitter wife in his divorce example, who expressed her newfound independent identity through using the litigation process; and the rival sisters fighting over the summer house on Cape Cod, whose reconciliation as siblings made a pragmatic property settlement possible.

Let me offer just one example of my own to conclude my case. In the famous White House debates surrounding the Cuban Missile Crisis in 1962, a crucial moment that helped tip the decision to avoid war came when Attorney General Robert F. Kennedy slipped his brother, President John F. Kennedy, a note that read, “Now I know how Tojo felt when he was planning Pearl Harbor” (Neustadt and May 1986: 6). This was an identity‐based, rational argument against making a surprise air attack against an unprepared enemy. Historians suggest that the Pearl Harbor analogy played just as much a role in the final outcome as did the cost–benefit analyses that were heard over those famous thirteen days. In other words, final decisions in hard cases involving war or peace, prison or freedom, or life and death should be made not by favoring rational cost–benefit analysis over moral intuitions. They should be made by engaging in two different kinds of rational thinking.

Thus, as I see it, Mnookin's four rules need to be supplemented to include an additional test, one that would assure that people confronting evil behave “rationally” not only as cost–benefit analysts but also as identity‐based decision makers. Perhaps these considerations could be added to Rule Three as constituting acceptable grounds for “rebutting the presumption” favoring negotiation. They could also be used to open up Mnookin's overly narrow requirement in Rule Four that negotiations are mandatory unless the “costs of resistance” are consented to by all affected parties. The kind of questions I have in mind might include the following:

1. Is the sense of duty, obligation, or identity on which I am tempted to decide against negotiation a deeply felt, authentic aspect of myself — something connected legitimately to my profession, family, religious faith, national identity, or other social group?

2. Does this sense of duty, obligation, or identity stand up to a reasoned challenge to the legitimacy of the social group I feel connected to — that is, is this social group honorable or is it evil, racist, or dedicated to harming others without justification?

3. What does a reasoned analysis of the history of my group and its past encounters with situations similar to mine tell me about my duty in this case?

4. Do I have competing duties, obligations, or identities to consider in reasoning about whether or not to negotiate? For example, does this situation require me to think about potentially conflicting duties I may owe to more than one social group or conflicting identities I may have as someone who occupies several different social roles?

5. Would my duty, obligation, or identity be compromised simply by speaking with and listening to my adversary — even if I could prevail without any compromises on all material points of our dispute?

6. Are there any ways for negotiations to take place between third parties that would not, if revealed, compromise my duty, obligation, or identity?

A book review essay is not the occasion to fully work out the set of identity‐based questions we would want to ask in a moral inquiry about whether or not to negotiate. But the list I suggest above may provide a start on this project.

On this much, Mnookin and I agree: in the kind of extraordinary, life‐and‐death circumstances presented in the vivid cases Mnookin gives us, we have a moral duty to decide as rationally as possible. But before we can call a decision truly wise, I think we need to expand the idea of rationality beyond the confines of cost–benefit analysis to include reasoned consideration of obligation, duty, and identity.

Ever since Roger Fisher and William Ury published the first edition of Getting to Yes: Negotiating Agreement without Giving In, the 1981 classic of the negotiation literature, everyone in the field (including me, a host of others, and both of Getting to Yes's distinguished original authors) has been trying to figure out how to get beyond it. Ury has given us Getting Past No (revised edition, 1993) and The Power of a Positive No (2007), among other great titles, and Fisher has coauthored Getting Together (Fisher and Brown 1989), Getting It Done (Fisher and Sharpe 1999), and Beyond Reason (Fisher and Shapiro 2006), among other great titles. With its genesis in a spirited debate between Fisher and Mnookin, Bargaining with the Devil is another shot at this elusive target. As I see it, Yes still rules.

Getting to Yes implies — and Fisher forcefully argued in November 2001 — that we should always negotiate if there is even a remote chance that we could gain something from engaging in dialogue with our enemies. Mnookin posits a theoretical exception to that approach: when our adversary is evil, then sometimes it is wrong to negotiate even if we have something to gain from doing so. Whereas Getting to Yes gives us a “principled” approach for conducting the negotiation process, Bargaining with the Devil presents principles for deciding whether, in morally compelling situations, we can walk away from negotiations altogether.

But, as I hope I have shown, Mnookin's new framework basically leaves us where Getting to Yes had already arrived — with a presumption that negotiation is almost always the best road. Indeed, Mnookin even withdraws his support for George Bush's treatment of the Taliban by the end of Bargaining with the Devil, because history had revealed that Bush was not really using cost–benefit analysis at all but had simply adopted a dogmatic stand against negotiation in all cases involving terrorists. (Mnookin quotes with disapproval Vice President Dick Cheney's statement, “We don't negotiate with evil; we defeat it.”)

So Bargaining with the Devil does not really open significant new territory for negotiation process theorists. But it does a magnificent job of opening the door to a whole new kind of moral conversation in both the negotiation literature and the classroom. And it brings us some compelling narratives told with drama and skill that provide a rich context for asking an important new set of questions. What role, if any, does identity‐based bargaining play in the toughest negotiations? Do moral rules that transcend utilitarian cost–benefit analysis have a role beyond the usual “bargaining ethics” aspects of conflict resolution? Should experts in interest‐based negotiations spend more time studying the origins of obligation, duty, and identity as sociological, anthropological, and historical aspects of conflict? And should they treat these factors as distinctly different from conventional “interests”?

Finally, can we get beyond the familiar tactics of interest‐based negotiations to discover a systematic set of best practices — of strategies, leverage dynamics, and psychological variables — that practitioners can use when negotiating with true “devils”? The outlines of such a roadmap are sketched in Bargaining with the Devil— factors such as timing negotiations so they will not be mistaken as signs of weakness, using third‐party channels, managing the media, maintaining the moral high ground, rallying one's own side in a show of temporary strength, and relying on character and moral authority as negotiation assets — but the tactical book for skillful negotiations with truly evil people is not fully developed. Thus, much work remains to be done to create frameworks for conducting negotiations with the likes of Osama bin Laden and the Taliban that will take us far beyond the conventional tactics of both hard‐nosed and interest‐based bargaining.

My hope is that, in such templates, we will begin to recognize that the claims of duty, obligation, and identity require new and different tools to negotiate successfully. They are not the tools we use in adjusting conflicts among different interests. They are the tools of influence, persuasion, metaphor, emotion, and imagery that we use when reaching out to change people's beliefs.