Overview

This chapter discusses the role of justice in negotiations between rival groups and the durability of resulting peace agreements. It draws on information about group negotiation processes and agreements concluded to end civil wars in countries around the world, mostly during the early 1990s. Possible relationships between the presence and importance of distributive justice (DJ) in the agreements, and their stability, were first explored. The difficulty of the conflict environment was shown to have the strongest impact upon stability. However, the DJ principle of equality was found to reduce the negative impact of difficult conflict environments on their stability. An emphasis on equality was also associated with more forward-looking agreements, which were found to be more stable than backward-looking ones. Next, the presence and importance of procedural justice (PJ) were examined in the negotiation processes that led to the signing of the peace agreements. Significantly more stable agreements occurred when a process based on PJ led to agreements emphasizing equality.

A close examination of how the equality principle was expressed in the agreements revealed three main types of provisions: equal measures, equal treatment, and equal shares. Agreements with equal treatment and/or equal shares were associated with forward-looking outcomes and high stability, and equal measures with a more backward-looking outcome and poorer stability. Third party roles were then assessed in four select cases. In both cases of high stability (Mozambique, Zimbabwe), third party intervention was central to the formulation of high equality agreements and to implementation. In the cases of low/no stability (Angola, Rwanda), third parties did not work actively to promote agreement based on forward-looking or any equality provisions. The findings suggest that negotiators and third parties should strive for agreements based on equal treatment and/or equal shares, since they are more stable during the implementation period and beyond, and that a variety of tactics and approaches (both facilitating and forceful) can serve that objective.

The research continued with an examination of the impacts of justice on macro-level societal peace. Sixteen durable peace variables were included in an expanded data set of 50 cases of peace agreements. Among the key findings were relationships between procedural justice and societal peace over the course of 8 years following the negotiated agreement. Specifically, we discovered a path going from PJ during the negotiation process to DJ in the outcome to stable implementation of the agreements and to societal peace. Of particular interest were the importance of two PJ principles, fair treatment and transparency, and the role of community activities during the post-agreement period. A discussion of possible explanations for these findings and implications for a culture of peace conclude the chapter.

Issues concerning the role of justice in negotiation have been addressed by scholars and practitioners in a number of areas in social science. These areas include the study of civil wars, international trade negotiations, historical negotiations on security issues, law, organizational management (see the chapters by Eden and Ackermann and by Eden, Kaur and Carreras for examples of the impact of PJ in the design and use of Group Support Systems), and social psychology. They focus attention on group decision processes that occur in this domain. We have learned from these studies about how justice influences negotiation processes, outcomes, implementation, and the durability of societal peace. A review of the justice literature precedes a discussion of our projects on stable and durable peace. We then develop implications of the findings for practice and policy.

How Justice Influences Negotiation Processes, Outcomes, Stability, and Durability

The influence of justice on negotiation processes and dynamics has been explored in interpersonal (e.g., Deutsch 1985), organizational (e.g., Konovsky 2000), and international (e.g., Zartman et al. 1996) contexts. A study of international negotiations across four issue areas (trade, the environment, ethnic-sectarian conflict, and arms control) found that negotiators regularly act upon justice considerations and that these can affect the process in numerous ways (Albin 2001). At the most basic level, they may guide the framing of the issues – proposals put forward, the exchange and evaluation of concessions, and the formulation of agreements – and thereby facilitate the trade-offs, particularly when parties share the same or compatible notions of justice.

An associated concept is the norm of reciprocity; that is, mutual responsiveness to each other’s concessions. Research has distinguished several different patterns of how concessions are made while negotiating. These include “comparative responsiveness” – that is, acting based on a comparison of one’s own and the other’s tendencies to concede (Druckman and Bonoma 1976; Druckman and Harris 1990) – and “diffuse reciprocity” – that is, acting to ensure that roughly adequate or sufficient, rather than specifically equal or comparable, concessions are made to establish a balanced agreement overall (Albin 2001).

Distributive justice considerations may complicate the bargaining process, cause deadlocks and stalemates, and become subject to negotiation themselves. This pertains to the situation in which parties endorse competing justice principles or interpretations (applications) of them. In the end, however, reaching agreement usually requires formulating terms that can earn the respect and voluntary approval of all parties and their constituencies partly by appealing to their sense of justice. Negotiators are thus motivated to act on terms that can be generally accepted as reasonable and balanced. The search for such terms frequently leads them to balance and combine several justice principles in the terms of agreements. This very act of balancing is also associated with justice, in a situation in which no distributive justice principle emerges as morally superior on its own and several are needed to take account of relevant factors and different circumstances (Zartman et al. 1996). Similarly, a study of how public resources and burdens are allocated demonstrates that decision makers must balance different justice principles and that major theories of justice fall short of capturing these real-world nuances (Young 1994).

The presence of procedural or process justice is also widely regarded as adding legitimacy to the results (Albin 2008). Beyond this, however, general systematic conclusions about how justice in the negotiation process influences the terms of agreements and the outcome are few. In an analysis of international trade talks, adherence to procedural justice while negotiating was found to increase the chances for mutually beneficial agreements (Kapstein 2008). In her study of the Liberian peace process, Hayner (2007) found that stable agreements depended on both procedural justice (fair representation of stakeholder groups) and confronting complex issues during the negotiation process. Along similar lines, Hollander-Blumhoff and Tyler’s (2008) field experiments showed that when procedural justice principles are evident in the process, the negotiators are: (a) more willing to disclose information, (b) more trustworthy, (c) more likely to attain an integrative agreement, and (d) be more durable. These findings were supported by Wagner (2008) in her study of a dozen historical cases of security talks and by Konovsky (2000) in her review of the management literature. Whether procedural justice promotes agreements based specifically on distributive justice is disputed in both research and policy debates. In the context of business organizations, Konovsky (2000) did find a relationship between procedural (process) justice and distributive justice in the outcome.

Justice and Negotiation: A Framework

A framework that organizes the justice and negotiation literature was developed by Druckman and Wagner (2016). Realizing that justice studies concentrated on particular stages of negotiation, they placed the research in one of four stages: prenegotiation, negotiation processes, negotiation outcomes, and implementation of agreements or postnegotiation.

With regard to decisions made during the prenegotiation stage, the research focused on the framing of group or community boundaries, which is important for expanding or limiting inclusion of various stakeholders in the negotiation process (Mikula and Wenzel 2000). It also provided insights into the way that negotiating norms are developed, the difference between self and collective interests (Lind and Tyler 1988), and preferences for alternative distributive principles (Deutsch 1985).

With regard to the negotiating process, we have learned about relationships between PJ and motivational orientations as self-interest or group values (Pruitt and Carnevale 1993), the importance of shared identities in building the trust needed to implement PJ principles (Tyler and Blader 2003), the interesting concept of false justice (Lind and Tyler 1988), and the way that PJ can balance the playing field between parties that are asymmetrical in power (Kapstein 2008). A particularly compelling finding is the co-varying relationships (as additive or interactive) among PJ, trust, and problem-solving behavior.

The justice research on negotiation outcomes has focused attention on distributional benefits in terms of the distributions accorded under different DJ principles (equality, proportionality, compensation, need). The research has probed the decision processes used to reach each of these types of outcomes and the impacts of these outcomes on the stability of agreements. Another popular research topic has been the relationship between PJ and DJ (Wagner and Druckman 2017). Questions explored have included the strength of the relationship between these types of justice, the relative contribution of PJ and DJ to perceptions of fairness, and the compensating effects of these principles in the sense of whether adherence to PJ principles during the process offsets disappointing distributional outcomes (Brockner and Wisenfeld 1996).

On implementing agreements, the research has concentrated on the role of PJ in restoring relationships between the parties during the postnegotiation period including long-term reconciliation (Tyler and Blader 2003), the role played by spoilers in preserving agreements (Stedman 2000), and implications for other negotiations being conducted in parallel or in sequence (Spector and Zartman 2003). A set of hypotheses were generated for each stage culminating in a path that depicts the way justice in the process leads to outcomes with consequences for compliance and reconciliation.

Another contribution made by Druckman and Wagner’s review is alternative perspectives on justice and negotiation. One perspective emphasizes systemic or nonlinear processes, another focuses on linear sequences. The former considers justice as part of a system of interacting variables that have bidirectional and circular effects. The latter construes justice as part of a chronological sequence as depicted by the stages idea discussed above. An advantage of systemic thinking is that it places a negotiation in a larger context of interacting influences. It reveals the dynamics of justice perceptions: for example, the circular relationship among PJ, trust, problem solving, and integrative outcomes. A question raised by this perspective is what sets the cycle in motion; does trust precede or follow adherence to PJ principles? This is a question of order, which is addressed by the sequential perspective. Thus, the perspectives can be regarded as being complementary. For example, trust reinforces PJ and problem solving which, working together, move the talks toward a durable agreement. Similarly adherence to PJ principles, encouraged by third parties, reinforces trust and problem solving leading to positive outcomes.

Research on the Role of Justice in Peace Agreements

The extensive literature on negotiations to end civil wars includes studies of cases from a variety of regions and countries (e.g., Zartman 1995; Stedman et al. 2002) and large-sample comparative studies (Fortna 2004; Hartzell and Hoddie 2007; Druckman and Wagner 2019). Findings from these studies shed light on the conditions – both within and outside the negotiating room – for concluding and sustaining peace agreements. An example of important findings comes from the comparative study conducted by Downs and Stedman (2002). Focusing on a set of 16 peace agreements concluded mostly during the early 1990s, these investigators showed that implementation was largely a function of the difficulty of the conflict environment surrounding the talks. Less successful implementation occurred in more difficult conflict environments. Another variable, willingness of neighboring powers to intervene, had virtually no impact on implementation. Missing from this study, and generally from research on settling civil wars, is the role played by justice. This gap was filled by several studies on justice and the stability of peace agreements. A first study focused on distributive justice (DJ) in the agreements. A second study concentrated on procedural justice (PJ) in the negotiation process.

Both studies utilized original systems for coding justice. The DJ codes consisted of key words in the agreements that signaled each of the four principles: equality, equity, compensation, and need. The PJ codes consisted of evidence in primary and secondary accounts of the negotiation process that reflected each of the four principles: fair treatment transparency, fair representation, and voluntary decisions. The development and implementation of these coding systems facilitate the evaluation of hypotheses about relationships among the justice and stability/durability concepts. The coding process converted the justice concepts into variables such as the extent to which each DJ or PJ principle was central to the agreement or was emphasized during the negotiation process. These “conversions” facilitate performing statistical tests that evaluate hypothesized relationships: For example, the more central DJ (or PJ) principles are in the agreement (or in the process), the more stable the agreement. The results of the statistical analyses can then be used to construct models that depict the way that the set of variables interact through time across the cases: for example, PJ principles in the process lead to DJ principles in the agreement which, in turn, results in a stable agreement. A more recent study expanded the scope of this research by adding the concept of durable societal peace measured over a period of 8 years (Druckman and Wagner 2019). This study also extended the sampling from 16 to 50 peace agreements. Each of these studies is reviewed followed by a discussion of implications for the connection between justice and peace.

Distributive Justice and Stability

Building on the Downs-Stedman data set, Druckman and Albin (2011) coded the 16 peace agreements for four DJ principles: equality, proportionality, compensation, and need. These particular principles are emphasized in both theoretical and empirical research, and actual negotiation practice (see e.g., Albin 2001; Deutsch 1985; Konovsky 2000). Druckman and Albin also developed coding categories for types of agreements, namely, whether they were forward-looking (FL) or backward-looking (BL). Complete texts of all the agreements were assembled from web documentation for coding DJ and FL/BL. The agreements varied in length from 5 (the agreement between the government of Nicaragua and YATAMA) to 52 pages (the agreement between the Republic of Rwanda and the Rwandese Patriotic Front). Although longer texts are likely to provide more opportunities for statements that relate to justice to appear, our emphasis on centrality of the principles, rather than frequency of their appearance in the text, reduces the problem. Each agreement was examined for the presence of each DJ principle – equality, proportionality, compensation, or need.

Three other variables were included in the data set: implementation (failure, partial success, and success), difficulty of the conflict environment (number of warring parties, likelihood of spoilers, number of soldiers, and access to disposable resources), and willingness of neighbors to intervene in the conflict or to provide resources.

This study evaluated a number of hypotheses. The literature presents competing hypotheses about how DJ relates to stability – that basing agreements on DJ either increases (based on arguments about root causes of internal conflict) or decreases (based on arguments about negotiating norms) their stability. These hypotheses were reconciled by including the difficulty of the conflict environment. The authors hypothesized further that the root causes argument holds in less difficult environments; the normative argument holds in more difficult conflict environments. Variation among the cases on the conflict difficulty variable provided an opportunity to evaluate these contending hypotheses.

Hypotheses were also evaluated concerning the effects on stability of each of the DJ principles, which we considered as being either forward (equality and proportionality) or backward (compensation and need) looking. In particular, the forward-looking principles were expected to occur more frequently than backward-looking principles in the agreements. They were also expected to produce more stable agreements. A final hypothesis posited that forward-looking outcomes – which may include forward-looking justice principles – would be more stable than outcomes which deal primarily with the past.

The results can be summarized as follows. The strongest relationship was between the difficulty of the conflict environment and stability: Less stable agreements occurred in more difficult environments (r = −0.65). A moderately strong correlation was obtained between justice and stability (r = 0.56). However, these relationships changed when partial correlations were calculated. A slightly reduced correlation between difficulty and stability was obtained when justice was controlled (from −0.65 to −0.57). A reduced correlation was also obtained between justice and stability when difficulty was controlled (from 0.56 to 0.46). Similar results were obtained from a regression analysis that included the difficulty, justice, and stability variables. These variables form a cluster as indicated by the results of a factor analysis. The willingness variable did not load on this factor, nor did it produce any significant correlations with the other variables.

These findings suggest that when justice principles are central to an agreement, the impact of more (less) difficult environments on stability is reduced (enhanced). In technical terms, justice was shown to mediate the relationship between the difficulty and stability variables. This means that DJ contributes to the stability of peace agreements. That contribution is indirect in the sense of reducing the negative effects of intense conflicts on stability or increasing the positive effects of less intense conflicts. These findings provide some support for the root causes argument: addressing issues of DJ in outcomes contributes to the shelf life of an agreement. They do not support the normative argument: addressing DJ issues did not interfere with implementation of the agreement. Further investigation provided additional clarification for these findings.

Analyses conducted on each of the four DJ principles revealed that one principle in particular accounted for the relationships between difficulty, DJ, and stability. This was the principle of equality, which was also the most frequently-occurring principle in the agreements. When equality was analyzed separately, the same relationships among the variables emerged: like DJ, equality was shown to mediate the relationship between difficulty of the conflict environment and stability. In fact, the relationships between each of the other variables and equality were stronger than they were when DJ (measured as an aggregate of the four principles) was used as the justice variable in the analyses – the DJ-stability correlation was 0.56; the equality-stability correlation was 0.76. The inclusion of the other principles actually depressed the relationships with the difficulty and stability variables. Each of the other DJ principles (proportionality, compensation, and need) showed very weak relationships with stability. Thus, equality accounted for (or mediates) the relationship between DJ and stability.

These findings are summarized in Fig. 1. They suggest that the relationship between difficulty (referred to as an independent variable [IV]) and stability (referred to as the dependent variable [DV]) depends on equality principles (referred to in Fig. 1 as the mediator [M]). Thus the negative effects of difficulty on stability are reduced when equality is central to the agreement; they are increased when equality is not central to the agreement.

Fig. 1
figure 1

The mediating effect of equality on agreements

Full size image

The variable referred to as forward and backward-looking (FL/BL) outcomes was also analyzed. The findings show a strong relationship between this variable and stability: More forward-looking outcomes are more stable (r = 0.66). However, that relationship was also shown to be accounted for by the equality principle: When equality was statistically controlled, the relationship between FL/BL and stability decreased dramatically (from 0.66 to 0.38). The mediator analysis showed a significant indirect effect for equality (Sobel’s z = 1.96, p <0.05). Thus, the impact of FL/BL on stability is due largely to the centrality of the equality principle in the agreement. More forward-looking outcomes occur when equality is emphasized: Most, but not all, of the forward-looking agreements contained equality provisions. However, the stability of the agreements depended more on equality than on FL/BL outcomes. (See Druckman and Albin (2011), for more details.)

Procedural Justice and the Stability of Agreements

The negotiations were also examined for the presence of PJ, defined in terms of four principles: transparency, fair representation, fair treatment and fair play, and voluntary agreement (Albin and Druckman 2012). These principles are widely recognized as key components of procedural justice in the research literature (e.g., Thibaut and Walker 1975; Lind and Tyler 1988; Konovsky 2000; Hollander-Blumoff and Tyler 2008). Moreover, they lend themselves well to being operationalized so that their role in particular cases can be assessed. Together the four principles define an ideal way of negotiating against which actual practice can be examined. Our coders were instructed to judge whether each of these principles was present and, if so, how influential (significant) it was in the process, even if not stated by name.

The amount of documentation available on the negotiation processes varied from case to case. At one extreme is the daily chronology of the Cambodian peace process assembled by Raszelenberg (1995) and the round-by-round discussions in the Mozambique talks described by Hume (1994). At the other extreme is the scarce documentation on the negotiations on Liberia and Angola (the Bicesse accords). For these cases, we sought the assistance of experts; for example, Herman Cohen, former US Assistant Secretary of State, coded PJ in the Angola talks. Confidence in the coding was bolstered by high agreement between Ambassador Cohen and our student coder. Overall, across the 16 cases, agreement between independent coders was high.

As discussed earlier, a number of studies have shown that PJ plays an important role in outcomes and their stability: When PJ principles are central in the negotiation process, outcomes are more likely to be mutually beneficial and lasting. Further, PJ may lead to outcomes that contain DJ principles and are more forward looking. These earlier findings were regarded as hypotheses evaluated in the context of the 16 peace agreements. Taken together, the set of hypotheses suggests a sequence: less difficult conflicts facilitate adherence to PJ principles that, in turn, lead to equality and forward-looking outcomes that endure.

The results provided partial support for this sequence. First, negotiators did, to some extent, adhere more to PJ principles in less intense conflict environments (r = −0.48). Second, more equality outcomes occurred when negotiators adhered to PJ principles during the talks (r = 0.60). Third, adherence to PJ principles was associated with more stable agreements (r = 0.58). And, fourth, adherence to PJ principles was associated with forward-looking outcomes (r = 0.53). Each of these relationships was, however, qualified by the results of additional analyses.

It turned out that the relationships between PJ and each of the other variables (difficulty, outcomes, and stability) was noticeably weaker when equality was controlled for in the statistical analyses: The correlation between PJ and stability decreased from 0.58 to 0.24 when equality is controlled; the correlation between PJ and FL/BL decreased from 0.53 to 0.26 when equality was controlled. This means that equality in the agreements accounted for effects of PJ on both outcomes and stability.

Once again, equality is the key variable. However, it operated differently in the two studies. In the DJ study, the equality principle reduced the negative influence of intense conflicts on stability (see Fig. 1). In the PJ investigation, this principle accounted for the impact of PJ on stability: without equality principles in the agreement, PJ principles have only a small impact on the stability of the agreement. Nor did PJ mediate the effects of the difficulty of the conflict on stability. Thus, both PJ and conflict difficulty are accounted for by equality: In technical terms, as shown in Fig. 2, this means that we have two separate models, one based on the chain from the difficulty of the conflict (referred to as IV1) through equality (referred to as the Mediator) to the stability of the agreement (referred to as the DV) and the other based on the chain from PJ (referred to as IV2) through equality (the Mediator) to stability of the agreement (the DV). Although these are complex findings, they have in common the key factor of equality in the agreements.

Fig. 2
figure 2

Mediated effects of the equality principle on agreements: difficulty of conflict and procedural justice

Full size image

Two intriguing questions are raised by the results obtained in both studies. One question is: Why is equality important for maintaining peace agreements? Another is: What did the negotiators and third parties do to achieve outcomes based on equality? Answers to both questions have practical implications for designing negotiation processes and for policy. They are discussed in the sections to follow.

The Equality Principle

The centrality of equality and three other principles of distributive justice – proportionality, need, and compensation – were assessed in each of the study’s 16 peace agreements. Fourteen agreements included the equality principle. These agreements were analyzed in terms of what forms the application of equality took; that is, what was to be treated equally and how. Three main types of equality emerged from the provisions across the agreements: equal measures, primarily backward-looking and concerned with military strength and disarmament/demilitarization; equal treatment, forward-looking and aimed to secure nondiscriminatory treatment and equal opportunities for all groups or peoples concerned on a long-term basis; and equal shares, concerned with shared political powers and decision-making on a transitional (time-bound) or longer-term (structural) basis.

Agreements in which equal treatment and/or equal shares are central were associated with highly forward-looking outcomes and high stability, while equal measures were associated with a more backward-looking outcome and poorer stability. Equal treatment specifically was central in all the agreements with the highest stability score, and marginal in the agreements with low stability. Thus, the importance of the equality principle is based primarily on the idea of equal treatment and, to some extent, on the concept of equal shares.

Third Party Roles in Equality Provisions

Questions addressed in this section are: to what extent do third party roles explain the presence of equality in agreements, and their implementation? Why did third parties succeed in achieving high equality outcomes that were implemented in some cases, and failed to do the same in others? These are intriguing questions given that most peace agreements are negotiated in situations of considerable power inequalities between parties – a context commonly thought to impede evenhanded outcomes and successful negotiation.

To shed light on these questions, third party intervention was examined more closely in four African cases, selected to provide contrasts on equality content and implementation success: Mozambique and Zimbabwe (high-equality content, successful implementation), and Angola 1/Bicesse Accords and Rwanda (low equality content, failed implementation). In all cases, questions were examined regarding the identity, status, and functions of third parties; types and stages of intervention; any explicit or implicit statements and efforts concerning the inclusion of equality provisions in an agreement; and the overall role and importance of third parties.

The process leading to the 1979 Lancaster House Agreements on the transition from Rhodesia to Zimbabwe was an unusual case of highly forceful and biased mediation resulting in a high equality outcome. The British mediation team, led by Lord Peter Carrington, controlled and steered the process with a heavy hand throughout, and the two conflicting parties, the Zimbabwe-Rhodesia government and the Patriotic Front, never negotiated directly with each other. Drawing fully on its leverage over the former colony’s rival factions, the mediators regularly threatened to go for a “second-class solution” involving formal recognition of the Zimbabwe-Rhodesia government to elicit concessions particularly from the Patriotic Front. Britain’s tactics, bias and obvious potential to deliver an agreement help to explain the high equality content in the final outcome: They helped to induce both parties to make (more) concessions, with the expectation that it would be rewarded and, in the case of the Patriotic Front, that the “second-class solution” would be avoided (see Davidow 1984).

The 1991–1992 General Peace Agreement for Mozambique resulted from a completely different type of mediation. A four-member team without any leverage to use threats or incentives served as impartial facilitators. They received significant support in the form of financing, expertise, guarantees, and assurances in connection with a signed agreement, and encouragement from international actors (the US, Italy, Portugal, Zimbabwe, and Russia). Drawing on this as well as their competence, relationship-building, creativity, and persuasion, the facilitators helped create dialogue, trust, and cooperation between the rival groups. This impartial yet active and important role paved the way for the conflicting parties themselves to seek reconciliation and peace, and endorse an outcome in which equality was the central concept (Morozzo della Rocca 2003; Hume 1994).

In the 1992–1993 Arusha peace process over Rwanda, official mediator Ambassador Ami Mpungwe of Tanzania initially played the role of facilitator and honest broker: He worked to facilitate dialogue and communication between parties in order to create a positive environment for reaching a mutually acceptable resolution of the conflict. However, growing frustration with the Government of Rwanda and increasing sympathy for the Rwandan Patriotic Front caused a shift toward a more partisan and forceful role by Tanzania (Jones 2001). Mpungwe, along with the USA and France as official observers, coerced the hard-line members of the Government of Rwanda to accept a critical provision: a 50–50 (equal) split in the command of the new armed forces, which benefitted the Rwandan Patriotic Front. This move disrupted the peace process. A sense of fairness and satisfaction appear to have been lost, particularly for the hard-line government members. They felt pressure to relinquish their considerable possessions in exchange for few concessions from the Patriotic Front. A very difficult conflict environment, competing interests among direct and third parties, and the absence of stronger forward-looking types of equality (equal treatment, equal shares), go a long way toward explaining the failure of this case.

In the negotiations leading to the 1991 Bicesse Accords on Angola, Portugal was selected as the official mediator because of its expected impartiality. Their facilitative approach quickly ran into problems, in part due to a lack of leverage over the process. The USA and the Soviet Union, by contrast, actively provided military support to the conflicting parties (the UNITA and the MPLA). Drawing on their influence over the rival factions, the superpowers – as unofficial mediators – became more effective in eliciting concessions and securing an agreement. However, none of the mediators promoted including equality provisions in the agreement. Their goal was to expedite the agreement. Little time was afforded to develop a workable peace plan in the agreement, let alone provide equality provisions. The peace process collapsed after elections had been held.

In all these cases, the third party roles explain much of the outcome. In both successful cases, third party intervention (in one case forceful, in the other facilitating) was central to the formulation of an agreement that featured equality provisions that were carried out during implementation. In the unsuccessful cases, third parties (in one case forceful, in the other both facilitating and forceful) did not work actively to promote agreements based on forward-looking, equality provisions. These cases provide a closer look at the way third parties maneuver to get agreements intended to end civil wars. They also illuminate the importance of equality principles in sustaining the agreements.

Justice and Durable Peace

The peace agreement studies reviewed above focused on the stability of the agreements in the years immediately following the conclusion of negotiations. Missing from the analyses were impacts on societal peace over time. These broader concerns were considered in the next study by Druckman and Wagner (2019). In this study we expanded the depth and width of the research. With regard to depth, we constructed a multidimensional index of durable peace (DP) and eight control variables. These dimensions included indicators of reconciliation, changes in security and governance institutions, and economic growth. Both aggregated (overall DP) and disaggregated (components of each dimension) analyses were performed. On width, we expanded the sampling from 16 to 50 cases of peace agreements. A representative sampling frame included cases from six decades (1950s–2000s) and most of the regions of the world (Africa, Europe, Asia, and North and South America). This robust sampling increased confidence in the generality of the findings about relationships between justice in the negotiation process and outcome (micro-level variables), the stability of agreements (meso level), and societal peace (macro level).

Two statistical models were evaluated. One, referred to as a vertical model, assessed the impact of justice on peace controlling for a host of other variables also hypothesized to influence peace. Results of regressions showed PJ strongly influenced DP. None of the control variables had a significant effect. Further analyses showed that the PJ effect was accounted for by two aspects of the PJ variable, fair treatment and transparency. These aspects are referred to as social conduct, which has been shown to influence negotiation outcomes in laboratory studies (Hollander-Blumoff 2017). Another model, referred to as horizontal modeling, examined a linear path from the microlevel negotiation to the macrolevel societal variables. The path consists of two mediating variables, negotiation outcomes in terms of distributive justice principles and stable agreements. It takes the following form:

$$ \mathrm{PJ}\to \mathrm{DJ}\to \mathrm{stable}\ \mathrm{agreements}\to \mathrm{DP} $$

This path was shown to be statistically significant.Footnote 1 Thus, adherence to the PJ principles of fair treatment and transparency set in motion the path that leads eventually to a durable peace. The justice-peace relationship is mediated by fair outcomes that are implemented with few breaches or violations.

Another finding from this study has implications for practice. The strongest impact of the justice variables was on the aspect of reconciliation referred to as community activities. Collaborative activities among former combatants during the implementation period and sustained through the years to follow were a key to transitional justice and preventing other outbreaks of violence. Indeed, this element was more important than such other features of reconciliation as trials and truth commissions, reception and reintegration, and provisions for refugees. This finding was reported also by the case-study authors in the special issue of the Journal of Conflict Resolution edited by Balcells and Justino (2014) on civil wars and violence prevention. It suggests that practitioners focus their efforts on encouraging joint activities that lead to continued cooperation. The key here is working together to achieve a shared goal such as restoring buildings destroyed during the war. But the activities would also contribute to another kind of “building,” referred to as building a culture of peace (Bar-Tel 2008). One by one, new relationships forged by cooperative projects could ramify into networks of cooperation which, taken together, would result in changed norms about justice with implications for changes in security and governance institutions. These are issues on the frontier of conflict resolution research and practice. (See Liebovitch et al. (2020) for perspectives on the dynamics of building social networks.)

Search for Mechanisms: Trust and Problem Solving

The results obtained in these studies of peace agreements support findings from other studies on the role of justice in negotiation. Those studies also suggest other variables that may operate with justice considerations in influencing the durability of agreements and the peace to follow. One of these variables is trust. Another is problem-solving behavior. In their simulation experiment, Hollander-Blumoff and Tyler (2008) showed that PJ principles correlated with both trustworthiness and willingness to disclose information. More PJ principles led to more acceptable agreements which were, in turn, more durable. In their comparative study of settlements to end violent conflicts, Irmer and Druckman (2009) found that comprehensive agreements depended on the development of trust through phases of the talks: specifically, movement from an early phase of mistrust through calculus-based and knowledge-based trust, culminating in identity trust in the later phases of the negotiations. Analyses of historical cases of negotiation involving the United States showed that PJ, problem-solving behaviors, and integrative agreements formed a correlated cluster: adherence to PJ principles (vs. a lack of adherence to these principles) in the process was strongly associated with problem-solving (vs. competitive bargaining) which, in turn, increased the chances of integrative outcomes which were durable (Wagner and Druckman 2012).

These findings suggest possible mechanisms that explain why justice principles, notably the equality principle, result in more durable agreements. Fair, durable agreements would seem to emerge from processes in which disputing parties build trust. This is more likely to occur when the process is guided by PJ principles and a problem-solving orientation. Less clear is the causal sequence of these variables: Does trust emerge from adhering to PJ principles and/or problem solving? Or, is trust a precondition for PJ and problem solving? These questions remain to be explored. Answers to them would also provide guidance for strategies used by third parties. For example, if trust is an emergent process, then focusing efforts first on establishing PJ rules would be advised. If, however, trust is a precondition, then an initial focus on creating conditions for increased perceptions of trust would be beneficial. But, if the trigger is problem solving, then encouraging these behaviors should lead to increased trust. It may be that this cluster of variables is intertwined or cyclical rather than sequential as discussed earlier. In this case, bolstering any one of them would have ramifying effects on the others. These are interesting challenges to be met in further research.

Another challenge is to sustain the trust developed during the negotiation process. Four interventions during the period of transitional justice may help to strengthen the relationships between former adversaries. One is to encourage adherence to fair treatment and transparency principles during the postnegotiation relationship building process. Another is to develop a problem-solving norm that guides interactions and decision-making. A third is to design and implement community activities that depend on cooperation from both sides. This is referred to as functional interdependence. It has been shown to increase cooperation without threatening group identities (e.g., Deschamps and Brown 1983). A fourth intervention is in the realm of truth and reconciliation processes. The key is to acknowledge (and apologize for) past atrocities while preventing escalation. Understanding how backward (restorative justice) and forward (more just institutions) looking dialogues are balanced remains a research challenge.

In addition to these interventions, we need to understand the role played by spoilers – parties with a stake in disrupting or nullifying agreements – in the postnegotiation period. Further work on spoiler typologies would be useful (see Stedman 2000). But it would also be helpful to probe situations where spoilers have served to unite the former combatants in a joint effort to undermine their activities in the interest of sustaining the new agreement. All of these efforts contribute to transitioning from a culture of violence to peace. The next challenge consists of sustaining the new culture in light of disappointed expectations that may arise as evidenced by the current developments in Mozambique and Nagorno-Karabakh.

Conclusion and Policy Implications

The results obtained from our early experiments are clear. Peace agreements that emphasize the principle of equality in their provisions are more stable than those that do not. This is particularly the case when the equality provisions are forward looking, by which we mean equal treatment for all parties or equal shares in terms of the distribution of power. When, however, the equality provisions are backward looking – concerned primarily with military strength – the agreement is likely to be less stable or no more stable than agreements that emphasize other justice principles. These findings suggest that negotiators and interveners should be guided by policies that stress the importance of seeking agreements containing provisions of equal treatment and/or shares. Agreements without these provisions may not last. Knowing this, we addressed the question of how to obtain these types of group agreements.

Lessons for strategy are suggested by close examination of selected cases from our data set. Agreements that proved to be durable provide advice about what to do; those that unraveled send a message about what not to do. It appears that the specific tactics used by third parties may be less important than their objective. Both forceful and facilitating approaches were effective in producing forward-looking agreements that lasted. Likewise, both approaches were ineffective in producing lasting agreements when equality principles were not included in the agreements. These observations are consistent with the well-known idea of a firm-but-flexible strategy: Pruitt and his colleagues demonstrated in a number of experiments that the best agreements occurred when negotiators or mediators were firm on objectives (or principles) but flexible on the means for achieving those objectives (e.g., Pruitt and Lewis 1977). The implication of this finding for policy is clear: Encourage third parties to actively promote the objective of forward-looking equality while giving them latitude on the tactics they use to accomplish this objective.

The more recent research examined a larger number of peace agreements cases and added indicators of durable societal peace. Of particular interest is the finding that shows PJ serves as a starting mechanism in the sense that it initiates a path that proceeds to agreements, implementation, and peace. This result has policy implications as well. Focusing on the key elements of PJ, transparency and fair treatment, we suggest that negotiation systems be designed to accommodate these features of the process. For example, a designer or third party may suggest a prenegotiation on the rules of the game. Although not binding, these rules would encourage circulating all documents and discourage private meetings. They would also spell out the meaning of fair play for all the delegations. With agreement on the procedures, the delegates can focus their attention on debating the substantive issues (see the chapter by Paur and Carreras). The idea is to build trust through early agreement on procedures.

The issues in peace negotiations are daunting. They include such aspects of reconciliation as provisions for refugees and community relations. They also include institutional reforms in the areas of security and governance as well as stabilization and growth of the ravaged economies. One challenge is how and in what order to discuss these issues. Another is finding a way to build agreement packages by trading concessions along the lines of multi-attribute value analysis (see Raiffa 1982; Ulvila 1990). These analytical approaches, well known in the group decision and negotiation community, are more likely to succeed when parties agree to adhere to the PJ principles of fair treatment and transparency and strive for an outcome that incorporates the DJ principle of equality.