The Power to Blame as a Source of Leverage: International Mediation and ‘Dead Cat Diplomacy’

Summary This article addresses a gap in the literature on international mediation by proposing the power to blame as an additional source of mediation leverage that had been hitherto largely ignored. The power to blame is framed here as ‘dead cat diplomacy’, a term originally coined by US Secretary of State James Baker to describe his threats to lay a figurative dead cat at the doorstep of a disputant to publicly signal its intransigence and thus force its acquiescence during the Middle East negotiations following the 1991 Gulf War. Drawing on the case studies of Baker and presidents Obama and Trump, the article presents three conditions necessary for the successful leveraging of the power to blame in international mediation: it must be used as a last resort, be perceived as credible by the targeted disputant and take place at a time when the targeted disputant’s bargaining capacity is limited.


Introduction
Why do some mediators threaten to publicly blame an intransigent party in order to save the negotiations from collapsing? What are the determinants of Siniver The Hague Journal of Diplomacy 17 (2022)  Beyond the contribution to our theoretical knowledge of the conditions necessary for the effective application of DCD, this article also fills an important gap in our empirical understanding of the occurrence and outcome of mediation. For example, existing studies of US foreign policy and Arab-Israeli diplomacy in the aftermath of the first Gulf War commonly take historical and systemic approaches to explain the diplomatic breakthrough as a moment of 'ripeness' following the collapse of the Soviet Union and the emergence of the United States as the undisputed principal third party in the region, or as a logical by-product of new realities at the national level which help reorientate domestic politics towards conflict resolution.6 While valid, such explanations do not tell us much about how Baker's style of mediation helped to bring three intransigent parties (Syria, Israel and the Palestine Liberation Organization [PLO]) to the negotiation table. Understanding Baker's unique mediation style in this context thus contributes to our empirical and theoretical knowledge of the application of this source of leverage in mediation beyond the ubiquitous systemic explanations of the 'New Middle East' and the reshaping of regional politics as functions of US foreign policy interests.7 The recourse to DCD is not privileged to great powers. Less powerful mediators may also resort to this action. Consider, for example, Kofi Annan's failed mediation efforts to end the Syrian civil war in 2012. As the UN and Arab League envoy to Syria, Annan's public blaming of the Assad regime (for refusing to implement his proposed peace plan), the opposition forces (for escalating the military campaign) and even the UN Security Council (for lacking unity and leadership) was so ineffective that he was eventually forced to withdraw his mediation.8 Other applications of DCD by non-US mediators (with mixed results) include the sequential mediation efforts of Julius Nyerere and Nelson Mandela during the Arusha process in the 1990s,9 Martti Ahtisaari's successful mediation in Aceh in 2005 and China's mediation between the United States and North Korea in 2005 over the latter's nuclear programme.10 I suggest that for DCD to be effective, three conditions must be met: the mediator must use it as a last-resort strategy, the targeted disputant must perceive the mediator's threat as credible and the bargaining power of the targeted disputant must be reduced due to internal constraints or the external 6 See, e.g., Rubin 1994;Pruitt 1997;Zartman 1997;Eisenberg and Caplan 1998;Shlaim 2001;Quandt 2005;Peleg and Scham 2010. 7 On 'mediation as foreign policy' , see Touval 2003. 8 Hinnebusch and Zartman 2016. 9 Southall 2006; Mediation Support Project 2008. 10 Baikoeni and Oishi 2016;Shea 2016;Kydd 2006b. environment in which they operate. Conversely, premature use of DCD, even by a powerful mediator, may detrimentally affect its credibility in the eyes of the disputants, while disputants who calculate that the costs of acquiescing to the mediator's threat do not outweigh the benefits of intransigence will choose to ignore it. This article proceeds as follows. In the first section, I survey the literature on power mediation and the merits of manipulative strategies in order to place the power to blame in the relevant theoretical context. Second, I frame the power to blame as DCD, based on James Baker's successful efforts at persuading Arabs and Israelis to attend the Madrid peace conference in October 1991. Third, I examine Barack Obama's and Donald Trump's failed engagement with the Israeli-Palestinian peace process to evaluate the ineffective use of DCD by other powerful mediators. Methodologically, this article adopts a qualitative comparative case study analysis. Relying on primary and secondary sources, the applications of the power to blame by Baker, Obama and Trump are used here as confirmatory cases to test the robustness of the DCD model. The Obama and Trump cases were selected because they represent the 'most different systems design'11 in that while both were US presidents and could be considered as 'power mediators' , their political world views, attitudes towards the conflict, diplomatic styles and approaches to negotiations were diametrically opposed, yet their application of the power to blame yielded similar results. I conclude by drawing out the theoretical and policy implications of the power to blame as a useful analytical tool to test the effectiveness of such manipulative strategies by third parties.

Power and International Mediation
International mediation is performed by an acceptable third party for the purpose of changing the course or outcome of a particular conflict and is widely regarded as the most common form of conflict management and resolution. It is distinguished from other forms of non-violent third-party interventions such as conciliation, adjudication and arbitration by its ad hoc, non-binding and non-legal nature.12 The vast literature on the relationship between antecedents and outcomes of mediation has long pointed to multicausality, from the motives for mediation Siniver The Hague Journal of Diplomacy 17 (2022) 216-247 and the mediator's roles and behaviour to notions of ripeness, credibility and leverage in mediation.13 This article is concerned with the pure-power typology as a source of leverage in the mediation literature. It proposes the 'power to blame' as an additional source of leverage in international mediation. Power in mediation is commonly understood as the degree of influence that a mediator can exert on the disputants in order to move them towards conciliation: the more powerful the mediator is, the more resources it can employ to change the disputants' attitudes, perceptions and motivations.14 Leverage is defined here as the capacity to influence the cost-benefit calculations of the disputants through the exertion of various sources of power in material as well as nonmaterial terms, whether positive (carrots) or negative (sticks).
Given the idiosyncrasies in the types of conflicts, the timing of intervention, level of violence and the identity of the disputants as well as the mediator, the literature is divided over which type of mediation is more effective and when. Some scholars maintain that the exercise of power in mediation does not sit comfortably with disputants' expectations of impartiality and neutrality, and that the exertion of power may impinge on the mediator's expected role of a credible and trustworthy 'honest broker' .15 Others point out that impartiality may get in the way of enabling the mediator to use its resources to exert concessions from the disputants and pursue more directive strategies to get the job done.16 As noted above, all mediators have power, though some have more power than others. Therefore rather than being mutually exclusive, the pure-power typology in the mediation literature represents two ideal types of intermediaries' roles and functions, neither of which is inherently 'better' than the other: a low-power, low-stakes, neutral mediator, and a high-power, high-stakes, principal mediator.17 The distinction here is not only between mediators' resources, their bargaining capability and interests in the conflict, but also between two styles of intervention. According to Princen,18 the goal of the neutral mediator is to 'create realistic empathy' by targeting the mode of interaction, whereas the principal mediator aims to enhance the incentives for agreement by affecting the 'payoff structure' in the bargaining process. Perceived impartiality is 13 Bercovitch and Langley 1993;Kleiboer 1996;Crocker et al. 2003;Savun 2008;Beardsley 2011;Wallensteen and Svensson 2014;Duursma 2014. 14 Zartman andTouval 1996b;Bercovitch andHouston 2000. 15 Young 1967;Moore 1996;Fisher 1996;Susskind and Cruikshank 1987;Skjelsbaek 1991. 16 Bercovitch 1986Touval 1982;Carnevale 2002;Kydd 2003, 2006a. 17 Princen 1992Smith 1994. 18 Princen 1992, 29. thus more important to 'pure' mediators whose mandate to mediate rests upon their reputation -and indeed the disputants' expectation -that they will act as honest brokers. For 'power' mediators, however, impartiality is not necessary, and indeed not expected by the disputants. What matters is the mediator's ability to deliver. As Wilkenfeld et al. suggest, 'more active mediation offers more to parties in conflict' .19 The pure-power typology is intrinsically linked to the type of strategies that third parties may apply to achieve their aims. Much of the literature on this subject draws on Touval and Zartman's classic typology of facilitative, formulative and manipulative strategies of mediation.20 Stretching along a passiveactive spectrum of mediation strategies, they correspond to different obstacles in the mediation process, which accordingly call for a different level of engagement by the mediator. On the most passive or 'pure' end of the spectrum, the mediator-as-communicator is required when direct channels of communication between the disputants are not yet established, thereby requiring the mediator to act primarily as a passive conduit of messages and provide 'good offices' . The more active role of formulation is necessary when the disputants cannot agree on how to move forward, thus requiring the mediator to assume a more substantive part in proposing new agendas and persuading the parties to think creatively about ways to move forward. This article is concerned with the most active mode of interaction, the mediator-as-manipulator, where the lines between mediation and coercion are sometimes blurred.21 While the literature is inconclusive over which mediation strategies are more effective in different phases and types of conflict, a large body of scholarship maintains that directive/manipulative strategies of mediation are more effective in bringing about a successful outcome.22 Moreover, even mediators who are perceived as impartial by the disputants may resort to certain manipulative tactics to control the situation.23 In its most extreme form, the manipulative mediator may become part of the solution, if not the conflict itself. This type of manipulative mediator is likely to have direct interests in the disputed issues, unlike Princen's 19 Wilkenfeld et al. 2005, 79. 20 Touval and Zartman 1985;Pruitt 2000;Wilkenfeld et al. 2005;Beardsley et al. 2006;Svensson 2007;Vuković 2011;Bergmann andNiemann 2015. 21 Favretto 2009. 22 Wallensteen and Svensson 2014. For peace practitioners, a more appropriate distinction can be made between facilitative/directive mediation and high-powered diplomacy, to reflect the acceptability of the mediator by the parties as well as the styles of mediation employed in the process (Baumann and Clayton 2017). 23 Kydd 2003.

Siniver
The Hague Journal of Diplomacy 17 (2022) 216-247 'principal mediator' who only has indirect interests such as regional security or access to natural resources. For example, the United States' historical role as a principal mediator in the Arab-Israeli conflict is undisputed given its special relationship with Israel and its regional strategic interests, although at times it has become an integral part of the bargaining process by intertwining its own interests with those of Israel. According to Aaron Miller, former Middle East envoy under the Bush Sr. and Clinton administrations, '[f]or far too long, many American officials involved in Arab-Israeli peace-making, myself included, have acted as Israel's attorney, catering and coordinating with the Israelis at the expense of successful peace negotiations' .24 Zartman and Touval's typology of mediation strategies is supported by five sources of power on which the mediator can draw to bring the parties towards an agreement:25 the purely communicative exercise of persuasion; the more active extraction of concessions; the threat of termination (of the process); and finally the power to use gratification (carrots) and deprivation (sticks). Elsewhere, Rubin proposes that mediators may draw on six bases of power to exert leverage over the disputants:26 reward and coercion (Zartman and Touval's 'gratification and deprivation'); expert power, deriving from the mediator's relevant knowledge and expertise; legitimate power, based on legal authority or international law; referent power, stemming from the relationship between the mediator and at least one of the disputants; and informational power, deriving from the mediator's access to, and use of, information to affect disputant behaviour. A more detailed yet not essentially different typology is Beardsley et al.'s aggregated model of mechanisms and tactics of mediation.27 None of these oft-cited typologies refers to the power to blame as an additional source of leverage in international mediation.

3
The Power to Blame I suggest an additional source of leverage which has been largely overlooked in existing studies of international mediation. The power to blame is defined here as a distinct type of manipulative strategy whereby the mediator threatens to blame one of the disputants for the looming failure of the process as 24 Miller 2005. 25 Zartman and Touval 1996a. 26 Rubin 1992. 27 Beardsley et al. 2006, 66. The Power to Blame as a Source of Leverage The Hague Journal of Diplomacy 17 (2022) 216-247 a last-resort effort to prevent the break-up of negotiations.28 Unlike Zartman and Touval's 'termination' , where the mediator threatens to withdraw from the process and leave the disputants to their own devices, here the mediator goes a step further to threaten to lay the blame for the failure of the process on one of the disputants, thus publicly choosing sides between the 'intransigent' party and the 'co-operative' party. While in reality both disputants may be equally intransigent, the mediator's apportionment of blame to one party at one time is more effective in building a narrative of 'responsibility of failure' while distancing the mediator from its own failings. A notable example of such a war of narratives took place in the aftermath of the failed Camp David Summit in 2000 between the Israelis and the Palestinians. The immediate narrative of blame was articulated by the Clinton administration and the Israeli government: it blamed the Palestinian leader Yasser Arafat for turning down the most generous peace offer ever made by an Israeli prime minister, and for instigating the bloody second intifada against Israel to detract attention from his opposition to peace with Israel. A counter-narrative soon followed, laying the blame for the failure of the summit on the close collusion between the American and Israeli teams throughout the negotiations, and on the inherent unjustness in Prime Minister Ehud Barack's 'generous' peace offer. Inevitably, later accounts point to failings on the part of all three parties to various degrees.29 The power to blame can be thought of as an extension of the mediator's use of carrots and sticks to reward the disputants for their co-operation or punish them for their stubbornness: by threatening to publicly blame one of the disputants, the mediator indicates not only that the bargaining process is over, but also that it is due to intransigence on the part of one of the disputants (thus distinguishing it from the mere termination of the process as a source of leverage). The power to blame as a source of leverage is similar to Kydd's blame casting,30 although Kydd's model assumes that the mediator is powerless and therefore cannot, unlike Baker and other powerful mediators, 'simply offer rewards or threaten punishments directly' , and so 'there must be some other powerful audience before whom the parties wish to appear accommodating' . Another study that looks at blaming in international conflicts is Herfroy-Mischler and Friedman's 'blame game frame' ,31 although their focus is 28 This practice is the exact opposite of scapegoating in mediation, where the mediator agrees to take the blame to allow the disputants to save face domestically. See, for example, Lord Carrington's mediation in Rhodesia in 1979 (Novak 2009 on how the blame game plays out in media coverage of the mediation process, rather than its application as leverage by the mediator against an intransigent party. Finally, Hafner-Burton,32 and Squatrito, Lundgren and Sommerer, refer to 'naming and shaming' as a strategy employed by international organisations in support of mediation activities.33

3.1
Conditions for Effectiveness I propose that the effective use of the power to blame rests on three conditions: the mediator must use it as a 'last resort' ultimatum; the targeted disputant must perceive the mediator's threat as credible; and the bargaining power of the targeted disputant must be low due to internal politics or the external environment in which they operate. Conversely, premature use of the power to blame, even by a powerful mediator, may detrimentally affect its credibility in the eyes of the disputants, while a disputant with little to lose by withdrawing from the process will not accede to the mediator's threat and may even win domestic support for standing up to pressure from a great power. There is no hierarchy of importance or sequential relevance as far as the interaction between the three conditions is concerned. What matters is how the act is perceived by the targeted disputant as part of the dynamic of the mediation process and its relations with the mediator.
Firstly, as a last-resort ultimatum, the threat to blame removes the disputant's capacity to bargain with the mediator for a better deal -it either returns to the table if it considers the threat (and the mediator) credible, or it is willing to accept the costs of intransigence if it believes that they are outweighed by the benefits of heeding the mediator's threat. As noted above, the threat also signals to the targeted disputant that there is no more room for bargainingit either acquiesces or faces the consequences of being publicly blamed for failing the process. The use of an ultimatum as part of the dead cat strategy is thus seen here as a hostile act on the part of the mediator, as it represents a willingness to hurt the feelings of a disputant through displays of resentment, anger, rudeness, spitefulness and negative evaluations of the other party.34 The issuing of an ultimatum by the mediator is commonly seen in the literature as an act of hostility which undermines trust and is likely to promote negative emotions between the disputants and the mediator.35 There is similar support in non-International Relations studies (including economics, psychology and 32 Hafner-Burton 2008. 33 Squatrito, Lundgren and Sommerer 2019. 34 Buss 1961. 35 Fischer 1991Susskind, McKearnan and Thomas-Larmer 1999. business) that rational actors perceive ultimatums as unfair and hostile acts and will thus respond in an angry or intransigent manner.36 Similarly, Deutsch found that the use of threat in a bargaining context limited the potential for an agreement.37 Other studies further suggest that a mediator's toughness (as manifested through threats) is likely to make the disputants more willing to offer concessions.38 A case in point is UN mediator Ralph Bunche who, facing the breakdown of talks between Israeli and Egyptian delegations during the mediation of the 1949 armistice agreements, presented the two sides with commemorative plates to celebrate the occasion -with the accompanying threat that he would break the plates over their heads if they did not reach an agreement.39 Another example is seasoned Finish Prime Minister Martti Ahtisaari's reputation as a 'hostile' mediator for his tried and tested tactic of listening with a sour expression followed by an outburst and him throwing his pencil on the table, signalling to the parties that he meant business.40 This leads us to the second condition for effective DCD. The most common explanation for the success of some ultimatums in negotiations and the failure of others is the perceived credibility of the mediator and their threat. There is consensus in the literature that credibility is intrinsically associated with mediation effectiveness.41 Maoz and Terris define mediator credibility as 'the extent to which disputants believe the mediator's statements, threats, or promises … Broadly speaking, the more credible the mediator is perceived by the disputant, the more accepting the disputant will be of her offers' .42 The mediator's credibility is not only a function of its perception by the disputants but also a reflection of the mediator's role and standing in the international community, or at least in the immediate community of conflict. As Kleiboer and Pressman note, the effective application of such 'coercive leadership' is predicated on the mediator's capacity to deliver on its threat or promise.43 While the essence of the power to blame is non-material, it is more likely to be exercised by 'stronger states' who possess the material as well as non-material resources to 'arm-twist' and 'bribe' the disputants.44 Similarly, Svensson points out that an important characteristic for mediator credibility is the capacity to deliver 36 Pillutla on promises, suggesting that great powers are more likely to be effective mediators than small states or organisations.45 However, the credible use of the power to blame is not only material-based, similarly to Beardsley's intangible leverage.46 It also echoes Reid's concept of credibility leverage which refers to 'the influence mediators wield with information, contextual knowledge of the conflict, and a perceived commitment to the peace process' .47 As Nathan notes, credibility also 'emanate[s] from moral stature rather than formal power' ,48 citing various examples of international mediation in African civil wars. Credibility is also closely tied to the mediator's relationship with the disputants, but it should not be treated as a simple function of outdated binary notions of impartiality (i.e., that a credible mediator must be impartial and that bias stands in the way of perceived credibility). Instead, credibility should be seen as one distinguished but related aspect of mediator characteristics, together with its perceived impartiality, interests in the dispute and inclination to exercise degrees of leverage on the parties. Accordingly, perceptions of mediator bias may explain the decision of the parties to accept its involvement, while perceptions of credibility can explain why the parties choose to accept the mediator's proposals. In other words, disputants may reject or acquiesce to the mediator's threat because they perceive it as credible (i.e., the mediator's willingness and capacity to deliver on the threat), regardless of whether the mediator is perceived as impartial or biased. Notable examples include American mediation in Southern Africa and the Middle East.49 Finally, credibility (like most other factors in mediation) is conflict dependent and must be understood within the wider context of relationships, history, bargaining behaviour and conflict cycle which together determine the effectiveness of the mediation process. As Crocker et al. suggest, DCD may be a logical strategy when the mediator wishes to prevent the collapse of the process by conveying a sense of urgency to the intransigent party that the train is about to leave the station with or without it.50 It is clear, therefore, that credibility 'has an impact on the occurrence, process, and outcome of mediation' .51 Finally, the credibility of the mediator and its timing of the threat are more likely to be effective if the targeted party's domestic or external standing is 45 Svensson 2009 weak. A disputant who cannot afford the costs of being blamed publicly (whether because it will be punished domestically or will suffer the reputational costs of intransigence on the world stage) is more likely to take heed of the dead cat lying on its doorstep. There is wide support in the literature for the argument that the potential reputational costs of defiance or intransigence in negotiations can modify state behaviour, although it is not a sufficient condition to explain foreign policy. Nevertheless, the likelihood that repeated intransigence may adversely affect the reputation of the state on the international stage, or the personal character of its leader, may contribute to at least tactical changes at the negotiation table. Some of these reputational costs may include perceived damage to a state's international standing and credibility, which may result in difficulties in alliance formation or maintenance; the increased likelihood of the militarisation of the dispute; and challenges to informal co-operation, such as collaborative voting at the UN or ad hoc political or military support.52 Non-compliance on the international stage may also raise questions about the government's commitment to respect the limits on its power domestically.53 I distinguish here disputant susceptibility to the threat as a different catalyst of ripeness in mediation -it is not wholly independent from the dynamics of ripeness but rather provides an additional useful prism to understand the motivation of disputants to return to the negotiation table. Ripeness is commonly treated as 'willingness' or 'readiness' on behalf of the disputants to engage constructively in a sustained process of negotiations.54 Here, however, ripeness is understood as a degree of vulnerability on the part of a disputant which drives it to acquiesce to the mediator's threat. Moreover, as others have noted, the concept of stalemate as sine qua non for mediation occurrence has been somewhat overstated in the literature.55 Whereas ripeness theory emphasises the need for a mutually hurting stalemate to bring both parties to the table, here what matters is not the symmetry of pain between the disputants but the mediator's credible and timely threat to blame an intransigent party in order the save the process from collapsing. More precisely, it is not the mediator's threat per se which informs the disputant's decision, but rather it is the constraining environment in which the disputant operates which makes it 52 Brewster 2009;Blandfold 2011. 53 Guzman 2002Downs and Jones 2002;Simmons 2002. 54 Zartman 1985, 2000, 2001Rubin 1991;Kleiboer 1994;Haas 1988;Crocker et al. 2003;Beardsley 2010. 55 Greig 2005

Siniver
The Hague Journal of Diplomacy 17 (2022) 216-247 more vulnerable to the threat. Here the disputant succumbs to the mediator's threat out of fear of the costs associated with intransigence at home or abroad, rather than actively seeking compromise out of sheer optimism in the process. In other words, a hurting disputant is more likely to acquiesce to the mediator's threat if it calculates that the costs of being blamed publicly for derailing the process outweigh the benefits of maintaining a status quo policy of intransigence. It also follows that the precondition of a mutually hurting stalemate to understand ripeness (and thus mediation success) does not apply here, since DCD can only work if the mediator targets one disputant at one time during the process. If both disputants find dead cats on their doorsteps at the same time, then the mediator's threat is no longer perceived as credible but rather is seen as desperate, and each disputant will calculate that it is better off with the collapse of the process than yielding to the threat.

James Baker: The Effective Application of 'Dead Cat Diplomacy'
The term 'dead cat diplomacy' was first coined by former US Secretary of State James Baker (1989Baker ( -1992 during his mediation efforts to persuade the Israelis, Syrians and Palestinians to attend the Madrid peace conference in the aftermath of the 1991 Gulf War. The Madrid conference was the first part of a threepronged US-led mediation strategy, intended to be succeeded by a series of bilateral and multilateral meetings in Washington on common issues such as water, the environment, arms control, refugees and economic development.
The multilateral track was favoured by the Arab states, who preferred to negotiate with Israel from a unified position, whereas Israel preferred to maximise its relative strength vis-à-vis individual Arab states through the bilateral track.56 The Madrid conference brought together Israeli and Arab leaders for the first time since the 1973 Geneva conference, while the bilateral track which followed saw the first direct talks between Israel, Jordan, Syria and Lebanon since the armistice talks of 1949. The Madrid conference was also the first appearance of the Palestinians at the negotiation Arab-Israeli conflict lacking purpose and direction, regional and global events soon set in motion a process that culminated in Washington finding itself in the unprecedented position of a principal mediator and an honest broker in the Middle East. First, the fall of the Soviet Union and the demise of bipolarity not only weakened Arab states such as Syria, who could no longer rely on Moscow's political and military aid, but it also moved the regional system from being dominated by superpower competition towards co-operation. Second, the Iraqi invasion of Kuwait in August 1990 and with it the threat to a significant portion of the world's oil reserves was effectively countered by a US-led coalition of 28 states. When the war ended in March 1991, the myth of Arab unity against Israel was shattered as Egypt, Saudi Arabia, Syria and the Gulf states joined the coalition forces. The Bush administration earned the respect and gratitude of the Gulf states for removing the threat posed by Saddam Hussein and, just as importantly, for stopping short of occupation or regime change. The Bush administration also proved capable of winning the support of the United Nations and many Arab states while protecting Israeli security interests at the same time. For the first (and only) time in Middle East history, all actors accepted the undisputed role of the United States as the indispensable and only respected mediator in post-war diplomacy. As Joseph Nye pointed out at the time, 'as the action over Iraq has shown, where America leads, others follow -this year, next year and for the following generation' .58 However, despite these propitious conditions for American mediation, a diplomatic breakthrough was slow to come. As post-war euphoria turned into diplomatic attrition, Baker had to draw on every resource and skill in the mediation toolbox to persuade the parties to come to the Madrid peace conference. He made eight trips to the region between March and October 1991, meeting with the leaders of Israel, Egypt, Syria, Jordan and Saudi Arabia, as well as with a non-PLO Palestinian delegation in Jerusalem.59 The Palestinians, Israelis and Syrians proved particularly difficult to work with. Seeking ways to exert maximum leverage on the parties and avoid diplomatic failure, Baker conceded that under those circumstances, '[t]he only leverage a secretary of state has in that kind of a situation is the ability to blame one party or the other for the collapse of the peace process' .60 Reflecting on his mediation efforts in the run-up to the Madrid conference some years later, Baker noted that '[o]ne of the strongest points of leverage, with respect to all parties, was the threat to, as I found myself saying all too often, lay the dead cat on their doorstep. No one wanted 58 Nye 1991, 48. 59 Baker 1999. 60 PBS 1999.

Siniver
The Hague Journal of Diplomacy 17 (2022) 216-247 to accept the blame for scuttling the process. Some days this felt like the only leverage I had.'61 Once Baker left the figurative dead cat on the doorstep of one of the intransigent parties, he would use the press (on 'background')62 to reveal who killed the cat -something that none of the parties wanted to be blamed for.63 According to Aaron Miller, one of Baker's closest Mideast advisors, the secretary of state could not trace the origins of this phrase, but it was apparent that throughout the negotiations dead cat diplomacy was one of Baker's 'central instruments in getting Assad, Shamir, and the PLO to agree on terms of reference for Madrid' .64 According to Kurtzer and Lasensky, along with President Bush's 'clearly defined presidential priority' and the full consultations with the parties throughout the process, it was Baker's 'sustained diplomacy' and 'determination during the negotiations' endgame' which led to the success of Madrid.65 Similarly, Shimon Peres recalled that Baker 'proved an adept and gifted arm-twister; he exercised precisely the right amount of pressure on all parties to ensure that the conference took place' .66 Others have described Baker as 'a masterly salesman and deal maker … [with] uncanny instinct for knowing how to achieve short-range tactical goals'67 and even suggesting that without Baker's 'express threat of the "dead cat award" … Madrid would not have happened' .68 All three components of effective DCD could be found in Baker's leveraging of the Israelis, Palestinians and Syrians: it was used as a last-resort strategy, it was perceived as credible and the bargaining power of the disputants was low due to internal and/or external constraints. Israel had reasons to be wary of the aftermath of the Gulf War. Struggling to absorb the influx of more than a million immigrants from the former Soviet bloc, it faced constant pressure from the Bush administration to stop settlement activity and to ease the lives of the Palestinians. From the outset, relations between Secretary Baker and the right-wing Israeli government, led by the intransigent Yitzhak Shamir, were far from cordial. Baker's call on Israel to 'lay aside, once and for 61 Baker 1999, 188. 62 'On background' comment to the press allows the source to speak freely, but without being identified by name (e.g., as 'senior official'). 63 Kurtzer et al. 2012, 70. 64 Miller 2009, 197. 65 Kurtzer and Lasensky 2008, 30. 66 Peres 1995 Christison 1994, 40. 68 Thomas 2007, 169.  The Syrians and Palestinians, too, were vulnerable to American pressure. The collapse of the Soviet Union ended decades of political, military and diplomatic support to the regime of Syrian President Hafez al-Assad, while the PLO's miscalculated support of Saddam Hussein during the war had cost it considerable economic and political support from the Arab states. It was isolated by the international community while Israel felt vindicated that, in his support of Saddam Hussein, Yasser Arafat had unmasked his true self as a terrorist. The PLO was also financially close to bankruptcy after Saudi Arabia and Kuwait stopped their payments to the organisation. Finally, the Tunis-based PLO's leadership had been undermined internally for several years by the growing popularity on the Palestinian streets of Hamas, the new Islamic resistance group which emerged in 1987 following the outbreak of the Palestinian 69 Friedman 1989. 70 Rubin 1995, 197. 71 Bush 1991. 72 Shlaim 2001, 485-487;Quandt 2005, 306-307.

Siniver
The Hague Journal of Diplomacy 17 (2022) 216-247 uprising (intifada) in the West Bank and the Gaza Strip.73 Yet for all these constraints on the Israeli and Palestinian leaderships, their respective positions remained unchanged throughout the negotiations on a single issue -the participation of the PLO at the peace conference. While Shamir flatly refused to deal with Arafat, even indirectly, the Palestinians insisted that the PLO was their sole legitimate representative.74 In his initial dealings with the parties, Baker focused mainly on procedural matters and on the development of a broad framework of an international conference followed by rounds of bilateral and multilateral negotiations between Israel and its Arab neighbours. Yet as the Israelis, Palestinians and Syrians continued to oppose Baker's proposals, and with the prospect of a historic missed opportunity at the height of America's prestige in the region, Baker's procedural mediation turned gradually into manipulation, with DCD as its most frequently used leverage to save the negotiations from collapsing. Baker's approach was underpinned by the sole aim of creating a situation 'in which no party could withdraw from the negotiations without appearing opposed to peace' .75 As Aaron Miller noted, 'no one wanted to be in that position ' .76 Soon, dead cats began appearing on the doorsteps of Baker's interlocutors at the most opportune moments. When President Assad backtracked from a previous agreement, the secretary of state 'looked hard at him across the table, slammed his leather portfolio shut, and said, "Mr. President, I don't think we can do business together"' .77 As he terminated the meeting, Baker told Assad: 'If I get the right answers … I'll be back in Damascus … If I get the wrong ones, I don't expect to see you again for a long time' .78 On another occasion, Baker reminded Assad who had the upper hand: '"Remember that in the end, this is our invitation" [to the Madrid conference], I said. "At some point, we will issue it -regardless of Syrian objections'".79 In all such instances, Baker made the threat to blame the Syrians publicly for their intransigence, repeatedly referring to Assad as the 'chief impediment to progress' in the media. Such tactics had the desired effect, with Syrian Foreign Minister Farouk Al-Sharaa pleading with Baker's aide, Dennis Ross, 'to never say such things about Syria again' .80 73 Indyk 1991. 74 Kurtzer et al. 2012, 23. 75 Bannerman 1992, 146. 76 Miller 2009, 218. 77 Djerejian 2009, 96-97. 78 Baker 1995, 457. 79 Baker 1995 Ross 2005 Baker applied his self-described 'tough as nails' approach with equal rigour to his dealings with Shamir.81 'I'm working my ass off, and I'm getting no cooperation from you. I'm finished … I've got to say I'm basically disinclined to come here again.' On the way to the airport, Baker told Dennis Ross, 'I'm going to leave this dead cat on his doorstep' .82 Despite the United States' special relationship with Israel, and unlike the vast majority of his predecessors and successors, Baker did not hesitate to lay the dead cat at Shamir's doorstep in public: appearing before the House Appropriations Subcommittee on Middle East Peace Mission, Baker repeatedly called out Israel's settlement activity as the biggest obstacle to peace, noting that 'nothing has made my job … more difficult than being greeted by a new settlement activity every time I arrive' .83 During another appearance before the House, Baker left little doubt about whom he blamed for the mothballed process: 'Everybody over there [in Israel] should know that the number of the White House is 1-202-456-1414. When you get serious about peace, call us' .84 Even the US ambassador to Israel, William Brown, found Baker's mediation style 'disturbing. It was tougher than I thought necessary.'85 Baker's treatment of the Palestinians was equally tough. According to Hannan Ashrawi, a member of the Palestinian delegation and its chief interpreter, 'Baker's most striking feature were those cold piercing eyes that seemed capable of looking straight into your mind and calculating alternative tactics for mental assault. This is a no-nonsense person, I thought to myself' .86 As with his dealings with the Israelis and the Syrians, he made explicit threats to publicly blame the Palestinians for failing the process. 'You don't want a story [in the press] that the cat died on the Palestinian doorstep' , he told them at the conclusion of one meeting.87 On another occasion, he told the Palestinian delegates: 'You're lucky I'm talking to you. Your guy backed the wrong horse' , in reference to Yasser Arafat's support of Saddam Hussein during the Gulf War.88 On the eve of the Madrid peace conference, as the Palestinian delegation raised some last-minute reservations, Baker again signalled his willingness to let them hang out to dry, telling them: 'I am sick and tired of this. With you people, the 81 Miller 2009, 220. 82 Baker 1995, 457-458. 83 Baker 1991, 184. 84 Friedman 1990. 85 Brown 1998. 86 Ashrawi 1995, 82. 87 Baker 1995 Miller 2009 As noted above, the mediation literature commonly places the power to extract concessions from the disputants and the power to terminate negotiations as the most extreme forms of manipulation. Such threats can only be deployed effectively if the mediator is respected, trusted or feared. Otherwise, this manipulation may backfire and only serve to further embolden an intransigent disputant, as well as to damage the prestige and credibility of the mediator, perhaps beyond repair. In Baker's case, DCD worked because the mediator's power to blame was perceived as credible by disputants whose power to haggle had been weakened in the aftermath of the Gulf War. According to Ashrawi, all the Arab delegations to Madrid (including the Syrians, Palestinians, Jordanians and Lebanese) viewed Baker's threats as credible, noting that '[o]bviously, he was not an easy man to fathom or to cross' .93 She recalled Baker shouting at her over the phone following a last-minute demand by the Syrians: "You just tell [Syrian foreign minister] Mr. Shara that the whole thing is off. I'm going home. I'm taking the plane this evening and he can go back to Syria. As far as I'm concerned, it's finished!" And he hung up on me … I went out into the living room and reported the gist of the conversation to the Arab group. Everyone was convinced that Baker was serious, and we urged the Syrians to accept an Arab compromise.94 In Baker's own words, the lessons from his diplomatic approach must not be forgotten: Even when the goal is lofty, the means to that end must sometimes be less than elegant. While appealing to a high purpose, I would also have 89 Baker 1995, 507. 90 Ross 2007 Ashrawi , 87. 92 Miller 2009, 197. 93 Ashrawi 1995, 82. 94 Ashrawi 1995 to cajole and coerce and, again, use leverage in every instance possible. If I didn't, peace would be the loser, but American credibility and prestige would also suffer.95

Barack Obama and Donald Trump: How Not to Lay Dead Cats
On the surface, a comparative analysis of the mediation approaches of presidents Obama and Trump to Israeli-Palestinian negotiations seems inapt given their stark differences: US-Israel relations during the Obama administration had often been portrayed as the most 'difficult' since the administration of George H. W. Bush, while Trump is widely considered the most pro-Israel president since Ronald Reagan. Similarly, Trump's close personal relationship with Israeli Prime Minister Benjamin Netanyahu is well documented, while the evidence of the mutual antipathy between Obama and Netanyahu is equally substantial. Finally, the gap between the two presidents' diplomatic styles and political world views can only be matched by the gap between Israeli and Palestinian positions on the core issues of the conflict.96 Yet despite their differences in substance and style, both Obama and Trump fell into the same trap in their dealings with the Israelis and Palestinians. Like Baker, they did not hesitate to use DCD to publicly blame an intransigent disputant for the failure of negotiations -yet whereas Baker's application of DCD was timely, credible and effective, both Obama and Trump exercised this ultimate leverage too early in the process, at great risk to their prestige and the credibility of their respective administrations, and at a time when the intransigent parties did not feel sufficiently hurt by the discovery of dead cats at their doorsteps to change their policies.
Observers of Baker's mediation have pointed to the meticulous decisionmaking process of Baker's experienced team of advisers, and the intensive preparation and deliberation which went into each part of the process.97 The use of DCD was thus not a desperate act of brinkmanship but a calculated exercise of American leverage at the most opportune time in the negotiations. Both Obama and Trump, however, were quick to point the finger publicly at the Israelis and Palestinians, respectively, even before procedural negotiations had commenced: in the spring of 2009, less than three months into his administration, Obama publicly called out Israeli intransigence and 95 Baker 1999, 192 In each case, DCD failed not only because the American president laid the dead cat at the Israeli or Palestinian doorstep too early, before the commencement of even procedural mediation, but also because neither party felt sufficiently weak to acquiesce to the mediator's manipulation. Whereas during Baker's mediation efforts the Israelis and Palestinians were economically and politically vulnerable and depended on America's goodwill, Obama and Trump faced Israeli and Palestinian leaders who not only had little political reason to acquiesce to American pressure, but who also scored valuable points at home for standing up to this pressure: Netanyahu had won the Israeli general election in February 2009, only a month after Obama entered the White House, and he had little reason to jeopardise his political gains by entering into risky negotiations with the Palestinians (and soon his attention turned to Iran's nuclear aspirations), while for President Mahmoud Abbas of the Palestinian Authority, any diplomatic gesture towards the Trump administration would have emboldened his Hamas rivals and led to popular unrest in Palestine and beyond. As a result, in both cases the disputants calculated that the costs of 98 Froomkin 2009;Siniver 2011. 99 Sanchez 2018Trump 2018. 100 Breslow 2016. 101 Erekat 2018Eriksson 2018. acquiescing to the mediator's exercise of leverage would be outweighed by the benefits of standing firm. As chief Palestinian negotiator Saeb Erekat noted, '[i]f the art of their negotiations is to put us in a position where we have nothing to lose, I think they succeeded' ,102 while Netanyahu repeatedly boasted of facing down the pressure from Obama to halt settlement activity, thus appealing to his wide right-wing support base in Israel and the United States.103 As a result of the poor timing of laying the dead cats and lack of sufficient constraints on the Israeli and Palestinian leaderships to acquiesce to American pressure, the remaining condition for successful DCD -the credibility of the power to blame -was also lacking in both cases. In their survey of American diplomacy in the Arab-Israeli conflict, Kurtzer and Lasensky conclude that ' [a] successful envoy needs to be viewed as credible by all sides' .104 This condition had clearly not been met -indeed, Obama's and Trump's lack of credibility in delivering dead cats was compounded by an external environment that was significantly less conducive to the exercise of American manipulation in the context of the Israeli-Palestinian conflict. Whereas Baker entered Mideast diplomacy at the height of America's prestige and credibility following the Gulf War and with the backing of the UN and an international coalition of allies, Obama attempted to mediate Israeli-Palestinian peace while battling a global recession and fighting to end the wars in Iraq and Afghanistan (and soon, facing the challenge of Iran's nuclear programme). While Trump's foreign policy agenda was less demanding than Obama's during his first year in office, his very election was met with suspicion, despair and fear among America's allies and adversaries alike (with Israel being one of the few exceptions, following eight years of Obama's perceived pro-Palestinian bias).105 Taken together, the failures of Obama and Trump to exercise this source of leverage (despite being power mediators) in their dealings with Israelis and Palestinians present a compelling case of how DCD should not be attempted.

Conclusion
The manipulation of third parties is not unique to international mediation or the United States. International organisations (including the United Nations), non-governmental organisations and the news media often resort to 'naming and shaming' practices to coerce human rights violators to improve their Siniver The Hague Journal of Diplomacy 17 (2022) 216-247 record or to bring them to the negotiation table,106 although the framing of the power to blame as a source of leverage remains largely understudied in the mediation literature. As a strategy designed to avoid a looming failure by threatening to publicly blame the disputants, DCD is undoubtedly positioned at the manipulative end of the spectrum of mediation strategies. Moreover, as a high-leverage, highstakes strategy it can most effectively be performed by third parties who possess the capabilities and resolve to manipulate the disputants in such a way while accepting that in doing so their perceived impartiality (if it exists at all) may be jeopardised. Indeed, as noted by Smith, power mediators may resort to such manipulative strategies to punish and reward behaviour, but their impartiality, 'while perhaps desirable, is unnecessary to secure an agreement' .107 Conversely, pure mediators, whose very essence is defined by their image as honest brokers, cannot, and must not, resort to such manipulative strategies.
While it is well established that the more resources the mediator possesses, the greater its ability to move the parties towards a solution, it is the non-binding and non-violent nature of mediation which places the ultimate power in the hands of the disputants, rather than the mediator. Faced with a stalemate, even the most powerful and resourceful mediator cannot force its will on intransigent disputants. In this regard, DCD is by definition a last-resort manipulative strategy -the only tool left in the mediator's toolbox which can prevent the irreversible collapse of the process. James Baker used this tool effectively and credibly only after other efforts to keep the parties on track had failed; Barack Obama and Donald Trump seemed eager to lay dead cats at the disputants' doorsteps even before procedural negotiations had commenced.
In his 1996 survey of five decades of US policy towards the Middle East, Michael Hudson observed that 'one must be impressed by the diplomacy of US president George Bush and secretary of state James Baker' . Referring to Baker's adroit 'exploitation' of the Israelis, Arabs and Palestinians in the run-up to the Madrid conference, Hudson wished that '[f]uture students of diplomacy will study the elaborate, multi-track structure of negotiations that Baker created' .108 This article has taken a modest step in understanding what lessons can be learned from Baker's effective mediation, and under what conditions similar practices of DCD are likely to be successful. However, beyond the contribution to our understanding of DCD as the power to blame, this article points to broader lessons and potential areas for future research about the conduct of The Power to Blame as a Source of Leverage The Hague Journal of Diplomacy 17 (2022) 216-247 mediation as a tool of foreign policy. For example, there is scope to examine the interplay between the power to blame and other sources of leverage, most pertinently the power to terminate; does the threat to blame necessarily follow a (failed) threat to terminate the process? And conversely, why do some mediators choose not to threaten to blame an intransigent party despite possessing the credibility to do so and when their other strategies to leverage the party have failed? Two other future research agendas that can build on the findings in this article are the comparative effects of DCD on state vs non-state actors, and the use of different media outlets/strategies by the mediator to affect the behaviour of the disputant party.