Creating Data on International Negotiation Strategies, Alternatives, and Outcomes John S. Odell 20 April 2001 International Negotiation, forthcoming 1 If greater efforts can be invested in generating uniform data on the international negotiation process, I would nominate at least three key dimensions for fresh attention: strategies, alternatives, and outcomes. Although these are not the only interesting ones, better empirical knowledge about them is surely central. The outcome is the main reason why scholars and practitioners alike are interested in negotiation in the first place. Strategies are the most encompassing representation of what players do when trying to influence the outcome, and for practitioners, strategies represent the range of choices available. Our knowledge of how diverse strategies affect international outcomes is still rudimentary and fragmented. The BATNA--the party’s best alternative to a negotiated agreement with the other party—is highly likely to shape choices during the talks and hence the outcome. This note takes materials from my recent book, Negotiating the World Economy, and extends them a bit. It emphasizes ideas concerning strategies and treats the other two concepts more briefly. This note is suggestive and preliminary in spirit; it does not attempt to resolve many of the technical questions that would arise in producing and using reliable indicators. Strategies The term “strategy” is pressed into service to mean many different things. Meanings range from broader to narrower categories of behavior. In international relations, one of the most common typologies has limited the options to two: cooperation and defection. The list of modifiers that have been attached to “strategy” also includes 2 deterrence, compellence, coercive diplomacy, reciprocity, enhanced contingent restraint, retaliation, bilateral, multilateral, and issue linkage. Pruitt’s and Rubin’s list (1986) postulates contending, yielding, problem solving, breaking off negotiations, and inaction. Young 1991 posits structural, entrepreneurial, and intellectual forms of leadership in institutional bargaining. Negotiating the World Economy, a book that concentrates on international economic negotiations, offers a theoretical strategy continuum that helps pull together some of the concepts that have been suggested separately. Let strategy mean a set of behaviors that are observable at least in principle, and associated with a plan to achieve some objective through bargaining. Strategies are part of the process of negotiation—a sequence of actions in which parties address demands and proposals to one other for the ostensible purposes of reaching an agreement and changing the behavior of at least one actor.1 Tactics are particular actions that make up a strategy. A strategy may remain constant throughout a bargaining episode, or the party may alter its strategy along the way. Any strategy will necessarily be adapted to the special features of the situation at hand; an ideal-type strategy will not manifest exactly the same tactics in every application. Thus defined, a strategy does not necessarily specify every possible response to every conceivable contingency. This notion is different from usage that is common among game theorists. Suppose all behavioral strategy options are represented along a theoretical continuum ranging between two polar ideal types: distributive or value-claiming behavior, and integrative or value-creating behavior.2 This continuum is not special to economic negotiations. At one pole is what I call the pure value-claiming or distributive 3 strategy, a set of actions that promote the attainment of one party's goals when they are in conflict with those of the other parties.3 Offensive claiming tactics attempt to take value from the others, whereas defensive claiming tactics aim to prevent the others from taking value from the first. Distributive behavior is a rational response to a situation in which no settlement can make both or all parties feel better off than they were prior to bargaining. Concretely, a strict value claimer insists on an agreement under which her side will be better off than before at the expense of the others, or at least not worse off. Failure to cite the others’ well being as a reason for negotiating is a sign of this strategy, even if we do not take a statement of such concern at face value. The claimer insists her side will make no concessions and resists doing so. The tactical menu also includes a high opening demand, concealing information about true priorities, and criticism of the other's positions. She may concede at a slow rate; delay; attempt to manipulate the others’ beliefs about their alternatives and her own; and attempt actually to worsen the others’ alternatives and improve her own. The explicit threat is a strong claiming tactic. With a threat the claimer attempts to establish a commitment to a particular demand, in order to lower the other's resistance point, shift the zone of agreement, and rule out certain points inside it. Various ancillary tactics are used to increase threat credibility. Seizing hostages is an even stronger move. The United States has employed relatively pure distributive strategies many times to reduce another country's exports of particular goods. (It is not the case that we observe only strategies in the middle of the spectrum). In these cases Washington demanded a concession from an exporting state but refused to make any compensating concessions of 4 its own, such as increasing its imports of other products from the exporting country. In October 1969, for instance, the U.S. government began a famous episode of claiming from Japan over textiles trade. Washington's aide-mémoire described its objectives concerning "the textile problem" not as seeking mutual benefit but exclusively by reference to American concerns. It notes apprehension on the part of American management and labor over the rapid rise in imports of textiles and the adverse effect this can have on investment decisions and on employment; the importance which the Government of the United States attaches to the creation and preservation of employment opportunities for minority groups in the United States. . . . The message called on Japan's government to accept a comprehensive five-year bilateral agreement establishing a quantitative limit on exports to the United States of every wool and synthetic fiber textile and apparel product. This document pointed to "strong sentiment in the United States Congress for a legislated solution should the attempt to achieve a negotiated solution fail."4 Later, after two years of fruitless bargaining with Japan, President Nixon escalated from this thinly veiled threat to a public threat to impose such limits unilaterally. Claiming is not restricted to the most powerful states. While offensive claiming by the weak against the strong is rare, defensive claiming is common among all states. A poor country requesting a gift and declining to reciprocate is attempting to shift value from one party to another. Delay and refusal to make unrequited concessions are forms of defensive claiming. 5 The claiming strategy can include the tactical retreat--agreeing to accept less than demanded earlier or give more than offered earlier, without a compensating gain. Falling back belongs fundamentally to a distributive process by which pies are divided, rather than a process that expands pies or bakes new ones, unless the fall back is part of a linked exchange of concessions.5 At the opposite pole is the pure integrative or value-creating strategy. It involves actions that promote the attainment of goals that are not in fundamental conflict, actions designed to expand the pie. Behaving as a pure value-creator, the first agent would propose that the two parties negotiate toward an agreement designed to make both, not just her side, better off. In Walton's and McKersie's well-known formulation based on U.S. labor- management negotiations, 6 the ideal-type integrative process moves sequentially through three phases. At first the parties explore, discover, and reveal the nature of the problem and their own objectives and priorities. Throughout, openness about information is favored (except information about the parties’ true reservation values). The negotiator will state positions so as to facilitate identification of, rather than to conceal, underlying concerns and priorities. She operates less like an institutional role-player with a fixed brief to read, and more like individual assigned to take the initiative in solving a problem that affects both sides. In a tense dispute, the pure value-creator listens carefully and probes for underlying reasoning rather than immediately responding to attack with equivalent attacks. The second phase is a joint search for potential solutions, which may include baking a different pie rather than splitting the existing one, and joint effort to estimate 6 options’ likely consequences. Desired solutions stated in specific terms replace sermons about abstract principles. Here too inventiveness is favored. In the final phase of an ideal integrative negotiation, problem solvers combine their utility functions in some way and test the invented alternative solutions against them. At this stage if not earlier, the parties explore for concessions that might be valued differently and thus exchanged. The integrative bargainer treats all partial agreements as provisional until all issues are settled, to avoid excluding a possible re-packaging that would move the outcome closer to the frontier. Note that the second pole does not mean giving in or yielding unrequited concessions. On the contrary, this concept denotes behaviors that are meant to make the negotiator’s own side better off while making the other better off as well. Nor do integrative behaviors necessarily indicate that the negotiator is motivated by altruism, though altruism is not excluded by definition. Nothing requires a negotiator to choose one of these pure polar strategies. Indeed, the ideal of pure integrative behavior has not been documented often in international economic negotiations--only occasionally, as a phase in a longer process.7 Negotiators sometimes mix distributive and integrative tactics, so that observed strategies vary by degrees between the poles. The strategy spectrum ranges theoretically from pure claiming to claiming diluted by minor integrative moves, to a balanced mix, to mostly value-creative tactics diluted by mild claiming moves. To negotiate by mixing, the agent will commit her government to search for an agreement that benefits all parties and not rule out possible concessions by her side. She 7 will invite the others to express their true concerns and priorities as opposed to their public positions, and will avoid the harshest claiming tactics. Yet she will also delay her own concessions and attempt to bias the distribution of joint gains in her direction. She will defend her proposal and resist claiming by others. She may propose a principle of fairness or a formula that would realize gains for both while also shifting their distribution her way.8 In principle mixing could be accomplished either sequentially or simultaneously. One negotiator could begin with a relatively pure integrative strategy and then in a second phase, when joint gains are in sight, shift to vigorous claiming over their division. Another could attempt the opposite sequence. A third negotiator might try to mix tactics from both sides of the continuum from the beginning and throughout.9 In international relations two mixed strategies from the integrative side can be seen in the 1942-1944 process that resulted in the Bretton Woods monetary agreement. Both the U.K. and the U.S. governments took the initiative, each putting forward a proposal for a new international organization designed to benefit many parties, not only itself. For the U.S. this relatively integrative position was a dramatic change from the historical record. The strategy was diluted with mild distributive tactics to protect special U.S. objectives. The American negotiator initially proposed that Washington must have a veto over the organization’s decisions. He rejected Keynes’s basic scheme to create a new international money and expose the U.S. to far greater financial liability. The GATT’s Tokyo round of trade negotiations in the 1970s also illustrated mixed strategies. This round has been described as "a rulemaking exercise of major proportions."10 While bargaining to create this larger pie, the typical party also attempted 8 to claim the largest possible share of the gain for itself. In still other cases, we observe mixed strategies from the other side of the spectrum--mainly distributive but diluted with mild integrative tactics.11 Theoretical tit-for-tat can be seen as one variant of a mixed strategy, wherein the negotiator rewards the other for favorable moves and punishes the other for harmful ones. Most, however, have conceived tit-for-tat more narrowly and mechanically, excluding the joint discovery and creative activity captured by the notion of integrative behavior.12 One previous operational scale of distributive and integrative negotiating behaviors also misses this key innovation in Walton’s and McKersie’s theory. Donohue and Roberto 1996 define a set of actions ranging in order from threatening to complying. The latter pole does not attempt to capture to exploratory information exchanges, brainstorming, or searches for acceptable logrolls. Turning the proposed theoretical continuum into a reliable indicator would require considerable technical work. As a first step, an appendix offers provisional operational definitions of these main strategy types. Subsequent steps would be based on the assumption is that a strategy is a pattern of actions with some stability through time, rather than behavior that is expected to change from one sentence to the next during a discussion. Thus coding strategies would mean looking for patterns of behavior in the flow of micro-acts, not classifying micro-acts one by one. At the same time, the idea of mixing elements from a continuum allows a way to represent some change in the observed pattern during a single negotiation. To reduce reliability problems in coding a relatively encompassing concept like this would naturally require development of appropriately complex coding instructions, defining precisely what would count as a shift 9 to a different strategy. They would need to show how to merge inferences from different forms of evidence, such as official statements and post-hoc interviews. The instructions would need to define the time boundaries of the negotiation, the forms of evidence to be coded (transcripts of secret talks, public statements issued by parties, press accounts, interviews with participants after the fact), the unit of observation or point of measurement, and the like. As usual, the instructions would need to be tested in pilot studies and refined. BATNAs A second worthwhile indicator would estimate the value of the party’s best alternative to negotiated agreement and changes in that value. The batna refers to the best course of action the party could take if the talks ended in impasse. Theoretically, behavior and ultimately outcomes depend centrally on the parties’ alternatives. The worse the party’s alternative, the lower its reservation value and the more eager it will be to sign a deal. An improving alternative outside the talks can be expected tend to shift the party toward harder distributive tactics inside. The party’s batna could change through time and its value could in principle be measured continuously throughout the negotiation. One approach here would be a set of material indicators, each of which proxies the most important factors for a particular issue area. For a party negotiating an armistice during an internal war, the proportion of the population under its control might proxy its prospects of winning the struggle without negotiation. A decline in this proportion during the talks would lead us to expect the negotiator to soften her demands in negotiations 10 over settlement terms. In Bosnia in 1995 as the battlefield tide turned against the Serbians, Slobodan Milosevic spoke for the Bosnian Serbs and softened their position in the bargaining, and the Muslim Sarajevo government stiffened its own. For some trade negotiations, a rise in the price of the country’s chief export commodities could indicate an improving alternative to negotiating an official agreement over that commodity. For a country with a pegged exchange rate negotiating with foreign lenders during a balance of payments crisis, the quantity of official financial reserves relative to the country’s import needs can indicate how badly it needs a deal. Countries’ gross power assets and ratios between these assets are probably less valid indicators, since countries allocate many power assets to particular purposes and cannot convert them to other purposes quickly or without substantial losses. Each of these is a tangible condition that we would expect to affect the value the party would place on its alternative at a given time. Any material indicator, however, is one step removed from the perceived batna, the subjective value the negotiator actually places on her alternative, her actual belief about how badly she needs an agreement. Perceptions could diverge from simple material conditions for both systematic and non- systematic reasons. Consider the 2001 Hainan Island incident, when American and Chinese aircraft collided and the Chinese government held the American crew and plane hostage while demanding U.S. concessions. For that negotiation one might read China’s or America’s batna off the distribution of power assets between the two countries, the bilateral economic dependence running in each direction, China’s physical control over the American aircraft and crew, and other material conditions. But evidence about what Beijing’s and Washington’s leaders actually believed to be their best alternative courses, 11 in light of their subjective priorities, would certainly get us closer to an understanding of their behavior. More generally, analysts would benefit if they had a set of valid generalizations identifying relationships between material conditions and the most likely beliefs, for example showing from evidence how the perceived value of one’s alternative tends to fluctuate in response to changes in material conditions, over time in the same country, or even in groups of countries. Of course creating uniform data to measure perceptions about alternatives or other cognitive variables is a complex task. Not only does official secrecy restrict outsiders’ access to evidence about negotiators’ true beliefs. When evidence is in hand, each form has at least some shortcomings. Scholars at different times have turned to public statements by leaders and negotiators,13 editorials of government-controlled newspapers, memoirs, participants’ unpublished personal papers, transcripts of historical meetings,14 interviews with participants during15 or after the events, questionnaires completed by expert observers, histories, and combinations of these sources. Young and Schafer (1998) review the state of the art in classifying and describing foreign policy leaders’ operational codes, cognitive maps, images, and conceptual complexity. Scholars have improved these methods over the years but these techniques still fall short in significant respects. Many of the applications have been qualitative rather than quantitative and many have not concentrated on negotiations as such. Thus there seems to be potential for additional fruitful work in this area, building on this foundation and probably using multiple complementary forms of evidence. Outcomes 12 Presumably the ultimate reason to study the negotiation process is the assumption that it affects the international outcomes. Yet many international relations scholars who concentrate most on international and domestic power structures and institutions remain skeptical that detailed negotiation process studies really add value sufficient to justify their costs, over and above what can be known about outcomes by studying only background conditions. For negotiation researchers, I believe a high priority should be placed on pinning down the difference the process makes to the outcome, net of other influences. This too is much easier said than done, of course. The negotiation outcome in the most general sense means whether the interaction ended in agreement or deadlock, and how much was gained or lost by each party and by the group of parties. The result could be either the shifting of value from one party to another, or the joint solving of a problem with gains for all. Outcomes are also sometimes compared as to their efficiency, their fairness, and their stability.16 How much a negotiator gained or lost is understood most often by reference to the status quo prior to the talks. Is the party better off, worse off, or about the same as before? In a single case study the terms of an armistice might be compared to the distribution of territory held by each fighting force when the talks began. The terms of an official “voluntary” agreement to restrict Korea’s textile exports can be compared to the higher value of its exports in a base period, measuring Korea’s loss due to the bargaining. When parties have intangible as well as tangible goals, a quantitative measure limited to the tangible will be an imperfect approximation. Furthermore, assessing an outcome relative to the status quo ante can be unfair to a negotiator whose batna worsened after the process began, for instance because another 13 party issued a credible threat. It can be rational to agree to a loss relative to the status quo ante if the best available alternative today is worse still. Defining gain and loss relative to the party’s changeable batna gets us closer theoretically to the decisions that negotiators really face. Thus in principle, a second outcome indicator could compare the terms of the settlement for each party with a batna estimate for that party at the time of agreement. Was the party better off, worse off, or about the same as it would have been had it walked away from the table at that time? Yet faithfully measuring outcomes relative to batnas would raise complex interpretive questions that arise when attempting to measure batnas empirically. To add another illustrative technical issue to the list, if party A has issued a credible threat against B, the value of B’s alternative logically should reflect a judgment about whether A was credible or bluffing, since B’s true alternative was much better if A was bluffing. Generally, assessing what A and B would have done had B walked away is inherently speculative, and so is the value of the counterfactual alternative. Counterfactual argumentation can in principle be carried out with some care in a case study and might be persuasive to many readers.17 But no one to my knowledge has developed procedures for employing counterfactual arguments uniformly across a large number of cases. A few attempts to measure outcomes over large numbers of cases illustrate possibilities and problems. Some have used the simpler approach of defining relative “success” operationally from one party’s standpoint. Hufbauer, Schott and Elliott 1990 classify 116 attempts to use economic sanctions for foreign policy purposes between 1914 and 1990 as to the relative success or failure for the sanctioner. Bayard and Elliott 1994 rate the degree of success in 72 United States attempts to extract trade concessions using 14 its section 301 law. Bercovitch 1996 classifies 593 attempts at international mediation between 1945 and 1990 as to their success or failure. Druckman et al. 1999, in contrast, classify 30 international negotiation outcomes as to how many disputed issues they covered and whether the parties expanded joint gains, split benefits symmetrically, split them asymmetrically, or ended in impasse. Others have observed outcome variations by means of comparative case studies. These projects employ what amounts to ordinal qualitative measurement. George, Hall and Simons 1971 contrast the greater U.S. success in using a strategy of coercive diplomacy against Laos 1961 and Cuba 1962 with the failure vis-à-vis Hanoi in 1965. Walton et al. 1994 observe negotiations between business managers and unions in 13 U.S. cases during the 1980s, and classify the outcomes uniformly as to the degree of change in the substantive contract and the degree of change in union-management cooperation. This study also measures management strategies qualitatively and reports which strategy tended to achieve the greatest gains. Evans et al. 1993, Milner 1997, and Odell 2000 illustrate the use of paired cases with outcomes that vary clearly but not quantitatively.18 Ordinal measurement in a comparative case design is less precise than in quantitative work and the sample is of unknown representativeness. But the conclusions reflect much fuller information about and analysis of the cases selected, and they still benefit from the leverage of comparison. Some mixture of qualitative and quantitative methods in negotiation studies will be needed, since each compensates for shortcomings of the other.19 * * * 15 In conclusion, new investments in creating data for analyzing international negotiations should, in my view, give high priority to measuring negotiation strategies, alternatives, and outcomes. Subsequent research could then study the effects of changing batnas on behavior, and the conditions under which a given strategy gains more and less. Such data creation projects would face many theoretical and technical dilemmas that would require careful thought. It would also cost significant amounts of time and money to measure negotiation dimensions satisfactorily over large numbers of cases--more than single researchers could mobilize separately. But the United States, other countries, and international organizations have spent much larger sums on other forms of scientific research and on military forces. In the larger scheme of things, if more advanced preparatory work showed that this form of research would be fruitful, allocating resources to a better understanding of international negotiation would be a matter of setting priorities. 16 Appendix Operational definitions of strategy types VALUE-CLAIMING STRATEGY. Code a party's strategy as "pure claiming" if any of the following actions are observed and no more than a small minority of the behavior fits the definition of "value-creating." BOTH DEFENSIVE AND OFFENSIVE VARIANTS. The negotiator criticizes the other country's (or countries’) actions or arrangements, blames it for the problem under discussion; attempts to exclude from the agenda issues on which her own country would probably have to make concessions; rejects or ignores demands for concessions or delays their consideration; avoids saying her own country is partly responsible for the problem under discussion, avoids expressing concern for the other’s objectives or a desire for a mutual-gain outcome, avoids making a proposal characterized a beneficial to the other parties or the world as a whole; manipulates information for her own advantage: avoids revealing information about own genuine objectives and priorities; makes arguments whose effect is to support her demands or refusal to concede and does not present information or arguments that are inconsistent with that position; e.g., argues that the other's alternative to agreement is worse for them than they realize, that our alternative is better than they realize, or that the other's forecasts showing future improvement 17 for us (in absence of agreement) are not convincing, or that she simply does not have the capacity to deliver what is demanded; or that the other's proposal would harm our side or others; establishes a commitment to a particular outcome, by means of some public action tied to that outcome such that accepting less would be costly to the negotiator or her country; denies that he or she believes the other's commitments. OFFENSIVE VARIANT: The negotiator also: demands concessions for the benefit of his or her own country; files a formal complaint against the other under global or regional rules; takes steps to worsen the other's alternative to agreement and improve her own-- e.g. unilateral actions or negotiations with third parties that would help compensate it for a breakdown in relations with the other or provide itself with a superior alternative, or raise the cost of a breakdown for the other; actions could include introducing draft legislation for official consideration at home or "talking the national currency down"; threatens to take action harmful to others unless they yield the desired concessions; actually imposes such penalties and implements its alternative to agreement. DEFENSIVE VARIANT. The negotiator also: brings a counter-complaint against the other under international rules; 18 threatens or imposes counter-threatens sanctions. VALUE-CREATING STRATEGY. Code a party’s strategy as "pure value-creating" if the following actions or tactics are observed and if no more than a small minority of the behavior fits “value-claiming:” The negotiator states that the two (or more) parties have an interest in common or expresses concern for an objective held by the other; proposes negotiations designed to benefit both sides, usually aiming to agree on a joint approach to a common problem or an exchange of concessions; praises the other and avoids public statements criticizing the other country or blaming it for the problem or issue under discussion; invites the other to state frankly its genuine concerns and objectives and their priority order, as distinguished from its demands and proposals; proposes and implements a series of meetings whose only or main purpose is to engage the parties in joint study of problems and objectives they have in common; uses and refers to information about the issue or problem without shaping it to her own side's advantage; engages in an "even-handed" discussion of all the facts whether favorable or unfavorable to her side; proposes an exchange of concessions for mutual benefit; argues that a different conception of other's interests or a redefinition of the issues could lead to an agreement that would benefit both parties; proposes an agreement described as helpful to other parties as well. 19 MIXED OR COMBINED STRATEGY. Code a party’s behavior in a conflict or negotiation as a “mixed” strategy if “claiming” and “creating” tactics are mixed in some proportion, either simultaneously, or in a sequence dominated by claiming in one phase and creating in another. 20 Reference List Bayard, Thomas O., and Kimberly Ann Elliott. 1994. Reciprocity and Retaliation in U.S. Trade Policy. Washington: Institute for International Economics. Bercovitch, Jacob. 1996. Understanding Mediation's Role in Preventive Diplomacy. Negotiation Journal. 12:241-58. Beyers, Jan, and Guido Dierickx. 1997. Nationality and European Negotiations: the Working Groups of the Council of Ministers. European Journal of International Relations 3: 435-71. Bonham, G. Matthew, Victor M. Sergeev, Pavel B. Parshin. 1997. The Limited Test-Ban Agreement: Emergence of New Knowledge Structures in International Negotiations. International Studies Quarterly 41:215-240. Destler, I. M., Haruhiro Fukui, and Hideo Sato. 1979. The Textile Wrangle: Conflict in Japanese-American Relations, 1969-1971. Ithaca: Cornell University Press. Donohue, William A., and Anthony J. Roberto. 1996. An Empirical Examination of Three Models of Integrative and Distributive Bargaining. International Journal of Conflict Management 7:209-229 21 Druckman, Daniel, Jennifer Martin, Susan Allen Nan, and Dimostenis Yagcioglu. 1999. Dimensions of International Negotiation: A Test of Iklé’s Typology. Group Decision and Negotiation 8:89-108. Duffy, Gavan, Brian K. Frederking, and Seth A. Tucker. 1998. Language Games: Dialogical Analysis of INF Negotiations. International Studies Quarterly 42: 271-294 Evans, Peter B., Harold K. Jacobson, and Robert D. Putnam, eds. 1993. Double-Edged Diplomacy: International Bargaining and Domestic Politics. Berkeley: University of California Press. Fearon, James D. 1991. Counterfactuals and Hypothesis Testing in Political Science. World Politics 43:169-95. Friedheim, Robert L. 1994. Negotiating the New Ocean Regime. Columbia: University of South Carolina Press. George, Alexander L., David K. Hall, and William R. Simons. 1971. The Limits of Coercive Diplomacy: Laos-Cuba-Vietnam. Boston: Little Brown. Goldstein, Joshua, and John Freeman. 1990. Three Way Street: Strategic Reciprocity in World Politics. Chicago: University of Chicago Press. Hopmann, P. Terrence. 1974. Bargaining in Arms Control Negotiations: The Sea Beds Denuclearization Treaty. International Organization 28, no. 3: 313-443. 22 ________ . 1986. The Negotiation Process and the Resolution of International Conflicts. Columbia: University of South Carolina Press. Hufbauer, Gary C., Jeffrey J. Schott, and Kimberly Ann Elliott. 1990. Economic Sanctions Reconsidered, 2d ed. Washington: Institute for International Economics. Leng, Russell J. 1993. Reciprocating Influence Strategies in Interstate Crisis Bargaining. Journal of Conflict Resolution 37, no. 1: 3-41. Lax, David A., and James K. Sebenius. 1986. The Manager as Negotiator. New York: Free Press. Levi, Ariel, and Philip E. Tetlock. 1980. A Cognitive Analysis of Japan’s 1941 Decision for War. Journal of Conflict Resolution. 24:195-211. Milner, Helen V. 1997. Interests, Institutions, and Information: Domestic Politics and International Relations. Princeton: Princeton University Press. Odell, John S. 2000. Negotiating the World Economy. Ithaca: Cornell University Press. _________ (forthcoming 2001). Case Study Methods in International Political Economy. International Studies Perspectives, and in The Analysis of International Relations, ed. Detlef Sprinz and Yael Wolinsky. Pruitt, Dean G., and Jeffrey Z. Rubin. 1986. Social Conflict: Escalation, Stalemate and Settlement. New York: Random House. 23 Rhodes, Carolyn. 1993. Reciprocity, US Trade Policy, and the GATT Regime. Ithaca, NY: Cornell University Press. Sebenius, James K. 1992. Negotiation Analysis: A Characterization and Review. Management Science 38, no. 1: 19-38. Tetlock, Philip E., and Aaron Belkin, eds. 1996. Counterfactual Thought Experiments in World Politics: Logical, Methodological, and Psychological Perspectives. Princeton: Princeton University Press. Walton, Richard E., and Robert B. McKersie. 1965. A Behavioral Theory of Labor Negotiations: An Analysis of a Social Interaction System. New York: McGraw- Hill. Walton, Richard E., Joel E. Cutcher-Gershenfeld, and Robert B. McKersie. 1994. Strategic Negotiations: A Theory of Change in Labor-Management Relations. Boston: Harvard Business School Press. Weber, Steve. 1991. Cooperation and Discord in U.S.-Soviet Arms Control. Princeton: Princeton University Press. Winham, Gilbert R. 1986. International Trade and the Tokyo Round Negotiation. Princeton: Princeton University Press. ———. 1977. Negotiation as a Management Process. World Politics 30: 87-114. 24 Young, Michael D., and Mark Schafer. 1998. Is There Method in Our Madness? Ways of Assessing Cognition in International Relations. Mershon International Studies Review 42, supp. 1: 63-96. Young, Oran R. 1991. Political Leadership and Regime Formation: On the Development of Institutions in International Society. International Organization 45:281-308. Zartman, I. William, and Maureen R. Berman. 1982. The Practical Negotiator. New Haven, Connecticut: Yale University Press. 1 I use “negotiation” and “bargaining” interchangeably. 2 This continuum is inspired by the investigations by Walton and McKersie (1965 and 1994) on labor-management negotiation, where the concepts have proven valuable empirically, and by Lax and Sebenius 1986, who coined the terms value-claiming and value-creating and made important additions to the theory. Also see Sebenius 1992. 3 Walton and McKersie 1965, chap. II; Lax and Sebenius 1986, chapter 2. 4 United States aide-memoire of October 2, 1969, rpt. in Destler, Fukui, and Sato 1979, pp. 339-41. 5 Some theorists, e.g., Pruitt and Rubin 1986, postulate yielding or making concessions as a strategy separate from other strategies. But reducing demands or making concessions can have more than one meaning and can overlap with other strategy concepts. Giving up concessions that are unrequited shifts value from one party to another on balance. Making a concession in exchange for a concession by the other can make both better off. To me it seems clearer, more useful, and consistent with other theory to recognize these two as parts of different strategies rather than to combine the two into a single strategy type. 6 Walton and McKersie 1965, chaps. IV and V. U.S. labor-management experience of the 1980s led these authors to modify their earlier theory and make it more complex in some respects (Walton et al 1994). The simpler 1965 formulation seems better suited to present purposes. 25 7 For example, see Winham 1986, pp. 169-174, on the unusual process whereby European Community and U.S. negotiators broke a deadlock over rules on trade subsidies and countervailing measures during the Tokyo Round. 8 See Winham 1977 for illustrations and Zartman and Berman 1982 for the useful concept of the negotiation formula. 9 Walton et al 1994, with strategy types defined slightly differently, find that mixing gained more than either of their polar types in 13 episodes of U.S. labor-management negotiation during the 1980s. 10 Winham 1986, p. 12. 11 Negotiating the World Economy, chap. 7, develops two examples. 12 Most empirical studies of tit-for-tat in international relations have concentrated on military-political issues; see Hopmann 1974, Goldstein and Freeman 1990, Leng 1993, and Weber 1991. An exception is Rhodes 1993. 13 E.g., Bonham, Sergeev, and Parshin 1997, Duffy, Frederking and Tucker 1998; also see Hopmann in the present issue. 14 E.g., Levi and Tetlock 1980, Friedheim 1994. 15 E.g., Beyers and Dierickx 1997. 16 Hopmann 1996, 28-30. 17 Fearon 1991 and Tetlock and Belkin 1996 discuss the techniques of counterfactual argument. 18 Also see Druckman, the present issue, for other examples. 19 Odell (forthcoming 2001) assesses the strengths and limitations of case study methods relative to statistical methods.