Introduction
Although mediation is accepted as a crucial component of conflict resolution in international relations, opinion is divided among scholars about which mediation strategies are most effective, and what the role of coercion should be. This article examines these issues in the context of intractable identity-based conflict, specifically the Israeli-Palestinian conflict. Those who argue in favour of ‘pure’ mediation, like John Burton and Frank Dukes, reject the idea that third-party mediators should use leverage and coercion, whereas these factors are central to ‘power’ mediation as understood by Saadia Touval and Peter Carnevale.Footnote 1 However, as Jeffrey Z. Rubin, William Zartman and Touval, and Luke Heemsbergen and Asaf Siniver have argued, this binary distinction overlooks the many nuances of power involved in mediation, which are better conceptualised on a spectrum from the material to the social-psychological.Footnote 2
The effectiveness of certain forms of power depends on what the mediator is trying to accomplish. If the parties’ positions are far from one another and their dialogue is not well developed, then something closer to ‘pure’ mediation is more relevant to build a relationship and increase mutual understanding.Footnote 3 If looking to secure an agreement, quantitative analysis of large conflict mediation datasets suggests that coercive use of leverage is the most effective.Footnote 4 Indeed, Harold Saunders has argued that the effectiveness of a particular mediation strategy can vary depending on the stage of the conflict.Footnote 5 However, studies on effectiveness have yet to focus specifically on identity-based conflict, which tend to be the most intractable, nor do they explicitly consider how the sequencing of mediation activities impacts on effectiveness.Footnote 6 Ronald J. Fisher and Loraleigh Keashly’s ‘contingency model’ of third-party intervention provides an important basis for thinking about the relationship between strategies, power, and timing, reflecting on the appropriateness of different types of third-party activity at different stages of a conflict. In addition to emphasising the importance of conflict context when determining a mediation strategy, they argue that ‘pure’ and ‘power’ approaches to mediation need not be adversarial, but can be complementary.Footnote 7
Drawing on Fisher and Keashly’s contingency approach, this article proposes a theoretically informed model to understand the applicability of coercive mediation strategies to identity-based conflicts such as the Israeli-Palestinian conflict, focusing in particular on the US as a mediator. While the US is a crucial actor, it has proved ‘unfailingly inept at launching successful initiatives’ in the realm of Israeli-Palestinian peace and has an extensive track record marked by failure.Footnote 8 Despite its leverage as a superpower and a major donor to both parties, the US has failed to exercise this coercive power effectively.Footnote 9 This poses important questions about the concepts of power, leverage, and coercion in the mediation of identity-based conflicts. First, what should the role of leverage be when negotiating an identity-based conflict? How and when should it be used? Nathan Thrall argues in favour of coercion as ‘the only language they understand’, but how appropriate is coercive mediation to solving identity-based conflict, where it is hard to reduce deeply cherished identity elements to bargaining chips?Footnote 10 As Oren Barak has argued, mediators have often erroneously treated the Israeli-Palestinian conflict as an inter-state conflict.Footnote 11 Thrall and other analysts of US mediation compare mediation of the Israeli-Palestinian conflict with that of the wider Arab-Israeli conflict, such as that between Israel and Egypt, but these parallels are problematic due to the difference in the nature of the conflicts. The relatively straightforward ‘land for peace’ concept that guided Israeli-Egyptian peace has proven impossible to translate from one conflict to the other.Footnote 12
Second, how does the power and leverage the US possesses affect its strategy and perception by the parties? Are US mediators able to employ alternative mediation strategies that do not centre around coercion? Heemsbergen and Siniver argue that the choice of what type of power to exercise and by extension what mediation strategy to pursue stems from ontological and epistemological understandings of the conflict and its causes.Footnote 13 This poses an important question: if strategy reflects a particular ontological and epistemological stance on the conflict, to what extent can one expect the same mediator to be adaptive and use different strategies effectively at different points within a set period of negotiations?
The article begins exploring these issues by establishing a theoretical framework to do with issues of mediation strategy, leverage, power, and coercion in the context of identity-based conflicts, and reflects on different theoretical approaches to US mediation of the conflict. It proposes a heuristic theoretical model, which illustrates the need to use particular strategies at certain stages of the negotiations. Initially, the parties may require a high level of coercion to overcome barriers to initiate dialogue, but coercion becomes less relevant once they have sat down at the negotiating table and a framework for negotiations exists. After extensive negotiations have narrowed the remaining areas of disagreement, significant leverage is likely to be required in order to finalise an agreement. In other words, leverage and coercion are key to the beginning and end of negotiations, but less intrusive strategies are crucial in between to ensure ownership of an agreement. Although this analysis centres on the Israeli-Palestinian conflict, the model and findings are relevant to other identity-based conflicts.
Existing empirical analysis of mediation of the Israeli-Palestinian conflict shows that the relationship between mediation strategy and effectiveness is far from linear, with less coercive mediation generating the groundbreaking Oslo Agreement and other important advancements in the peace process. This breakthrough also showed that using the right mediator and mediation strategy at the right time is an important determinant of success, and that multiple approaches are required either simultaneously or sequentially.Footnote 14 However, Oslo was not a permanent status agreement. It is important to better understand this relationship in the context of negotiating a permanent status agreement in order to increase the likelihood of successful mediation of a conflict, which continues to defy resolution.
To illustrate the relevance of the model empirically, this article then uses the Annapolis process, which included: 36 meetings between Israeli Prime Minister Ehud Olmert and Palestinian President Mahmoud Abbas (Abu Mazen) from December 2006 to mid-November 2008; the 2007 Annapolis conference; and the negotiations that followed led by then Israeli Foreign Minister Tzipi Livni and longtime Palestinian negotiator Ahmed Qurie (Abu Ala). This is a particularly interesting case for two reasons. First, the broad strategy pursued by Secretary of State Condoleezza Rice was, as she herself stressed, different from previous American approaches.Footnote 15 Kurtzer et al. have described it as ‘facilitation from afar’, generated as a compromise between the president’s limited view of a US role, scepticism from the rest of the neo-conservative administration, and the secretary’s own commitment to a deal.Footnote 16 Rather than coercing the parties into substantive trade-offs at the negotiating table, Rice was more focused on guiding and supporting the parties in their own negotiating process. This stood in stark contrast to the Clinton administration, who were intimately involved as a broker in practically every detail, including Clinton himself who devoted untold hours to the negotiations.Footnote 17 Second, the case contains elements of success and failure during different phases of the negotiations, which help to illustrate the model. While Rice successfully restarted final status negotiations in a challenging context through committed diplomacy, she failed to generate bridging proposals to achieve either a framework agreement on permanent status or a conflict ending agreement.
This analysis is based on author interviews with two former senior members of the US negotiating team who were directly involved throughout the entire process, and analysis of the leaked primary documents (transcripts of negotiation meetings and notes from the Palestinian Negotiations Support Unit) known as the ‘Palestine Papers’, in addition to existing secondary sources. Compared to previous periods like the Oslo process and the failed Camp David summit, this period has not yet been comprehensively analysed. While the George W. Bush administration’s policy towards the conflict has received warranted criticism from both within and outside, this tends to centre around a lack of willingness to engage coercively without considering in detail the type of engagement pursued or required, or when a certain form of engagement would have been most appropriate.Footnote 18 A micro-level empirical analysis of mediation strategies is necessary to better understand the lessons to be learned for effective mediation, and is a necessary complement to broader quantitative studies.Footnote 19
The merits and limits of coercion
In the context of mediation, coercion should not be viewed in absolute terms. It is a behaviour that is not inherently positive or negative, but one that may help mediators accomplish certain goals at certain points of a mediation process. However, it can also negatively affect a mediator’s relationship with the parties and possibly the outcome of mediation, depending on how and when it is employed. This section will consider the theoretical merits and limitations of mediation strategies that employ varying levels of coercion, and their applicability to different phases of the mediation process.
Mediation strategy is ‘an overall plan, approach, or method a mediator has for resolving a dispute … it is the way the mediator intends to manage the case, the parties, and the issue[s]’.Footnote 20 The choice of strategy by a mediator is a rational response to a specific conflict situation, taking into account the needs, interests, capabilities, and perceptions of the parties, as well as those of the mediator, and the relevance of various sources of power.Footnote 21 A prominent categorisation of mediator behaviour used by conflict resolution scholars like Jacob Bercovitch, Touval, Zartman, and others, identifies three main strategies that range from the least to the most coercive: communication-facilitation, formulation, and manipulation.Footnote 22
First, there is the role of ‘facilitator’, where the mediator serves primarily as a host for communication between the two parties. A facilitator can communicate messages between them if their relationship makes direct interaction impossible, arranges interactions between the parties, and gives them a secure space to clearly identify interests, issues, and points of contention or agreement to improve mutual understanding. There is, however, no substantive contribution to the negotiations themselves, and the mediator is not normally present at the table. This strategy represents ‘pure’ mediation.Footnote 23
Second, similar to the role of facilitator but more involved is the role of ‘formulator’. Here, the mediator exercises a greater degree of control over the meetings in terms of their frequency, pace, formality, protocol and procedures, and the physical setting. A formulator thus continues to develop the relationship between the parties, but can also make some substantive suggestions or proposals if they deem it appropriate or requested by the parties.
Third, there is the role of the ‘manipulator’. It is a coercive strategy, in that the mediator often uses material power and leverage to guide the negotiations, by for example pressuring the parties to be flexible, rewarding concessions by promising resources, or threatening to withdraw from the process. The mediator offers ‘carrots’ or ‘sticks’ to the parties; it can offer incentives to encourage cooperation and compliance, or punish intransigence and make clear what the cost of a non-agreement would be. Such a strategy allows the mediator to put forth proposed solutions, help change the parties’ expectations, and take responsibility for concessions that could otherwise be unpopular with respective domestic audiences. Another key function can be to verify the compliance of both sides with any agreement, thereby ensuring accountability and implementation. This strategy represents ‘power’ mediation or, as some refer to it, a directive strategy.Footnote 24
Clearly, some mediators are limited in the strategies they can pursue. Facilitation is commonly associated with influential individuals, non-governmental organisations, or small states who lack the material capability to coerce.Footnote 25 A manipulator requires the necessary resources – military, economic, or political – to generate leverage and thereby the capacity to affect bargaining positions and, potentially, mediation outcomes. However, as Marieke Kleiboer and Heemsbergen and Siniver argue, this strategy reflects a rationalist, positivist understanding of conflict and power, whereas power can also be understood in a subjective, post-positivist, less tangible way.Footnote 26 Reflecting a social-psychological understanding, Kleiboer argues that although we reflexively think of leverage in material terms, it can also be immaterial and rely on moral or psychological pressure.Footnote 27 Fen Osler Hampson and Zartman argue that the ability to intellectually convince the parties of the need for compromise is a key component, which can be done by a wide variety of actors.Footnote 28 Successful use of leverage, whether in terms of tangible or intangible resources, also depends on the willingness of the mediator to employ it against the parties, and the skill with which this is done; leverage alone is not a sufficient condition for achieving successful mediation outcomes.Footnote 29
Although a powerful state could in theory act as a facilitator, this can prove problematic due to its nature. The parties can come to expect a certain type of behaviour from a mediator due to the very possession of material power and leverage. As Thomas Princen argues, ‘the degree to which the United States can foster communication and trust – between itself and each of the parties and between the two parties – is highly constrained by the structural relationship between the third party and the disputants’.Footnote 30 Moreover, the disputants will develop expectations on the basis of previous interactions. Heemsbergen and Siniver argue that once a coercive mediator has exercised this power in negotiations, the parties will tend to respond favourably only when these resources are used again, and can come to expect them to be used. This makes it difficult to effectively transition from a more to a less coercive strategy.Footnote 31
Whether one accepts this argument or not, leverage and coercion nonetheless need to be considered within the temporal dimension of a peace process. Negotiation is a process that can be broadly conceived of in two stages: problem diagnosis and formulation.Footnote 32 If the parties’ positions are far from one another and their dialogue is not well developed, then the first two less intrusive strategies can help to bring the parties together to improve understanding, narrow down points of disagreement, and develop a relationship between the parties (or at least the negotiators) in order to foster trust. The latter function means this stage will often be required between individuals even if the issues in dispute are well known. Zartman explains that ‘the nature of the activity lies not in conducting the combined search for a/the solution but in arriving at and in convincing the other party to arrive at the conclusion that some joint solution is possible’.Footnote 33 Through this process, the non-negotiable can become negotiable (even though underlying basic differences may not be reconcilable) when conflicting parties learn to listen, understand, and empathise with the other party’s position, interests, and feelings, if these can be communicated effectively. Though not implying agreement, this can indicate an openness and responsiveness, which reduces hostility and defensiveness, allowing the other to be more open and responsive.Footnote 34
While this initial stage is sometimes referred to as ‘pre-negotiation’, this term can obscure the fact that mediation and negotiation can be required even to reach that initial stage.Footnote 35 Given the nature of this stage, there is an implicit assumption that facilitation or formulation are the central strategies, without acknowledging the fact that leverage and coercion might be required to enable the dialogue to take place. For example, Secretary of State James Baker ‘practically dragged’ an obstructionist Israeli Prime Minister Yitzhak Shamir to the 1991 Madrid Conference, which eventually paved the way for the Oslo Agreement.Footnote 36 Even if parties are willing to engage in indirect exploratory talks, a mediator who exercises some degree of coercion can provide welcome cover for what is often a controversial decision. The same is true of any concessions that will ultimately have to be made in order to secure an agreement.
However, the presence of a mediator at the table can also be a distraction. In negotiations between Israel and Syria in early 1996, Israeli negotiator Uri Savir noted that the Syrians were more interested in the American ‘mailman’ than their Israeli ‘suitors’.Footnote 37 Ultimately, it is the parties themselves who need to be the focal points of any understanding, and so their relationship should be paramount; they must feel a sense of ownership over any agreement if it is to be embraced and implemented. Indeed, Beardsley et al. found that the most durable agreements that secure a lasting peace are the ones reached with minimal external input, observing that ‘the more substantively intrusive styles of mediation – formulation and manipulation – may actually encourage a more passive stance regarding peace on the part of the adversaries’.Footnote 38 This is particularly relevant for identity-based conflicts or, to use a term coined by Azar, protracted social conflicts.Footnote 39 Such conflicts, which involve religious, cultural, or ethnic identity where basic needs such as security and communal recognition and developmental needs such as access to social institutions are not being met, may not lend themselves to coercive mediation.Footnote 40 In a situation where bargaining is not nearly as straightforward as in many inter-state conflicts, less intrusive strategies may be more appropriate than resource-based leverage to deliver concessions.Footnote 41
In their 2006 study of how different mediation strategies relate to crisis outcomes, Beardsley et al. argue that facilitation was preferable when resolving the parties’ commitment problems and reducing tensions, while manipulative strategy was the most effective in securing an agreement. They found that mediation is more effective when multiple strategies are used rather than just one, which reflects the importance of different strategies at different stages.Footnote 42 ‘The key, it seems, is to increase the incentives of agreement by structuring the costs of the conflict with a manipulative style while using the more integrative styles of facilitation and formulation to reach an agreement as consistent with the true distribution of capabilities as possible.’Footnote 43 This suggests that a more coercive strategy is appropriate towards the end of negotiations, although how to identify this point is notoriously difficult and only becomes clear after the fact. Process and sequencing remain vital, however. Zartman reflects that a proposed solution ‘which seeks to obviate the lengthy and inefficient process of negotiation … is unlikely to be welcomed in situations of sharp conflict over indivisibles or over non-tradables, situations where process ownership is important, or situations where the stakes include high political commitment.’Footnote 44
These theoretical considerations are summarised in the model proposed in Figure 1, which seeks to balance all the competing requirements. A highly coercive strategy like manipulation can initially be required to engage the parties in pre-negotiation, while these are normally mediated using less coercive process strategies. Once parties are at the negotiating table, these remain the most useful to ensure ownership of the talks. In order for an agreement to then be finalised on the basis of these talks, bridging proposals and a greater level of coercion is likely to be necessary.
Figure 1 A model of levels of coercion used over the course of negotiations.
A few caveats are in order. First, strategies may need to vary in response to immediate changes in context. A crisis such as the outbreak of violence, for example, may lead to a suspension of talks and suddenly require a manipulative strategy to resume them. The model does not deny the need for addressing such events, but illustrates the broader trends of mediation strategy usage. Second, mediation is a human activity, and as such is influenced by intangibles such as personalities. Qualities such as empathy and a sense of humour are often identified as key characteristics to building strong personal relationships, but these are difficult to account for in any model.Footnote 45 To provide further theoretical grounding for the case study, the article will now consider the particular characteristics of US engagement as a mediator in the Israeli-Palestinian conflict.
The US and the Israeli-Palestinian conflict
In the Israeli-Palestinian conflict, Israel, the stronger party, has traditionally always preferred bilateral negotiations, while the weaker Palestinians have always preferred the presence of a mediator.Footnote 46 The US is by far Israel’s most important ally, with whom they share a cultural and democratic affinity unlike any other in the region. Consequently, public opinion polls among Israeli Jews indicate that America is the international actor whose intervention in the conflict is considered the most desirable.Footnote 47 Despite this apparent bias, why have the Palestinians traditionally accepted the US as a mediator?
While perceived impartiality rather than utopian ‘neutrality’ is emphasised by a number of conflict resolution theoreticians,Footnote 48 Touval and others argue that ‘confidence’ in a mediator is the key desirable attribute, which can stem from a variety of different factors.Footnote 49 A mediator with strong links to one side can indeed jeopardise their suitability by appearing biased, but at the same time a certain degree of influence over one party can also be useful to the other if it means the third party can convince it to make concessions. Said party could potentially be assured that an ally would not suggest a solution detrimental to them, and would not allow them to suffer as a result of an agreement, and could so be swayed to yield on certain issues. Zartman and Rubin argue that ‘biased mediators … can be effective in assisting negotiations only if they deliver the party toward whom they are biased. In negotiation, external intervention rides the diplomatic equivalent of a Trojan horse.’Footnote 50
In an asymmetrical conflict, a mediator can thus serve to counterbalance the power of the stronger side.Footnote 51 As a result, even if originally negotiating from a position of weakness, a party may feel that a mediator who is traditionally biased towards the other may deliver an outcome preferable to the alternatives.Footnote 52 Moreover, the weaker party may feel that they cannot reject a powerful, biased mediator for fear of the consequences of saying no to such a superpower and moving them even closer to the opposing party. The potential of material rewards in terms of monetary aid, and the opportunity to improve relations with the superpower are also strong incentives to include them or accept them as mediators.
To understand the nature of US engagement in the Israeli-Palestinian conflict, Dennis Ross and David Makovsky suggest a typology of epistemological and ontological approaches to the conflict and its resolution, which has since been further developed by Thrall. First, there are the neoconservatives, who Thrall labels ‘Skeptics’. They argue that the Palestinians will not make peace with Israel, either because Palestinians completely refuse to accept the existence of a Jewish state, or because they will not acquiesce to the minimum terms acceptable to a majority of Israelis. Consequently, Skeptics consider negotiations pointless and do not believe the US should expend political capital on mediation, including brokering substantive compromises on key issues, and should certainly not coerce the Israelis in particular. Rather, any engagement should focus on easing the tensions on the ground caused by the occupation and improving Palestinian institutions.Footnote 53
Second, there are the political realists, who Thrall labels ‘Reproachers’. They argue that the US needs to take an active role, devise the parameters of a peace agreement to bridge the gaps between the parties, and pressure Israel into accepting them. As the more powerful party, Israel holds the cards and is the key to progress. Since the occupation has become an acceptable status quo for many Israelis, the US needs to show ‘tough love’ to compel them to act in their own best interests, and make the necessary concessions for peace.Footnote 54
Thrall also identifies a third category, the ‘Embracers’, who ‘combine the unconditional love of Israel of the Skeptics and the unwavering faith in the peace process of the Reproachers’.Footnote 55 They argue that the US is central to reaching a peace agreement, but that this must be done by embracing and reassuring Israel in order to give them the support they need to take the necessary risks for peace.Footnote 56
As noted earlier, the strength of a less or more coercive approach cannot be absolute but depends on context. While Ross and Makovsky make a case for adapting mediation strategy to suit the given context, they do not attempt to delineate which strategy is effective in which context or at which point in negotiations, nor do they reflect on the ability of the US to change strategies or execute them effectively.Footnote 57 The article will now go on to consider the mediation strategies employed by the US during the Annapolis process, 2007–08.
Pre-negotiation
After roughly a year of lobbying, Secretary of State Condoleezza Rice convinced a reluctant President Bush in July 2007 to endorse an international meeting of Israelis, Palestinians, and Arab states to restart permanent status negotiations. This was a significant change of policy. At the beginning of his first term, it quickly became clear that Bush had no desire to get embroiled in the Israeli-Palestinian conflict in the same way that President Clinton had; Afghanistan and particularly Iraq assumed centre stage on the US foreign policy agenda as part of the ‘war on terror’.Footnote 58 Although Bush became the first US president to officially endorse the idea of a Palestinian state,Footnote 59 he was unwilling to commit any political resources to achieving it. Diplomatic initiatives to address the conflict by Secretary of State Colin Powell like the creation of the Quartet (a multilateral body composed of representatives from the US, the EU, Russia, and the UN) and the Road Map were not given the requisite support by the White House to have any real impact.Footnote 60
This was a product of the Bush administration’s ontological and epistemological understanding of the conflict. It was viewed through the lens of the ‘war on terror’, and Prime Minister Ariel Sharon’s argument that they shared a common enemy in this regard resonated strongly after 9/11.Footnote 61 High-ranking neoconservative ‘Skeptics’, including Vice President Dick Cheney, Secretary of Defense Donald Rumsfeld, and National Security Advisor Elliott Abrams, were deeply sceptical that the Palestinians would ever be willing to make peace with Israel on terms the Israelis could accept. To them, the fundamental cause of the conflict continued to be the Arab world’s unwillingness to accept the legitimacy of Israel.Footnote 62 They explicitly vetoed any diplomatic initiative, were against putting pressure on Israel to make any concessions, and assiduously avoided using the term ‘peace process’.Footnote 63
Instead, the administration preferred to support Prime Minister Ariel Sharon’s unilateral withdrawal from the Gaza Strip, which was not coordinated with Abu Mazen to strengthen him against his Islamist rivals, Hamas.Footnote 64 They in turn gained power, prestige, and popular support at his expense due to an impression that armed resistance had prompted this withdrawal. Together with disillusionment with Fatah corruption and political failure, this helped produce a Hamas victory in the January 2006 Palestinian elections, which prompted a policy reconsideration. Abu Mazen and moderate Palestinian forces needed to be strengthened, particularly following Hamas’s June 2007 takeover of Gaza. Rice argued that the Palestinians needed a ‘political horizon’ in order to generate hope and positive momentum: ‘People are lost now in the Middle East and we need to act, she said; we need to make a big proposal.’Footnote 65 Furthermore, US credibility and reputation in the Arab world was at an all-time low as Iraq was unravelling and, following British Prime Minister Tony Blair and the State Department’s longstanding line of argument, serious commitment to the Israeli-Palestinian peace process could ameliorate their image.
Given years of violence during the second intifada and diplomatic stagnation, convening the Annapolis Conference at all was an accomplishment. Under difficult circumstances marked by a turbulent and often violent relationship between Fatah and Hamas, Rice and her team made eight trips to the region in 2007 prior to Annapolis to ensure Israeli and Palestinian commitment but also that of key Arab states like Egypt, Jordan, and Saudi Arabia. Ensuring wide participation and the presentation of a joint statement by the parties required considerable effort and constituted impressive diplomacy.Footnote 66 Although her overall mediation strategy was not centred around coercion, Rice did exert significant coercive influence during this period to resume negotiations, focusing predominantly on process and confidence-building rather than substantive content. Much like the 1991 Madrid Conference, Annapolis highlighted the continued importance of dedicated US diplomacy to bring the parties to the table.
As a starting point, Rice managed to convince the sceptical Olmert who had pledged to continue Sharon’s unilateralism through his hitkansut (‘withdrawing into oneself’) policy.Footnote 67 On 6 September 2006, Rice and her team had what Abrams describes as ‘one of the worst meetings ever with the Israelis’. The Road Map had been built on the idea of parallel phases of implementation of key commitments by each party, eventually leading to final status negotiations in Phase III.Footnote 68 The White House, however, overruled the State Department and concurred with Sharon that Israeli obligations such as redeployment from the West Bank and a settlement freeze were to be conditional on Palestinian security improvement and reform.Footnote 69 When Rice pushed the idea of parallelism rather than conditionality and simultaneously called for a jump to Phase III of the Road Map, Olmert’s advisors Tourgeman and Turbowitz objected strongly, arguing that this was a reversion to Clinton’s policy of continuing to negotiate while Palestinian violence continued.Footnote 70 Rice, however, persisted and insisted.Footnote 71
In mid-January and late March 2007, Rice pushed for a US role to supplement the bilateral talks that had begun with her encouragement between Olmert and Abu Mazen to give them greater seriousness and credibility among Arab states. The Israelis were dubious about this trilateral format, arguing that they did not want to end up negotiating with the US, but relented and agreed that Rice could join their meetings sporadically for a trilateral.Footnote 72 During this period, she ensured that Olmert and Abu Mazen continued to meet despite what Israel saw as the unacceptable (and brief) unity government between Hamas and Fatah agreed in Mecca.Footnote 73 In this way, Rice’s engagement was more akin to formulation in that she was exerting influence over the frequency and nature of the meetings. But she also continued to put pressure on the Israelis and pushed for confidence-building measures like prisoner releases to the point where Abrams complained that ‘the road to Annapolis was to be paved with Israeli concessions’.Footnote 74 In conversations with the Americans, Olmert emphasised that he wanted to reach a deal with the current leadership on all sides, remarking, ‘You don’t need to pressure me for that.’Footnote 75 As the more reluctant party, significant energy was expended to get them to Annapolis. However, the same was also true of the wider Arab world.Footnote 76
Up until the very last moment, there was ambiguity about the exact nature of Annapolis – whether it was going to be a crowning event to endorse a negotiated document, an actual negotiating session, or a launching pad for future negotiations – and who would be attending.Footnote 77 Official invitations were only issued a few days before the meeting. Abrams derides this period as one of Rice ‘substituting motion for progress’, reflecting a neoconservative lack of faith in negotiations, while Ross and Makovsky and Schiff argue that this is evidence of a poorly prepared event and a lack of strategic direction from Rice.Footnote 78 While this latter criticism appears warranted, there was some merit to this operative ambiguity; amid the uncertainty of what was possible, failing to meet an ambitious, publicly declared target would be far worse for the process.Footnote 79
Israelis and Palestinians fundamentally wanted the event to accomplish very different things; the Israelis wanted something broad that would relaunch permanent status negotiations, while the Palestinians wanted specifics on elements of a framework agreement on permanent status.Footnote 80 Attesting to these differences, Abrams recalls that in September 2007, just two months before the event, there was little enthusiasm for Annapolis on either side.Footnote 81 While their meetings were described as positive and making progress,Footnote 82 Olmert and Abu Mazen only delegated the negotiation of a joint understanding for Annapolis to their subordinate teams after their 1 October meeting. These negotiations went down to the wire, with repeated disagreement between the parties about the language it contained.Footnote 83 Their second meeting on 10 October showed the simultaneous ambition and limited parameters of the talks, in that they sought to include language on each of the final status issues but predominantly pave the way for negotiations after Annapolis. Tzipi Livni observed, ‘The key is to find two things: common ground, and a way to continue the process.’Footnote 84 By their seventh meeting on 12 November, it was clear that substantive issues were to be negotiated after Annapolis.Footnote 85 While US input on language was evident as the negotiations mention US drafts, none of this activity extended to coercing concessions from either party on substantive issues.Footnote 86 US officials did, however, on the morning before the beginning of the conference, have to put their foot down and threaten to cancel the whole event if the language in the joint statement was rejected.Footnote 87
Critics of Annapolis contend that the US should have pushed the parties harder and delineated a framework agreement to be endorsed at the event, much like the US could have done but failed to do prior to the Camp David summit in 2000.Footnote 88 Rice, however, had agreed that President Bush’s speech at Annapolis ‘should envelop the process and be encouraging, but it should not be directive’, which was in line with the neoconservative epistemological and ontological understanding of the conflict.Footnote 89 Such a move would have risked collapsing the negotiations prematurely at a point when they had yet to even agree to officially resume bilateral permanent status talks. Rice and her team looked to strike the inevitable balance between the ideal and the possible, erring on the side of caution to capture what had been accomplished up to that point and create momentum for the future.Footnote 90 According to a former Israeli official involved in the talks, Rice ‘was pretty nervous and happy to have any results’.Footnote 91 While the broad contours of a possible framework agreement are well known from previous talks, immediate acceptance of these without extended negotiations would have been politically very difficult for either leader to agree to and would have sought to obviate an important part of the process.Footnote 92 Olmert made it clear in late October that he had to avoid this and would resist any such imposition.Footnote 93
Bilateral negotiations
Significantly, Annapolis served as a launching pad for direct permanent status negotiations, which consisted of three tracks. The first track was between Olmert and Abu Mazen, who agreed to meet bi-weekly. The second track was a ‘steering committee’ of the two negotiating teams, led by Livni and Abu Ala, who agreed to begin meeting on 12 December 2007 and met regularly until September 2008.Footnote 94 The third track consisted of committees of subject experts to discuss the more technical matters of the final status issues. The most important of these, however, was the track between Olmert and Abu Mazen. Accounts of their discussions show that these negotiations between the two leaders were serious and in good faith, aiming to produce a framework agreement on permanent status to affirm the principles that would guide the more detailed solutions to the difficult issues like Jerusalem, refugees, borders, and security.Footnote 95
This was a positive change from previous patterns of negotiations in that the two leaders were focused on these issues from the beginning. A significant weakness of the Oslo process was that there was no clear ‘political horizon’ outlining a detailed two-state vision, and this weakness was being addressed at the highest level.Footnote 96 Moreover, they were talking directly in a way which was unprecedented, evidence of a ‘real relationship’ between the two leaders that had grown in pre-negotiations, allowing them to build trust and more clearly understand each other’s positions, constraints, and red lines.Footnote 97 Following the complete erosion of trust due to the violence of the second intifada, this was remarkable.
Kurtzer et al. write that, ‘Surprisingly, despite the weakness of American diplomacy, the negotiations did heat up after Annapolis.’Footnote 98 This disparaging view ignores the fact that a more facilitative strategy was appropriate at that stage. At Annapolis, the parties structured their own negotiating process and did not specify a US role at the negotiating table.Footnote 99 Palestinian negotiator Saeb Erekat seemed to emphasise the importance of Israelis and Palestinians retaining control of the process: ‘Annapolis is like a fire – if we control it, it will warm us and bring us energy[;] if not, it will burn us.’Footnote 100 Livni emphasised the centrality of the bilateral process and said that there could be no substitute for this.Footnote 101 Furthermore, the two leaders were committed to bilateral final status negotiations for different but complementary reasons, just as Rabin and Arafat before them, creating a mutually enticing opportunity.Footnote 102 Abu Mazen had long been the moderate vanguard of the PLO, committed to the two-state solution since the mid-1970s, and central to any negotiations under Arafat.Footnote 103 With Fatah continuing to lose ground to the ever stronger Islamist challenge, a peace agreement and deliverance of Palestinian statehood would turn the tables in their favour.
Olmert had come to the conclusion that a negotiated peace with the Palestinians was necessary for the long-term security of Israel. As an advocate of ‘Greater Israel’, during his ten-year tenure as mayor of Jerusalem from 1993 to 2003 Olmert subscribed to the view of the city as the indivisible capital of Israel, and supported settlement construction in the east of the city and the West Bank beyond. In later years, however, his stance became increasingly pragmatic, and in 2005 he left Likud and joined the centrist Kadima party.Footnote 104 In an interview with Yedioth Ahronoth (translated and reprinted in The New York Review of Books) following his resignation in 2008, Olmert acknowledged the need for territorial compromise, even in Jerusalem:
We must reach an agreement with the Palestinians, meaning a withdrawal from nearly all, if not all, of the [occupied] territories. Some percentage of these territories would remain in our hands, but we must give the Palestinians the same percentage [of territory elsewhere] – without this, there will be no peace. … Including Jerusalem – with, I’d imagine, special arrangements made for the Temple Mount and the holy/historical sites. Whoever talks seriously about security in Jerusalem … must be willing to relinquish parts of Jerusalem. … I was the first person who wanted to maintain Israeli control over the entire city. I confess. … For a significant portion of those years I wasn’t ready to contemplate the depth of this reality.Footnote 105
Predominantly, the US role involved troubleshooting and ensuring that the talks survived numerous episodes of violence in Gaza and elsewhere – Abu Mazen, for example, suspended negotiations in early March 2008, but Rice convinced him to resume them.Footnote 106 Once again, this veered into formulation rather than facilitation, which was also reflected in Rice’s insistence that she meet with the negotiating teams sporadically for a trilateral meeting (although she did not do this with the Olmert-Abu Mazen track). Former US consul general in Jerusalem Jake Walles argues that these were not negotiating sessions where Rice brought any new ideas or proposals, but chances for Rice to hear updates from the parties: ‘she would sometimes try to steer them in a certain direction, but it was not a very hands-on US role’. Still, Rice, ‘in effect, forced the parties to meet with her in a trilateral setting to talk about the negotiations’ when ‘they did not want US officials in the room or managing their agenda’.Footnote 107 Minutes of the negotiations show that although the US did not have a specific agreed role in the negotiations at the parties’ insistence, they found it difficult to say no and tended to accommodate Rice’s requests for meetings. As Livni put it to her Palestinian counterparts, ‘When they ask we obey but not as part of an understanding.’Footnote 108
Abrams recalls one such trilateral meeting with Rice, Livni, and Abu Ala during a trip to Jerusalem in June 2008, where the two lead negotiators ‘got into an argument that would have been avoided had we [the US] not been there. Both of them played to the US audience; there was no negotiating, and it was clear to me that there never would be while we were in the room.’ When Rice pushed for another trilateral, ‘both resisted, Livni with considerable energy. This was the only issue on which the Israelis and Palestinians truly saw eye to eye.’Footnote 109 He goes on to argue,
The serious negotiations are bilateral, and indeed in the case of Oslo were purely bilateral and kept secret from the United States. During the Bush years, efforts to insert the United States actually made bilateral Israeli-Palestinian negotiations harder. It is one thing to press the Israelis and Palestinians to negotiate but quite another to think that things will go more smoothly if we are physically present.Footnote 110
Historically, Abu Mazen certainly preferred a direct bilateral track, whether official or unofficial Track II, as did his negotiator Erekat who told a meeting of EU Heads of Mission, ‘Palestinians are better off sitting with Israelis alone. Personally speaking, I like to sit with Israelis alone. I have yet to see a third party play a role that is not biased to Israel.’Footnote 111 Olmert too shared this preference, believing that direct talks with Abu Mazen would be more productive without the presence of a mediator at the table, as did Israeli negotiator Udi Dekel, who thought that the Palestinians regressed with an American presence in the room, opting to ‘wait it out’ as they thought Rice’s position on the final status issues was closer to theirs.Footnote 112
In a departure from the Bush administration’s epistemological and ontological understanding of the conflict, at times the US role did appear to engage on substantive issues. In a bilateral meeting with the Palestinian delegation in Washington, DC on 16 July 2008, Rice spoke of securing certain territorial percentages with the Israelis, and that at the next trilateral she would present new ideas for a mechanism to address the issue of Palestinian refugees. She also reflected on the US role, which included ‘to step back and say what we think is possible’ and ‘try to bridge the gaps’.Footnote 113 However, as will be shown in the next section, Rice failed to do this authoritatively and effectively on any of the negotiating tracks.
Although the negotiations ultimately failed and significant gaps remained, multiple accounts attest to the progress made on issues including security and borders, specifically that the baseline for swaps would be the June 1967 borders.Footnote 114 This has been identified by a US official involved as ‘a major breakthrough’, a first in the history of negotiations and one of Rice’s biggest achievements.Footnote 115 Golan notes that Olmert’s assessments have shifted and varied over time, mostly in terms of apportioning blame for the failure to suit political interests, but his overall judgement appears to be very positive.Footnote 116 Other participants, however, claim that there was no explicit agreement on any of these issues. Olmert’s advisors Shalom Tourgeman and Yoram Turbowitz have both denied it, and Palestinian officials visiting Moscow in October 2008 reported to Arab ambassadors there that ‘there is no serious progress on any of the permanent status issues’.Footnote 117
Nonetheless, Olmert verbally outlined a significant proposal to Abu Mazen on 16 September 2008 and insisted that he sign an accompanying map then and there. Abu Mazen did not, but instead requested a copy of the map to study further, which Olmert refused. They agreed to a meeting of advisors the following day but the Palestinians cancelled and the meeting never took place.Footnote 118 Although this episode has become another staple in the Israeli narrative of Palestinian rejectionism, the reality is more complicated. Abu Mazen never rejected it or made a counter-offer, but rather, as Abrams, Kurtzer et al., and Thrall argue, had a number of questions about the proposal that were never answered. The proposal remained vague and there was inconsistency in the details, such as territorial percentages, that were not clarified.Footnote 119 Notwithstanding the fact that it was poorly handled, the very fact that Olmert reached this point and went further than all previous leaders, suggests that progress had been made and that there was some merit to the mediation strategy used.
Bridging proposals
While documents from the Palestinian Negotiations Support Unit (NSU) emphasise that the parties did not want bridging proposals, in a 2011 interview Abu Mazen suggested that US bridging proposals were required.Footnote 120 Indeed, at a certain point more coercive US mediation could have been welcome. During the 2008 negotiations, that point appears to have been some time between June 2008 and January 2009, when Bush was due to leave office.
Olmert had made Rice aware of his vision of a deal during her May 2008 visit to Jerusalem, and he gradually communicated this vision to Abu Mazen in their subsequent meetings, culminating in his proposal of 16 September.Footnote 121 Following one of their bilateral meetings in early August, a document entitled ‘President Bush Presents a “Trilateral Understanding”’ was drawn up based on their talks, and was to be presented during the first week of September. This seven-point document resembles a framework agreement on permanent status, in that it delineates the basic approaches to territory, Jerusalem, refugees, and other issues, although it remains (perhaps deliberately) vague in parts and leaves details to be negotiated.Footnote 122 While the author and origins of it are unclear, it suggests that the US was considering presenting a framework agreement to summarise the fruits of negotiations up until that point, possibly to the UN General Assembly in September.Footnote 123
Golan and Abrams both recount a somewhat confusing final instance at the end of 2008 where the US could have pushed to finalise a framework agreement. President Bush has explained that the idea was for Olmert to deposit his proposal with him in Washington, DC in November, Abu Mazen would then agree during his December visit, and a deal would be finalised in January. However, the Palestinians came to Washington, DC on 18 December and Abu Mazen did not agree to Olmert’s proposal. The Palestinians claim it was agreed that the negotiators would come back on 3 January 2009, to finalise a framework agreement. Olmert claims that he was never invited to such a meeting, and in any event, it was cancelled due to violence in Gaza.Footnote 124 An opportunity for intensive diplomatic engagement to formulate and mediate a bridging proposal based on Olmert’s offer was missed.Footnote 125
Bush has been described by a former senior American official as a president reluctant ‘to get involved in that kind of detailed, mind-boggling nuanced negotiation and diplomatic activity’, which fit the neoconservative ‘Skeptic’ mould.Footnote 126 This could in fact have been turned into a positive. Aaron D. Miller argues that Clinton’s commitment was such that it reached the point where his availability and presence became taken for granted. While this showed intense commitment, it also meant that a toughness, an ability to intimidate, and the credibility of threats to walk away in the face of intransigence were lost.Footnote 127 Although Bush may have accumulated goodwill and credibility, he was unwilling to use it to secure any type of agreement.
Contextual factors have been identified as the main reason for the failure of the negotiations. Although committed, both leaders were weak, and although this weakness may actually have strengthened their commitment, it did little to inspire confidence. Abu Mazen was politically weak, struggling to effectively confront the challenge from Hamas, and his popularity among Palestinians was low. Similarly, Olmert had been irreparably damaged by the poorly managed 2006 Lebanon War against Hizb’allah and had the lowest approval ratings of any Israeli Prime Minister in history.Footnote 128 Furthermore, corruption accusations and charges were being brought against him throughout 2008 and he was forced to effectively resign on 30 July, announcing that he would not stand in the Kadima party primaries. His official resignation came on 21 September 2008, and his mandate to negotiate as a caretaker prime minister was questionable.Footnote 129 Thus, on both sides there were serious doubts about whether the other would be able to deliver their respective publics and implement an agreement.
Moreover, division and political intrigue within the Israeli government led to mixed messages being sent to the Palestinians. Multiple sources report that from September 2008, Abu Mazen was being encouraged by Livni and former Prime Minister Ehud Barak not to sign anything with Olmert due to his weak standing, and that he should wait until Livni had achieved her ambition to become the new Israeli prime minister.Footnote 130 Even though the terms might not have been better – the Palestine Papers suggest that Livni was a tough negotiator, particularly on refugees,Footnote 131 and Olmert went further than she probably would have done – she would have had a better mandate to implement a deal. Political competition between Olmert and Livni is also likely to have informed Olmert’s preference for a separate negotiating track with Abu Mazen and the lack of synchronisation between that track and Livni’s negotiations.Footnote 132 Livni’s leadership bid for the Kadima party also meant that her negotiating posture changed, with Abu Ala complaining to Rice that she was not focused.Footnote 133 During an August negotiation session, she was unwilling to discuss Jerusalem, lest anything should leak and harm her campaign.Footnote 134
President Bush puts these political difficulties under the rubric of timing, identifying this as the key reason for failure: Abu Mazen ‘did not want to make an agreement with a Prime Minister on his way out of office’.Footnote 135 While reasonable, these contextual weaknesses could, in part, have been addressed by anchoring an agreement with the US and international organisations, making it a legacy for the next administrations, but this was not seriously pursued. Renewed conflict between Israel and Hamas in the Gaza Strip in December made this difficult, as diplomatic conflicts emerged over the pursuit of a Security Council resolution.Footnote 136 While these were hardly propitious conditions for the finalising of a framework agreement, an earlier initiative could have pre-empted this.
Conclusion
This article has proposed a new theoretical model for approaching levels of coercion in mediation of identity-based conflicts, illustrating it empirically using a set of negotiations in the Israeli-Palestinian conflict. As the model proposes and analysis of this case study has shown, engagement should not be uniformly constant but adapted to what the process and parties need at each stage, evidencing Beardsley et al.’s emphasis on the need for multiple strategies and Zartman’s emphasis on process. This is the primary weakness of the typologies identified by Ross and Makovsky and Thrall. Rather than arguing for any one type of approach, this model argues for a theoretically informed diplomacy, which shows adaptability by using different elements of each typology at different stages.
For all of the warranted criticism levelled at the second Bush administration’s policy, significant elements of Rice’s mediation strategy were theoretically sound and broadly fit the proposed model. Those who argue that the administration was not deeply enough engaged in the peace process obfuscate an important distinction between different types of engagement at different points of the negotiations. Following her coercive push for a resumption of final status negotiations, Rice’s more hands-off process strategy of ‘facilitation from afar’, which could sometimes more accurately be described as formulation, was appropriate for significant portions of the 2008 negotiations when you had committed leaders working seriously towards an agreement. This strategy was appropriate for both the Olmert-Abu Mazen and the Livni-Abu Ala tracks. This extended period of bilateral negotiations was critical, as this created a positive relationship between the two leaders, which enabled Olmert to make his offer and showed ownership of the concessions it entailed.
When bilateral negotiations seemed to reach an impasse in September 2008, a more substantive and coercive mediation strategy could have been pursued in order to clarify Olmert’s offer, generate bridging proposals, and finalise a framework agreement. However, the transition from a process strategy to a more coercive strategy was not made. Although there can be no certainty as to the outcome, such an approach would fit Beardsley et al.’s quantitative findings regarding the need for a more coercive, manipulative strategy to secure an agreement. US officials involved in this process and former officials who have analysed it have also advocated this approach.Footnote 137 Due to its power and relationship with both parties but particularly Israel, US leverage will be needed when finalising any agreement in order to create a less attractive alternative to a negotiated agreement. Israel has little incentive to take the risks involved when the fallback option of a largely cost-free occupation is perfectly acceptable.Footnote 138 Daniel Levy, a former Israeli negotiator, argues that leaders ‘need to be able to say, “Washington is holding my feet to the fire on this”’.Footnote 139 This pressure will need to be accompanied by US assurances and commitments on key issues like security.
The evidence casts doubt as to whether one mediator can be sufficiently adaptable to use the different strategies required. The Bush administration’s ontological and epistemological view of the conflict and its own role in it meant that the coercive mediation needed to bridge the remaining gaps, on substance as well as process, was beyond them. If indeed the US is required to play this coercive role, as both theory and evidence suggest, then perhaps it should not play the less coercive roles, which are not the best use of its energies and resources. Although Rice’s predominant blend of facilitation and formulation appeared fruitful, clear frictions were created by the presence at the table of a powerful actor such as the secretary of state. An alternative envoy, whether an individual, government, or non-governmental organisation, could perhaps have played the role of facilitator or formulator better, as was the case at several points throughout the Oslo process.Footnote 140 This type of multi-party mediation would have required a willingness from Rice to give up a certain measure of control over the process, and then been approved by the parties. It would have created an additional relationship to manage between different mediators and the parties, with implications for the security of communication and prevention of leaks, but it nonetheless remained an option that could have benefited the process.Footnote 141
While it is beyond the scope of this article to consider the strengths and weaknesses of the multi-party approach, it warrants further research in this conflict context. Could an alternative actor working in tandem with the US provide the benefits of the less coercive mediation strategies? Would this structure be acceptable to both parties? Beyond the Israeli-Palestinian context, it would be useful to apply these theoretical arguments to other identity-based conflicts in order to empirically examine the viability of the proposed model.
Some have argued that mediation as a conflict resolution strategy has exhausted itself in the Israeli-Palestinian context, and that it is time to try other strategies.Footnote 142 However, bizarre as it may sound after decades of mediation, this seems premature. The essence of the conflict may not change, but the individuals involved and the political context do, which may provide future opportunities. When those opportunities arise, the right constellation of strategies employed by the right actors is one crucial component of any possible success. Even though mediation continues to fail, one should not be too quick to dismiss repeated attempts as evidence of Einstein’s definition of insanity.
Acknowledgements
The author would like to thank Nina Caspersen, Liam Clegg, Marcus Schultzke, Asaf Siniver, and two anonymous reviewers for their helpful comments on previous drafts.
