Contents
-
-
-
-
-
-
Willingness to Accept the Costs and Risks of Mediation Willingness to Accept the Costs and Risks of Mediation
-
Supply-Side Factors Supply-Side Factors
-
Empirical Record Empirical Record
-
Illustrative Cases Illustrative Cases
-
Kissinger and Carter in the Middle East Kissinger and Carter in the Middle East
-
Carter in North Korea Carter in North Korea
-
Roosevelt at Portsmouth Roosevelt at Portsmouth
-
-
-
-
-
-
-
-
Cite
Abstract
This chapter examines the conditions under which mediation is most prevalent. The first two hypotheses are that mediation occurs when the costs of conflict are high and the shadow of the future is short. In such situations, the short-term benefits of mediation are extremely attractive, with any long-term risks being more of an afterthought. A third hypothesis is that mediation is more likely to occur when there are strong pressures from third parties. Results from the International Crisis Behavior (ICB) and Issue Correlates of War (ICOW) data confirm that third-party peacemaking is more likely during severe conflict in unstable or new democratic regimes and in democratic communities. It also appears, however, that mediation is less likely to occur when the third parties are too eager to intrude and shape the outcomes as peacemakers. The cases of Kissinger and Carter in the Middle East, Roosevelt at Portsmouth, and Carter in North Korea illustrate these important factors that shape preferences for mediation.
If it is true that the breakdown in diplomacy leads to war, it is also true that the breakdown of war leads to diplomacy.
To explore the impact of mediation on the conflict bargaining environment, we can, as I do in chapters 4 and 5, observe how the variation in conflict outcomes changes in the presence of third-party involvement. Alternatively, we can look at the conditions that give rise to mediation to understand the purpose for which the actors intend to use it. We can learn much about the intended purposes of a treatment by studying the symptoms that lead to its adoption.
By forming and testing our expectations about when actors find mediation in their interest, we can deal head on with the puzzle that arises when considering the long-term risks of mediation in conjunction with its frequent occurrence. If mediation is expected to do relatively poorly in the long run, why would combatants or third parties ever sign up for it? If there are both short-term benefits and long-term struggles associated with third-party involvement, then the mediation selection process should reflect that dilemma. It is doubtful that the actors involved are ignorant to the risks of third-party conflict management; therefore, when mediation occurs, the participants must have interests that supersede the potential for any resulting peace to be more fragile than what could be attained in the absence of outside involvement.
The argument I focus on in this chapter simply posits that any long-term instability is more worthwhile when the short-term benefits of peace are substantial. When the actors are desperate for a reprieve from fighting, even if only temporary, they will avail themselves of that option and accept the potential long-run risks of mediation. Moreover, a short-term break from conflict appears even more attractive when the future is heavily discounted, such that the benefits today are substantially preferable to the benefits tomorrow.
The supply-side incentives for mediation also become relevant. Third parties with strong interests in the outcome of a dispute or in the achievement of short-term peace can influence the viability of mediation. This can also help explain why mediation so frequently occurs even though the combatants are left with a less durable settlement as a consequence. The observable implications of these arguments can be quantitatively tested and applied to historical examples.
Aside from taking a first cut at assessing the dilemma that actors face in considering mediation, we gain an additional benefit from examining mediation selection, in that it allows studies of mediation outcomes (such as those in the following chapters) to address problems of causal inference that arise from the nonrandom assignment of mediation. Studying the effects of mediation without a firm understanding of when mediation occurs would be like studying the effects of a medical treatment without knowing the baseline characteristics of the population to which the treatment has been administered. The set of crises that lead to mediation is likely to be very different, in terms of the ex ante likelihood of mediation success, than the set of crises that does not lead to mediation or that leads to intervention of another type. For example, if mediators generally intervene in intractable situations, the conflicts that do not undergo mediation are likely to be more amenable to full resolution and mediation will appear to have less of a positive impact than it really does. This is a form of endogeneity bias or selection effect.1
Once the selection processes are understood, analysts can then carefully interpret the potential for biased inferences and can construct appropriate research designs to assess the relationships more accurately. Note that this is true of both quantitative and qualitative research. Both types of analyses must account for the nonrandom selection of mediation when forming causal inferences about its effects. The appendix at the end of the book provides more discussion of how selection effects are likely to apply to many of the common data sets used in the study of mediation.
Willingness to Accept the Costs and Risks of Mediation
Given the potential long-term risks of third-party involvement, we expect mediation to occur when the costs of conflict are high and the future is substantially discounted. In such situations, the short-term benefits are extremely attractive, with any long-term risks more of an afterthought. Mediation rarely occurs in the absence of violent conflict, that is, as a preventative resolution mechanism or between conflict spells. A common explanation is that the actors often need to feel the sting of bargaining failure before they turn to alternative dispute resolution mechanisms.2 When actors face costly and seemingly unending conflict, they are more likely to try other options, even if the alternative mechanisms are themselves costly and without guarantees of success. Relative to the costs and risks of severe conflict, mediation can appear cheap and nearly risk-free.
This explanation thus contends that the choice of mediation is one of desperation and relates to the expectation in the ripeness literature that mediation is more likely to occur during times of hurting stalemate. When the bilateral path has produced heavy costs, the combatants may feel that staying the course will surely bring more pain, whereas seeking third-party assistance at least will provide the chance of resolving the remaining bargaining barriers at less cost. If mediation fails to produce a stable peace, the disputants might not find it much worse than the nasty state of affairs prior to mediation. It becomes a worthwhile gamble. And even if mediation leads to long-term instability, the participants could actually be better off when conflict resumes because the third-party influence may curtail the severity of conflict, if not the likelihood, and the third party may have provided positive incentives as a part of the negotiation process.
It should not be assumed, however, that combatants seeking mediation want a full resolution of their costly conflict as an end in itself. Conflict is not something into which actors just accidentally fall; it is a chosen path that they take to maximize their payoffs in coercive diplomacy. They will have understood the risks of high conflict costs when refusing to settle at an earlier point in the bargaining process, and they will not want to pursue an alternative, less costly course of action if it offers no hope of more efficient bargaining. The argument here is that costly conflict makes other conflict management strategies relatively less costly or risky, but it is crucial that an alternative strategy such as mediation at least have some potential to decrease the barriers to bargaining that led to the onset of conflict in the first place.
Consistent with this view, high costs of conflict can lead disputants to want a limited cease-fire, during which they can explore their options for a full peace or renewed conflict. In other words, they may simply desire a breather during which negotiations can take place with some chance of success but with no delusions about the prospects for full resolution. In this case, participants recognize the costs associated with mediation and are willing to risk potential backsliding because it at least gives them a chance to make progress at the bargaining table, which has ostensibly not happened via bilateral bargaining. The negotiating parties may have also built the potential for such cease-fires into their decision calculus even before choosing to enter into conflict.
Hypothesis 3.1Mediation is more likely in disputes with high levels of severity.
The shadow of the future refers to how strongly actors discount future outcomes today. When the future is relatively meaningless, we say that an actor has a short shadow of the future, and this is precisely the type of situation that will make mediation more attractive if its benefits are primarily in the short run. With such preferences, actors could enjoy the immediate benefits of mediation even though they may eventually give way to greater instability further down the line.
One variable that can shape the shadow of the future is regime stability. In unstable regimes, leaders’ major concern is their own political survival. Transitioning regimes are adopting new institutions that they are unfamiliar with, and a small misstep may result in the current leaders being whisked out of office along with the failed institutions. Moreover, the presence of domestic instability typically indicates a high degree of contentious politics such that any decision of major importance, such as to fight or concede, will be hotly contested by opposition groups that want to portray the regime as weak or not acting in the national interest. Making decisions in the long-term interest of lasting peace is likely to be a low priority for such leaders because they will be unable to enjoy the downstream benefits of a policy if domestic instability persists and they soon lose office. As one example of such eagerness to achieve the short-term stabilizing benefits of mediation during times of desperation, in early 2001 Joseph Kabila, president of the Democratic Repbulic of the Congo (DRC), immediately accepted the prospect of talks in Washington and sought an end to the ongoing war enveloping the Great Lakes region of Africa after replacing his assassinated father as president. Options that have a high probability of short-term gains, particularly when accompanied by a period of stability, will thus be highly regarded by leaders in unstable regimes even if there is a low probability of sustained benefits.3
A similar logic applies to leaders early in their tenure. New leaders have less ability to think in terms of long-term benefits and, instead, need to focus on consolidating power. Many new leaders are new precisely because general instabilities exist that preclude the ability to plan far in advance. Although revolutions and coups are obvious cases of instability, even new leaders that accede to power because of the collapse of a coalition government in a parliamentary system can experience domestic instability that will make immediate results more salient than long-term ones. Other new leaders may come to power via normal periodic transitions, as in many presidential systems. New leaders in such cases may not be symptomatic of instability per se, but they are still prone to have short shadows of the future. For instance, they will want to capitalize on elevated political capital—which can be lost quickly as U.S. President Barack Obama and French President Nicolas Sarkozy have recently discovered—following an electoral victory before the opposition parties can strengthen. As an example of new leadership injecting a sense of desperation for results, the peace process in Northern Ireland moved past a critical impasse in 1997 in part because both the United Kingdom and Ireland experienced leadership transitions as Tony Blair and Bertie Ahern became the new premiers and immediately pushed for progress in the talks (Chastelain 1999). The point is that new leaders have a greater desire for immediate results, with less emphasis on outcomes further down the line or on bolstering their legacies.
The idea that regime stability and leadership tenure shape a leader’s discounting of the future and thus his or her preference for mediation relates to the saving-face function of third-party involvement. The short shadow of the future generated by new and unstable leadership leads not only to a general preference for short-term benefits but, much more specifically, to a need for immediate political cover. Notice that mediators may be sought to provide such short-term political cover both sincerely and insincerely. When a leader is unable or unwilling to make the concessions necessary to end a conflict because of domestic opposition, then he or she is likely to sincerely seek third-party assistance. Existing research has found that new leaders are more prone to make concessions,4 which means we expect new leaders to have a strong need for short-term political cover that makes the concessions palatable at home. To the extent that a mediator can provide political cover for concessions—through owning responsibility or signaling the prudence of concessions—this preserves the type of short-term political capital that new leaders and leaders in otherwise unstable regimes covet.
New and unstable leaders, however, may also face the opposite problem—having a domestic audience that is overly eager to end a conflict when the leader believes he or she will benefit more from fighting than conceding. Because leaders with a tenuous hold on power are less able to bear the political costs of fighting an unpopular war, they may seek mediation insincerely for a different type of political cover. In this case, the leader will try to go through the motions of a mediated peace process to convince the opposition at home that compromise is earnestly being pursued when, in actuality, the leader fully intends to remain intransigent at the negotiation table and return to conflict while blaming the other side for being unreasonable. For example, as we explore later, the DPRK is often suspected of pursuing mediation solely to play this part to domestic and international audiences, and Israeli and Palestinian leaders frequently accuse one another of the same. Opponents and third parties should be hesitant to pursue mediation when they suspect an actor of having such incentives, but uncertainty of the motives can allow for such imperiled mediation to occur nonetheless. Moreover, skeptical participants face a difficult choice between dealing with a negotiating partner potentially acting in bad faith and giving that actor the ability to use any refusals as evidence that it is not the side responsible for the ongoing hostilities.
Both types of political cover, because they merely help the leaders stay in power, meet more of a short-term need. So, leaders with a high demand for political cover will care less about conflict management strategies that primarily accrue in the long run. Their immediate needs of political survival will trump the desire for long-term gains.
Hypothesis 3.2Mediation is more likely in disputes involving unstable regimes or new leaders.
One implication of the political-cover logic is that the need to save face should be stronger for the challengers in a dispute than for the defenders of the status quo. Challengers are the actors that most clearly have chosen to enter into a dispute, and they must justify that choice to their publics, which can make backing down costly. The act of initiating a dispute signals to domestic audiences that the leader is optimistic about the chances of revising the status quo; however, then subsequently backing down can indicate incompetence and otherwise go against the audiences’ own perception of what an acceptable agreement looks like as informed by the leader’s initial justification (Bueno de Mesquita, Siverson and Woller 1992). Defenders, on the other hand, do not generally choose which disputes are brought against them and have fewer political constraints in making concessions. Because the audience costs of concessions are typically higher for challengers, we expect the challengers’ need for political cover to be greater. This logic is consistent with the argument from existing studies that democratic leaders are especially more selective in choosing when to initiate a conflict because they face stronger political sanctions at home if they challenge and lose.5 Following this argument, we can test whether regime stability and leadership tenure of just the challenging states do well in explaining the involvement of third parties in disputes.
Supply-Side Factors
As mentioned in the previous chapter, third parties can have a variety of selective incentives for serving as mediators. In particular, some can benefit from attenuating the conflict spillover costs, reducing a humanitarian crisis, or shaping the outcome in their favor. Even when third parties anticipate the long-term trade-off involved in mediation, they can still find serving as a conflict manager worthwhile. This is especially true if they primarily benefit—strategically, politically, economically, or morally—from helping to resolve the conflict immediately but do not bear many of the costs when the conflict recurs. Saadia Touval actually advocates that third parties discount the future and intervene forcefully whenever possible: “I believe that mediators ought to give greater weight to the likely near-term consequences of their choices because predictions of the near term are generally more reliable than those of the more distant future. Mediators can be certain that an ongoing war will produce casualties. Less certain is the proposition that ceasefires tend to break down, leading to the renewal of war and causing higher casualties in the long term” (2002, 183).
When outside actors will benefit from reducing negative externalities, have humanitarian interests, or have preferences about the final allocation of the good or issue under dispute, they can lobby the combatants to let them mediate.6 Even when the combatants are hesitant about mediation because of the costs or risks involved, third parties may be able to compel the disputants to undergo assisted negotiations. Such enticements may include some form of tangible aid or threat, but probably they more often involve less tangible promises of improved relations and preferential treatment that can be consequential, especially if the third party is an important player in the international system. For example, the offer of U.S. mediation in Honduras following the June 2009 coup presumably was accepted because the key players in Honduras recognized that maintaining favorable relations with the United States was important and that the United States had strong interests in the outcome related to both its desire to strengthen democracy in Latin America and to stand firm against Venezuelan meddling. Also, the pressure to mediate does not have to come from the would-be mediator per se if interested third parties are content to have another actor broker a deal. For example, from 1967 to 1969, U.S. inducements in the form of arms assurances provided an incentive for Israel to agree to participate in the negotiations under the auspices of Gunnar Jarring, special representative of the UN (Touval 1982, 149).
Conversely, an absence of interest by third parties in providing mediation can lead to a free-rider problem and an under-provision of mediation even when the combatants desire it. Either way, it should be the case that mediation is positively associated with third-party interest. Heightened outside-actor interest in conflict management should increase the propensity for mediation to occur while indifference should decrease the propensity.
Hypothesis 3.3Mediation is more likely in disputes with a high degree of salience to potential third parties.
Empirical Record
To assess these hypotheses, I use quantitative methods to explore aggregate relationships while accounting for potentially confounding variables. This statistical approach not only allows the analyst to test whether the implications drawn from the arguments are actually observed in the real world; it also allows us to see the substantive effects of the relationships. The purpose, then, is not just to test if there is some relationship between two variables in the observed data but much more than that. The methods are useful in understanding the magnitudes of the relationships. For this reason, my focus here is on the substantive interpretations of the findings.
I next present a brief overview of the data and variables. More information regarding the coding of these variables and the model specifications, as well as the full regression results, are presented in the appendix at the end of the book.
Two data sets are used: the International Crisis Behavior (ICB) data and the Issue Correlates of War (ICOW) data (version 1.1, created by Paul Hensel and Sara Mitchell). The ICB data are useful because they cover the entire globe from 1918 to 2006 and include cases in which militarized violence is a perceived possibility but does not necessarily occur—its threshold is neither too restrictive to only include conflicts that have escalated nor too relaxed to include conflicts with no potential for escalation. Of the 729 crisis dyads in the data, 170, just over 23 percent, experienced mediation.7
Because the ICB data contain only one observation per crisis dyad, and thus are unable to provide information about dynamic third-party activity within a crisis, I also use the ICOW data. These data contain information on territorial claims in the Western Hemisphere and Western Europe from 1816 to 2001; river claims in the Western Hemisphere, Western Europe, and Middle East from 1900 to 2001; and maritime claims in the Western Hemisphere and Western Europe from 1900 to 2001. The principal justification for using the ICOW data is that they allow the analyst an opportunity to assess each conflict management event over the course of a claim at different points in a conflict. Using an event-history approach with these data, it is possible to model not only whether third-party conflict management occurs but also when it occurs. Factors related to conflict severity, a short shadow of the future, and third-party pressure that are expected to have a positive impact on the incidence of outside involvement should also decrease the time until conflict management occurs. Note that some of the third-party conflict management attempts in the ICOW data involve arbitration and adjudication. Although the results are robust to a competing-risks setup that separates the occurrence of mediation from the occurrence of arbitration and adjudication, for simplicity of interpretation, the results presented here relate to models of the occurrence of general third-party conflict management.
To capture conflict costs, the analyses with the ICB data use four variables: an indicator of violence level, the length of the crisis (log transformed), the presence of an ethnic dimension in the conflict, and the latent military capability ratio (log transformed) that the strongest actor in a dyad has. A four-point indicator of violence most directly measures conflict costs. Crisis length and the ethnic conflict indicators primarily capture the difficulty of resolving the crisis peacefully, whereas the capability ratio measure captures the potential for the crisis to be mutually hurting. The ICOW analyses similarly use the number of fatal militarized interstate disputes (MIDs) that have occurred in a dyad over an issue in the current year, the number of fatal MIDs that have occurred in a dyad over an issue in the past five years, and the natural log of the capability ratio within the dyad.8
The models run with the ICB data use three variables to account for the potential interests of third parties in mediating. The first variable counts the number of neighbors of each actor in a dyad and sums them together to get a total count of the number of neighbors.9 The logic is that having more neighbors implies having more actors that are potentially adversely affected by spillover or that have particular interests in the outcome of the dispute. When conflicts occur in relative isolation, there should be fewer third parties interested in becoming involved as a conflict manager. The second variable counts the number of democratic neighbors. Sara Mitchell (2002) has argued that there is a higher preference for third-party conflict management within democratic communities.10 In addition, domestic political incentives to intervene in salient neighboring conflicts will weigh more heavily with democratic leaders, who will also tend to be more active in international organizations that frequently supply mediation. A third variable uses a five-point measure of the salience of a crisis to the international system, with low values indicating isolated regional relevance and higher values indicating widespread ramifications. The ICOW analyses also use the number of overall neighbors and the number of democratic neighbors as indicators of third-party incentives to become involved, although an analogous geostrategic salience variable does not exist for these data.
To measure instability, the number of years since a substantial change in a regime’s institutions is taken as one indicator. The number of days (log transformed) since executive leadership turnover is taken as another measure of the extent to which the future is discounted. Some models use the averages of these measures to characterize the stability and leadership tenure of the actors within each dyad—because the choice of mediation must be mutual, both sides in a dyad have to see some need for it. To demonstrate robustness, other models use the measures that pertain only to the challenger, which is, recall, the actor most likely to need political cover for backing down or making concessions.11
To get an even better grasp on the role of instability, it is useful to isolate the set of democratic dyads to explore how the effect of instability is conditioned by regime type. That is, democratic leaders who are early in their tenure or in unstable polities should perceive an even shorter shadow of the future than authoritarian leaders. Authoritarian leaders often can anticipate being in power indefinitely, whereas democratic leaders face strict institutional constraints. Moreover, democratic leaders, by the nature of their greater accountability, should need political cover more for both conceding and remaining in conflict when there are substantial portions of the polity that object to such important
decisions. For these reasons, we expect the tenure and regime durability variables to have an even stronger effect on mediation incidence when democracies are involved. Democratic leaders with a short tenure or with a tenuous hold on power should be the most likely to seek mediation, whereas authoritarian leaders with a long tenure and in a stable regime should be the least likely.
The results generally confirm our expectations. Figure 3.1 presents the substantive effects of the results using the ICB data. These are relative risks of mediation occurrence, calculated as the difference in the predicted probability of mediation when moving the independent variable as indicated, divided by the baseline predicted probability.12 Interval-level and ordinal variables are increased by two standard deviations (2SD) and dichotomous variables are increased from 0 to 1.13 The shaded bars indicate statistical significance, and empty ones indicate statistically insignificant relationships.
All the variables indicating severity and costs of conflict are statistically significant and in the expected directions. Mediation is more likely in a violent war, when there is ethnic conflict, the crisis is long, and there is power parity. When actors are relatively desperate for a reprieve, they are more willing to accept the costs and risks associated with mediation. Many of these relationships are also substantively meaningful. Mediation is almost three times more likely in ethnic conflicts, twice as likely when the natural logarithm of the crisis duration increases by 2SD, and less than half as likely when the indicator of capability imbalance increases by 2SD.
Regime durability has an unexpected positive relationship with mediation, and the tenure variable is statistically insignificant. But so far we have not accounted for the potential conditioning effect of regime type. Leaders in democracies should respond differently to issues of instability and tenure than other leaders. The results in figures 3.2 and 3.4 later in the chapter take into account the potential conditioning effect of regime type.
The evidence does not demonstrate as much support for our supply-side expectations. The total number of neighbors and the geostrategic salience have a substantial negative correlation with mediation, and the number of democratic neighbors is not statistically significant. An increase by 2SD (rounded) in the number of neighbors of a crisis dyad—which is an increase by thirteen—decreases the probability of mediation by almost half, and an increase in geostrategic salience from its median to its maximum decreases the propensity for mediation by even more. These findings are interesting because they suggest that when neighboring and powerful states in the international system have strong interests in a crisis, the disputants actually resist mediation. On further consideration, this is understandable because many disputants prefer to limit meddling by external actors wishing to shape the regional and global orders in their favor. As we see later, more intrusive interventions tend to produce greater risks of long-term instability, and so disputants have reason to be especially hesitant about allowing very eager third parties to become involved. When outside interests are too strong, the disputants are wary about third-party conflict management. Indeed, many of the crises with high geostrategic salience and many neighbors involve great powers, which are best able to keep external actors from interfering. Presumably for this reason, mediation rarely occurs in conflicts between great powers.
Figure 3.2 presents the relative risks of mediation from a model that is restricted to the set of democratic dyads. Regime durability and tenure, as expected, are both negatively correlated with outside involvement when the dyads are restricted to joint democracies. Mediation is more likely among relatively unstable and new democratic regimes. These findings corroborate the argument that actors find mediation most attractive when they strongly discount the future. In competitive
democracies, new leaders and those governing through infirm institutions have relatively short shadows of the future because of the need to consolidate power, the restrictions of term limits, and the needs for political cover from any unpopular foreign policy decisions. It is worth noting that the regime stability variable has much more of a substantive impact than the tenure variable: an increase by 2SD in regime durability decreases the probability of mediation by 78 percent, whereas a 2SD increase in the measure of leadership tenure decreases the probability by just 22 percent. This should not be surprising because regime stability is a more direct measure of how precarious a leader’s political standing is, and many new democratic leaders could conceivably feel quite comfortable if they recently received a strong electoral mandate.
Using the ICOW data, figure 3.3 displays the relative risks of outside involvement when the independent variables increase by 1 unit for the fatal MID counts and by 2SD for the other variables.14 The results presented are from two different models. One model measures tenure and durability using the averages across the actors in each dyad; this model is the basis for most of the results reported. The other specification, which produces robust results as seen in the appendix, measures tenure and durability using the values from the challenging state. Figure 3.3 indicates that the risk of experiencing third-party conflict management increases by 200 percent (i.e., it triples) when there is a fatal MID in the current year of the dispute and that it increases by 48 percent when there has been a
fatal MID in the previous five years. These findings confirm the expectation that actors are more desperate for at least a short-term reprieve when they are suffering heavy conflict costs. In such instances, the immediate needs of each actor to reduce costs become stronger than long-term concerns about whether any lulls in hostilities will only be temporary, intentionally or not. Interestingly, the capability ratio variable is not statistically significant when the ICOW data are used.15 One potential explanation for this is that this indicator is not as useful of a measure of a hurting stalemate in disputes that are predominantly low level and not militarized.
Even though the measure of the total number of neighbors is statistically insignificant, the results show that disputes in democratic communities are more prone to experience outside assistance; this effect is both statistically significant and substantively meaningful. An increase by nine democratic neighbors for a dyad (four or five for each actor) increases the probability of third-party conflict management by 52 percent. In democratic communities such as Europe and, more recently, the Americas, there is stronger pressure for third-party conflict
management. This at least provides some indication that neighboring interests can have a positive effect on the propensity for outside involvement, not just a negative one. In light of the discussion about hesitation by combatants to involve overly eager third parties, it is likely that disputants in democratic neighborhoods are less threatened by the potential meddling of other democracies and are more likely to acquiesce to their advances, especially if the democratic neighbors work through international organizations.
To better assess the role of political instability using the ICOW data, figure 3.4 again restricts the sample to just the democratic states and focuses on the tenure and durability variables.16 The results confirm our expectations—both relationships are negative and statistically significant in each model specification. We observe that outside involvement is less likely as stability—measured by both tenure and institutional durability—increases among democratic disputants. Moreover, the effects are quite strong because an increase by 2SD in three of the measures leads to a decrease of well over 50 percent in the risk of outside involvement. That is, short shadows of the future and greater needs for political cover correlate with more third-party conflict management, whereas better political stability correlates with less. It is interesting that the relative magnitudes differ between the ICOW and ICB analyses. When we use the ICB data, regime durability appears to have a much starker impact than leadership tenure. One explanation may relate to existing findings, which show that vulnerable leaders need political cover less in more severe conflicts (ICB crises tend to be more severe than ICOW disputes).17
To sum up the findings from the quantitative analysis, we find that, first, mediation is more likely to occur when the actors face costly and difficult conflict situations, as when the fighting is intense, has occurred for a long time, involves ethnic dimensions, and is between actors of similar strength. Second, democratic states with domestic instability and new leaders are more likely to seek mediation. These findings indicate that actors tend toward mediation when they are desperate for peace or otherwise discount the future greatly in favor of short-term political gains. We thus begin to get a sense of how states face a trade-off when considering mediation and are much more willing to take on the long-term risks when the immediate cessation of hostilities is highly attractive. Third, it appears that supply-side factors do matter, but not necessarily in ways initially expected. We see some evidence that democratic communities can encourage the practice but that disputants may actually resist mediation by heavily interested third parties.
Illustrative Cases
Here I introduce three illustrative examples in which the logic just discussed and tested has played out in practice. In subsequent chapters, I will continue to glean information from these cases. Although each of these cases did experience mediation, we see in chapter 5 that they had different outcomes with regards to the long-term stability of their mediated peace; we then focus on key factors that maximize or minimize the trade-off in mediation outcomes. For now, I present these cases as examples of the various mechanisms that drive the choice for costly and risky mediation.
Kissinger and Carter in the Middle East
Although Henry Kissinger and Jimmy Carter had different approaches to peace in the Middle East, I treat their efforts together here because the final EgyptianIsraeli peace treaty cannot be understood without following the peace process from 1973 to 1979. This case demonstrates how the high costs of conflict, need for political cover, and third-party pressure can shape the incentives for mediation to occur. Because of these incentives, Israel and Egypt were willing to accept the risks from third-party involvement.
Conflict costs are the first important factor driving the incentives for mediation. As the hostilities in the 1973 October War wound down, the immediate aims of Egypt centered on protecting the Third Army from complete collapse, recovering the Sinai Peninsula, and resolving issues related to the occupied territories and the Palestinian question.18 The need to save the Third Army, and generally to avoid further military losses, made its initial desperation for negotiations quite high. Even though Sadat claimed to have achieved his goals in the October War, he was keenly aware of the burdens of the war, and this motivated him in his peace initiative.
Why did I think we could achieve so much through peace? By a simple calculation: how much war had cost Egypt and the Arab world since 1948. Until the October War, 99 percent of the economic burden was borne by Egypt. Even after the October War, when the entire Arab world made a lot of money out of oil and added to their wealth, Egypt by contrast was drained of its resources. So whenever the Israelis created problems during the peace negotiations, my thoughts would go back to the burden we had to bear, and I would opt for peace.
I also thought of the direct results of the October War. What did that war achieve for us? We regained a very small portion of the Sinai and we managed to reopen the Suez Canal. Against this we have to set the cost to Egypt of 14 billion pounds, plus all the losses in men and equipment. (Sadat 1984, 106–7)19
The Egyptian economy was heading toward ruin because, in part, it bore the greatest security burden in the protracted conflict between the Arab states and the Israelis. It was also becoming clear that continual conflict with Israel was not ever likely to yield much of a payoff, given that the strategic advantage for Egypt at the start of the 1973 war was potentially as good as could be expected and yet the result was, objectively, closer to defeat than to victory.
Israel too had high immediate costs of conflict, even as it encircled the Third Army. The Israeli military suffered heavy casualties at the start of the war, and the people became desperate for peace as the war unfolded in the midst of a stalled economy.20 Kissinger (1982, 748–49) recognized the growing support for a settlement on disengagement, noting that Israel would have needed to maintain the deployment of several divisions to continue holding the territory gained on the west bank of the Suez and that Israel needed to find a way to recover its prisoners of war sooner rather than later.
The mounting price of war, not only from the immediate fighting in October 1973 but also from the toll of a quarter century of nearly constant hostilities, thus helps explain why Israel and Egypt embarked on a disengagement process with Kissinger at the helm. Both sides were relatively desperate to try something new and would not have thought that the costs and risks of mediation were all that significant in relation to the burdens of the status quo. Both sides, and especially Egypt, were also eager to tap more into U.S. technology and aid as a means to bolster their economies, which would further increase the attractiveness of pursuing a peace process under Kissinger’s auspices.21 The rising desperation on both sides was not lost on Kissinger, who claims, “From the outset, I was determined to use the war to start a peace process” (1982, 468).
The costs of conflict were certainly salient to the disengagement talks in the immediate wake of the October War, but they also were influential in later initiatives that included the Camp David summit.22 In that regard, as bilateral talks stalemated after the initial promise of Sadat’s bold trip to Jerusalem in 1977, the risk of renewed conflict resurfaced as impatience grew. Moreover, Egypt had been spending substantial amounts on its security forces during the lull in hostilities, with over half of its GNP going to military expenditures in 1975 (Telhami 1990, 69; Baker 1978, 130). With such high security expenses, the costs of nonagreement were substantial enough to keep the parties eager for settlement and more willing to bear the risks and costs of mediation.
Another factor motivating mediation throughout the step-by-step and the Camp David peace processes was the importance of political cover for the leaders in the midst of fragile domestic and international political situations. Aaron David Miller notes that the leaderships on both sides of the Arab-Israeli conflict have been “held accountable 24/7 by domestic politics, public opinion, and their own conception of how the terms of any agreement would shape their current political situation and place in history” (2008, 37). Negotiating with Israel, widely viewed by Arabs at the time as an illegitimate state, was politically costly domestically and regionally for Arab leaders. Sadat recognized that if he could not go all the way with Israel to reach a comprehensive settlement that resolved the issues related to the Palestinians and the occupied territories, he would certainly need substantial political cover to make an Egyptian-Israeli treaty viable.23 Moreover, both Israel and Egypt claimed victory after the October War, so it was politically delicate for either side to consider concessions that would be necessary for a peace agreement.
More specific to the Israeli leaders’ need for political cover, this period saw the end of the relatively firm control of power by the Labor Party and can be characterized as politically unstable (i.e., not conducive to a long shadow of the future). Public support for government policies on salient issues such as national security had to be carefully nurtured and protected. The Israeli domestic public was closely attuned to the U.S. leaders’ stance toward Israel because the fate of Israel had been so intertwined with its superpower ally. This actually created competing incentives for pursuing U.S. mediation. On the one hand, Israeli leaders could win political support when making decisions that improved U.S.-Israeli relations. On the other hand, by giving the United States a prominent role in the negotiations, the leaders faced a risk of stirring domestic concerns whenever they ran afoul of U.S. interests. In fact, Yitzhak Rabin (1979, 300) faults the aggressive posture of the Carter administration toward him as the key reason why the Labor Party lost control of parliament in 1977.24 The ability for a third party to hold sway with a domestic public can offer the opportunity for saving face, but it also can increase the risk of losing face when preferences differ from those of the third party.
The importance of political cover can be seen through comparing these mediation initiatives to other conflict management attempts. An interesting facet of the Arab-Israeli conflict during this time period is that we do observe some of the alternatives to mediation that were considered, and one reason why they failed to make progress was the abundance of rampant political posturing in the absence of political cover. That is, Sadat and Begin tried a bilateral initiative that began with lower-level negotiations through secret Moroccan and Rumanian channels and that became public with Sadat’s speech to the Knesset in 1977. But the bilateral talks soon bogged down, in part because neither side could make the moves necessary for progress when under constant domestic and international scrutiny.25 In addition, the parties tried the Geneva Conference, initially created as a charade of sorts. One problem with the Geneva Conference, as witnessed in the opening statements on December 21, 1973, was that the parties had strong incentives to grandstand and show off to their domestic and international audiences instead of negotiating sincerely with one another. Kissinger (1982, 794) writes, “Each of the contending parties could sustain its experiment with peace only by proving its constant vigilance to the hard-liners back home.” Another problem, and ultimately what doomed a return to the conference later, was that the representation of the Palestinian Arabs became a political dilemma: the Israelis would lose substantial political capital if there was a formal Palestinian delegation because this would imply recognition of the Palestine Liberation Organization (PLO) as a legitimate entity; on the other side, the Arab states would bear immense criticism if they consented to the exclusion of a Palestinian delegation and tried to speak on their behalf.26
Through mediation, especially the type characterized by Kissinger’s shuttle diplomacy and Carter’s secluded summit, the greater public was left in the dark about the content of the negotiations and thus could not decipher who offered which concessions and whether the Palestinian Arabs were getting a fair shake or not. In contrast to the bilateral initiative and the Geneva Conference, negotiations under the auspices of Kissinger and Carter could more easily proceed and deal with very difficult substantive issues instead of being sidetracked by political posturing. Mediation in these cases was thus motivated in part by actors confronting dicey political situations and needing to save face.
Turning to supply-side interests, the United States was initially eager to become involved as a third-party conflict manager to improve its influence in the Middle East relative to that of the Soviet Union and to get the oil embargo lifted.27 There was also fear that renewed hostilities between Israel and Egypt would bring the United States and USSR into serious confrontation, giving the United States a substantial stake in reducing regional tensions.28 We can see directly how the superpower rivalry—easing during this period of detente but still a strong concern—motivated U.S. incentives to mediate in Kissinger’s adroit denial of substantial Soviet involvement in the peace process. Keeping the Soviets at arm’s length, thereby enhancing U.S. sway in the region and minimizing the friction points between the superpowers, was an important goal for Kissinger. To do this, he had to create the perception that the Soviets were involved in the peace efforts while separately making progress on his own. A key function of the Geneva Conference in December 1973 was thus to involve the USSR and the UN as collaborators and to give the peace process legitimacy when the intention all along was to make progress on the disengagement agreements between Israel and Egypt after the conference and without Soviet involvement.29
The motivations for Camp David are arguably even more closely tied to third-party incentives than the motivations for the disengagement talks because the latter occurred in the shadow of massive battle losses, while the former occurred during a period of relative calm in relations between Israel and Egypt. Almost immediately on taking office, Carter began a Middle East peace initiative, and during his tenure he was the most willing of any U.S. president to take political risks for peace in the Arab-Israeli conflict.30 His zeal for peace strongly shaped the incentives for Israel and Egypt, both of which highly valued alignment with the United States at this point, to come together under his auspices. The U.S. president clearly saw his role as important in making the Camp David process happen, as Carter (1982, 316) writes in his memoirs, “There was only one thing to do, as dismal and unpleasant as the prospect seemed—I would try to bring Sadat and Begin together for an extensive negotiating session with me.” Cyrus Vance (1983, 164), secretary of state at the time, also highlights that the administration believed, even before it took office, that progress on peace in the Middle East depended on the United States taking a proactive stance: “President Carter and I agreed that his administration would take an immediate, leading role in breathing new life into the Middle East peace process, building on the foundations laid by our predecessors.”
Note, however, that the Camp David negotiations did not occur until September 1978, almost two years into Carter’s term. So, Carter’s eagerness to mediate the conflict did not immediately produce a third-party assisted peace process and may have even delayed the type of negotiations seen at Camp David. Just as we have seen in the quantitative analysis, disputants often resist the advances of third parties that are too interested in becoming involved; Carter was actually rebuffed early in his presidency. Israeli and Egyptian frustrations with Carter’s push for a renewed Geneva Conference, as well as their aborted bilateral attempt, are suggestive of the incentives and disincentives that disputants consider when choosing conflict management approaches. Mainly, they indicate that parties often have a preference to keep the negotiations as limited as possible if they can. Both Israel and Egypt wanted to avoid the spectacle of another Geneva Conference because of all the compromises that would be needed just to convene the conference, chief among them agreeing on how to incorporate Palestinian representation and how to limit Soviet and UN involvement.31 Israel and Egypt initially preferred a bilateral attempt to reach a comprehensive settlement and only later, after the bilateral initiative fizzled, saw merit in another U.S.-led peace process. Disputants will often prefer to start with a bilateral approach because of the costs and risks associated with including an intermediary.32 This also suggests that forum shopping for conflict management is sometimes a trial and error process. Disputants may start with a bilateral initiative and then later incorporate a third party when their initial attempt fails and they become more impatient for resolution and the assistance that third parties can bring.
Carter in North Korea
A later example is Jimmy Carter’s mediation as a private citizen in the North Korean nuclear crisis. This case, too, demonstrates the importance of conflict costs and supply-side incentives in choosing mediation but in important new ways. Here the high costs were anticipated and not actually related to heavy battlefield casualties. So, we see some evidence that expected costs alone can increase the incentives for the actors to desire mediation. Moreover, the supply-side pressure was not from a major power using its geostrategic influence but, rather, from a private citizen using less tangible political influence. Soft power can be relevant in addition to traditional power when we consider the incentives for mediation.
The DPRK signed the Nuclear Non-Proliferation Treaty (NPT) in December 1985, but delayed signing the mandatory International Atomic Energy Agency (IAEA) safeguards for six years.33 Only after the United States removed its nuclear forces from the Korean peninsula in 1991 did North Korea finally agree to the safeguards of seven declared sites in January 1992. Also, in December 1991, the DPRK and the Republic of Korea (ROK, or South Korea) agreed to a “Joint Declaration on the Denuclearization of the Korean Peninsula,” in which both sides promised to not pursue nuclear weapons and to allow mutual inspections. Between May and January 1993, the IAEA conducted six ad hoc inspections of the declared DPRK sites. The inspectors found discrepancies between the declared inventory and the observed levels, creating the suspicion that the DPRK had diverted fissile material prior to inspections. But the IAEA inspectors were denied the necessary access to verify the accuracy of this suspicion. In addition, after a tip from U.S. intelligence, the IAEA requested special inspection of two suspected nuclear waste storage facilities that were not included in the list of declared sites. On March 12, 1993, the DPRK refused to allow the special inspections and declared its intent to withdraw from the NPT. On April 1, the IAEA Board of Governors reported the DPRK noncompliance to the UN Security Council, which passed Resolution 825 on May 11 calling for North Korea to comply with its safeguards agreement.
After a year of high tensions and failed diplomatic initiatives, the situation escalated in mid-May 1994 after the DPRK began extracting the 8,000 spent fuel rods in the Yongbyong 5-megawatt reactor. On June 13, the DPRK officially submitted its intent to withdraw from the NPT, and on June 15, the United States began consultations with other Security Council members, as well as with Japan and South Korea, about a two-phase sanctions plan. The United States had also begun mobilizing its forces in the region and strongly considered a tactical strike on the Yongbyong complex because it feared that the DPRK might launch a preemptive military strike in response to the threat of sanctions and in advance of a full U.S. deployment (Wit, Poneman, and Galucci 2004).
Amid this growing crisis that was threatening to spiral out of control, Jimmy Carter acted as a mediator on June 15–18, 1994. On June 6, Kim Il Sung, who knew that the former president kept an open mind toward compromise with the DPRK, confirmed a standing invitation for Carter to mediate. The Clinton administration debated whether to allow Carter to go because he held a more lenient stance on whether the DPRK should be allowed to reprocess nuclear material—it was permitted under the NPT, but prohibited by the North-South joint declaration—and was much more dovish on sanctions.34 Ultimately, Clinton allowed Carter to go, hoping that his status as a former president might afford Kim the opportunity to back down from his nuclear program without losing face at home.
The expected costs of conflict shaped the propensity for mediation in this case. Even though hostilities never actually broke out, their heightened potential to do so, perhaps the highest since 1953, and the prodigious destruction they would entail made the attempt at mediation worth a shot. In June 1994, the Clinton administration was in the final stages of both seeking sanctions from the UN Security Council and mobilizing forces in the region. If the crisis had continued along the course it was taking, the risk of full-scale war would have increased substantially and the costs of such a war would have been enormous. Thus, although the direct immediate costs of conflict were not high in June 1994 (violent hostilities had not yet erupted), the expected costs of continuing in crisis—the probability of war multiplied by its costs—were high.
As discussed in chapter 2, disputants may have both sincere and insincere motives for mediation. Thus far I have discussed the high expected costs in this crisis in terms of sincere motivations—the attempt to find an alternative path of reaching a settlement in the midst of the potential for severe escalation. But the costs also probably created incentives, especially on the DPRK side, to pursue mediation insincerely with the end of stalling. The DPRK, with its nascent weapons program and foundering economy, presumably saw an opportunity to use the peace process as a means both to delay the threat to its weapons program and to extract aid from the international community. We will return to this likely motivation later as a key explanation for why tensions eventually renewed after this bold peace initiative and also consider how the Clinton administration tried proactively to reduce the DPRK exploitation of the process.
The United States was not oblivious to the potential for the DPRK to use the mediation initiative as a charade, but its perceived expected costs of the crisis made the mediation gamble worth taking. Moreover, supply-side pressure on the U.S. side helps explain the occurrence of mediation in this case. Carter was primarily motivated by the possibility of renewed war on the Korean peninsula if additional sanctions were levied, as well as the humanitarian consequences of such sanctions, and felt that he could use his experience and profile to, at a minimum, open up a line of communication and potentially help the actors find a way out of the escalating tension.35 To achieve these ends, Carter would have gone to the DPRK in June 1994 even without permission from the Clinton administration (Creekmore 2006, 44, 68). This threat to visit with Kim Il Sung without Clinton’s blessing placed pressure on the administration to consent to the initiative. Had Carter gone to the DPRK—as a third-party consultant, not a mediator—without Clinton’s support, Clinton might have looked weak and he would not be able to take credit for any progress that resulted. Realizing that it would be better for Carter to go with permission than without, Clinton eventually, but reluctantly, consented to having Carter play a mediation role. Again, this untraditional supply-side pressure is noteworthy because the likely devious objectives on the DPRK side would have created substantial counterincentives to avoid mediation. Carter helped convince Clinton that the mediation initiative was still worth the gamble, unlikely as it was to produce a satisfying settlement.
Roosevelt at Portsmouth
Theodore Roosevelt’s mediation between Russia and Japan in 1905 at the Portsmouth Naval Shipyard also demonstrates the dynamics discussed in this chapter. Foremost, this case confirms that mediation becomes quite likely when both sides are hurting and become eager for a way out. One of the interesting elements of this case is that the actors reached a point of mutual suffering even though it did not look like a traditional hurting stalemate. That is, Japan eagerly sought mediation immediately after one of the most resounding victories in military history because of just how costly the conflict had become. This case also provides insight into the role of supply-side influence in leading to mediation.
The conflict between Russia and Japan that eventually led to the outbreak of war in 1904 was set in motion in 1895, after Japan defeated China in the first Sino-Japanese War.36 The Treaty of Shimonoseki, signed on April 17, 1895, terminated this war. In an earlier agreement between Japan and China, Japan was to receive the Liaodong (Liaotung) peninsula and Port Arthur, important acquisitions for the Japanese establishment of a powerful presence on the Asian mainland. This, however, threatened the interests of the European powers who also had plans to increase their influence in Asia. As a result, Russia, Germany, and France pressured Japan to give up its claims to the Liaodong peninsula in exchange for more of an indemnity from China. This gave the European powers greater access to China and helped bring about the lease of Port Arthur to the Russians in 1897. The Japanese perception of Russia as a threat continued in the coming years as Russia solidified its position in Manchuria after the Boxer Rebellion in 1900. Meanwhile, Japan began establishing itself as the dominant power in Korea, which brought Japanese and Russian interests in East Asia into close competition. Japan wanted to be recognized as the dominant power in Korea while limiting the Russian presence in Manchuria.
On February 8, 1904, Japan launched a surprise attack on Port Arthur, thus beginning the Russo-Japanese War. With most of its military forces based in Europe, Russia experienced early difficulties in the conflict. Japan defeated the Russians at the battle at the Yalu River and then crossed the Yalu to defeat the Russians again at Liao-Yang in late August 1904. After months of struggle, Japan captured Port Arthur on January 1, 1905 and then defeated the Russians again at Mukden in March. This set the stage for the Russian defeat at the Battle of Tsushima Straits on May 27–28, 1905. Russia had sent its Baltic fleet around the world to cut Japan off from its land forces, but Russia lost sixteen warships in the battle, Japan virtually annihilating the entire fleet.37 This capped a series of confrontations in which the Japanese were able to defeat the Russians, but with high casualties and without completely devastating the Russian ground forces.38
Realizing that pursuing the war was taking a huge financial toll, on May 31, 1905, Japan asked Theodore Roosevelt to mediate. His invitation to host peace talks was formally accepted by the Japanese leadership on June 10 and by the Russians on June 12.39 It is worth noting that Roosevelt had made gestures toward mediation in January 1904, even before the war began, but Japan initially preferred to seek a bilateral agreement to end the conflict.40 Japan began testing the waters for mediation in March 1905, after Mukden, and entreated Roosevelt to mediate immediately after its victory at Tsushima. Russia became amenable to negotiations after its public became quite critical of the conflict and the ongoing Russian Revolution became more of a threat to Tsar Nicholas’s rule.41 The formal negotiations began on August 9 at the Portsmouth Naval Shipyard in Kittery, Maine. Komura Jutaro, the Japanese foreign minister, served as the chief negotiator for Japan; Sergius Witte, president of the Council of Ministers, served as the first plenipotentiary for Russia.
The possibility of continued conflict costs generated the primary impetus for the peace initiative and for greater involvement by Roosevelt as the initiative stalled. Although Roosevelt initially distanced himself from the negotiations—he in fact had pledged not to mediate when he sent the initial invitations to the peace conference—and retired to his summer home in Oyster Bay, New York, the threat of the failure of the negotiations and ongoing conflict motivated the Japanese delegation to request that he take a more active mediation role.42 Moreover, both Witte and Komura pushed back the end date of the conference multiple times to avoid leaving empty-handed.43 In other words, Roosevelt did not have to do very much to encourage the sides to stay at the bargaining table; the adversaries demonstrated significant resolve on their own to prolong the negotiations in the face of impending failure.
It is quite significant that Japan was so eager to enter into mediation after its successes at Mukden and Tsushima. The conflict had become so costly that Japan did not want to continue its successful campaign against Russia. Even in victory, Japan suffered many casualties fighting against the well-entrenched Russian positions. In the future, the costs of the war for Japan would only have increased because it was not well situated for a protracted land war with Russia outside a limited portion of Manchuria and it could never feasibly conquer Russia.44 In the first 240 days of the war, Japan had suffered nearly 58,000 casualties, compared to 28,000 for Russia (White 1964, 186). The Japanese economy was also struggling, worsened by the ongoing warfare that was financed completely through foreign loans (Princen 1992, 116). Had Japan borne fewer costs, it is doubtful that Japan would have been as eager for a quick end to the conflict or to stay at the bargaining table despite not being able to get all that it had expected.
At the same time, third-party pressure, especially on Russia, also shaped the preferences for mediation, particularly in regard to choosing the mediator. The Russo-Japanese War generated much international interest because it involved two great powers fighting for dominance in a part of the world where many other great powers were interested in securing trading and territorial rights. The terms of settlement of this conflict would not only affect the state of affairs in East Asia, but it would also lead to a change in the relative power held by both Japan and Russia. After the Spanish-American War, Roosevelt’s interest in the global standing of the United States heightened, and he particularly focused on the power struggle in East Asia (Trani 1969, 7). He understood what was at stake and had a clear preference for a certain outcome. Demonstrating that a mediator’s aims are often far from altruistic, Roosevelt remarked to a French ambassador, “From my point of view, the best [outcome] would be that the Russians and Japanese should remain face to face balancing each other, both weakened” (quoted in Princen 1992, 109).45
Roosevelt’s eagerness to mediate eventually awarded him the opportunity to do so. Japan initially spurned his offer, demonstrating the reluctance of many combatants to allow foreign actors with strong interests to meddle in their security affairs. Nevertheless, Japan and Russia eventually warmed to the idea of Roosevelt mediating. France, Germany, and Britain all might have been suitable mediators, but Roosevelt had actively cultivated a good rapport with the Japanese.46 Moreover, Russia sought U.S. financial backing, and Witte hoped to use the peace initiative as a gesture of goodwill to court support.47 Leaning on Russia further, Roosevelt appealed to Kaiser Wilhelm of Germany to urge Tsar Nicholas II to consider Roosevelt’s mediation offer. As a result, in a June 3 letter, Wilhelm expressed to Nicholas II his belief that Russia should admit defeat after Tsushima and consent to U.S. mediation (White 1964, 209). The combination of pressures, from both within and without, on the parties to negotiate led to the acceptance of Roosevelt’s invitation one week later.
To understand the impact of mediation, we must first understand what the actors hope to achieve through it and the conditions under which it is most prevalent. Studying the incidence of mediation also helps preempt a question that arises when the short-and long-term trade-off of mediation is explored in chapters 5 and 6: Under what conditions would actors consent to mediation if it decreased the stability of peace? In this chapter, I have examined the conditions under which mediation is most prevalent. The first two hypotheses are that mediation occurs when the costs of conflict are high and the shadow of the future is short. In such situations, the short-term benefits are extremely attractive, with any long-term risks being more of an afterthought. A third hypothesis is that mediation is more likely to occur when there are strong pressures from third parties. The results from the ICB and ICOW data confirm that third-party peacemaking is more likely during severe conflict in unstable or new democratic regimes and in democratic communities. It also appears, however, that mediation is less likely to occur when the third parties are too eager to intrude and shape the outcomes as peacemakers. The cases of Kissinger and Carter in the Middle East, Roosevelt at Portsmouth, and Carter in the DPRK illustrate these important factors that shape preferences for mediation. We will return to these cases in subsequent chapters to observe the impact of mediation in both the short and long terms.
For a more thorough account of selection effects with regard to the study of mediation, see Gartner and Bercovitch (2006). Note that this is different from selection bias, which is a problem in studies that examine the set of mediated cases only to determine what causes mediation success.
For analyses that find that a costly and lengthy conflict heightens the probability of mediation, see Greig (2005); Greig and Diehl (2006); Bercovitch and Jackson (2001); Svensson (2008). For additional studies relating desperation to the desirability of mediation, see Bercovitch (1996); Terris and Maoz (2005); Böhmelt (2010a).
Note that the logic here is not necessarily at odds with Goemans’s (2000) notion of gambling for resurrection in which leaders that are almost certainly going to lose office and then face severe punishment will attempt to seek a military victory to regain political control. The Kabila example provides a case of a leader that has a tenuous hold on power and is looking for short-term stability but is not at the point of needing an unlikely victory to stay in power.
In this vein, Gelpi and Grieco (2001) argue that leaders early in their tenure have a greater willingness to make concessions because they are vulnerable and unable to take on the burden of sustained military combat. Chiozza and Choi (2003) and Beardsley (2010) indicate that the impact of tenure on a leader’s preference for mediation is conditioned by regime type.
See Bueno de Mesquita and Siverson (1995); Reiter and Stam (1998); Gelpi and Griesdorf (2001). In a similar vein, Reiter (1995) argues that leaders often avoid preemptive wars because of the political burdens of initiating conflict. Slantchev (2003, 2004) and Filson and Werner (2002) also argue that the initiators of conflict are the actors most prone to revise their expectations of success downward as war progresses.
For an assessment of when third parties will offer to provide mediation in interstate and intrastate conflicts, see Greig (2005); Greig and Regan (2008). For a network perspective of how close ties with third parties can influence the incidence of mediation, see Böhmelt (2009).
The value 23 percent is for all crisis dyads, which translates into approximately one-third of all crises.
ICOW dispute length is not included here as an indicator of costliness because so many of the disputes never became militarized.
This approach intentionally double-counts actors that are neighbors to both combatants because such actors are especially affected by the dispute.
See also Crescenzi, Kadera, and Mitchell (2007).
The challenger-specific models are possible only with the ICOW data because challengers and targets are not identified in the ICB data.
Multiplied by 100, the value for relative risks gives the percentage change in the risk of occurrence. The baseline configuration has all the interval-level independent variables set at their means, all the ordinal variables at their medians, and the dichotomous variables at zero.
For the ordinal variables, the changes are rounded to the nearest integer. The geostrategic salience variable is increased to its maximum because that is the closest possible value to the two-standard-deviation (2 SD) increase.
These values were calculated from Cox duration models and are simply the relevant hazard ratios minus 1. A hazard ratio is the ratio between the hazard rate with the treatment and the hazard rate without the treatment—where the hazard rate is the probability of experiencing the event conditional on having not experienced the event yet and thus can be thought of, in this case, as a type of risk of experiencing third-party conflict management.
This variable does become statistically significant, with the expected sign (negative), when the sample is restricted to the set of democratic dyads.
When the variables capture the dyadic averages, the sample consists of the democratic dyads; when the variables capture the challenger’s characteristics, the sample consists of the democratic challengers.
See Beardsley (2010). This explanation is not complete, however, because it does not explain why regime stability does not also matter less in the ICB analysis.
See Touval (1982, 236).
Carter (1982, 283) also recognizes Sadat’s war weariness as a crucial factor in bringing him to the negotiating table. See also Baker (1978, 130, 243).
See Stein (1999, 97).
See Baker (1978, 137–38).
In fact, the lengthy quotation from Sadat’s memoirs expresses his justification for his peace initiative that began with his trip to Jerusalem in 1977 and that led to the Camp David process in 1978.
On Sadat’s need to save face, see Carter (1982, 345); Quandt (1986, 203).
It is, of course, difficult to assertain the truth of such a claim made by someone assigning blame for a failed bid for election.
On the importance of domestic political constraints in the bilateral initiative, see Stein (1999, 231).
For a thorough discussion of the importance of superpower competition in shaping the incentives for U.S. involvement, see Telhami (1990). In his memoirs, Carter (1982, 278) cites oil dependence as crucial in U.S. incentives to mediate. See also Quandt (2001, 125–26); Horne (2009, 237); Ben-Ami (2005, 149).
See Kissinger (1982, 749–50, 755, 843); Isaacson (1992, 538–39); Quandt (2001, 135). Brzezinski (1985, 83, 87) notes a similar motivation for the push by the Carter administration for a return to Geneva.
Ezer Weizman, the Israeli defense minister, was especially leery of maintaining a high degree of dependence on the United States and preferred initially to develop face-to-face negotiations with the Egyptians (Weizman 1981, 126–27).
For an overview of the DPRK nuclear crisis, see Mack (1994); Cronin (1994).
On the deliberations by the Clinton administration regarding Carter’s visit, see Wit, Poneman, and Galucci (2004, 203, 243).
For an overview of the motivations for war and of the war events, see Auslin (2005); Randall (1985); Schimmelpenninck van der Oye (2005); Teramoto (2008).
Perhaps this was not so surprising; Russia had not established itself as a naval power, as seen in Rudyard Kipling’s remarks just prior to the war: “I’ve been hearing details of the Russian fleet that are rather quaint. They seem to have all the makee-look but very little of the makee-do aboard vessels” (1903).
On the costs to the Japanese, see Steinberg (2005); White (1964, 206, 288–90).
For accounts of the invitation and responses, see White (1964, 208–13); Dennett (1959, 189). For the formal language of the invitation, see Papers Relating to the Foreign Relations of the United States (1906, 807, 817). Trani (1969) and Princen (1992) also provide accounts of earlier considerations of a mediation initiative.
See Lukoianov (2008, 41).
On the relationship between the instability in Russia and the peace process, see Saul (2005, 494).
See White (1964, 237–38).
See Trani (1969); Princen (1992).
On the precarious position of Japan, see White (1964, 185–205).
See also Dennett (1959, 202).
See Princen (1992); Trani (1969).
On the Russian incentives, see Lukoianov (2008, 50).
Month: | Total Views: |
---|---|
March 2024 | 2 |