CONCEPTUAL DEVELOPMENTS

Relatively new fields of study must first identify and define the core concepts around which research will center. In the case of war, the conceptual components and operational measures of that term are now relatively well accepted, but that was not the case in earlier decades when international conflict research was in its nascent stage.

Definitions and taxonomies

Peacekeeping analyses are notorious for their conceptual muddles. It is common for the terms peacekeeping, peace-building, peace enforcement, peacemaking, and a host of other terms to be used interchangeably. Indeed, the United Nations website (www. un.org/Depts/dpko/dpko/index.asp) labels a broad set of operations over its history as ‘peacekeeping,’ implying similar attributes. Even when distinctions are made, there is not necessarily agreement among scholars and practitioners on the conceptual com­ponents of a given term. Much of the problem is attributable to the metamorphosis of peace operations since the late 1980s. Prior to that, peacekeeping (now what is commonly referred to as traditional or cold war peacekeeping) consisted of lightly armed troops whose primary purpose was to separate combatants, following a cease-fire, but usually prior to final conflict resolution. Such operations were based on the so-called ‘holy trinity’: host state consent, impartiality, and minimum use of force (Bellamy and Williams, 2005a).

How does one sort out the different terms and components that make up the broad category of peace operations? Auseful place to begin is in the definitions put forward by the then UN Secretary-General Boutros Boutros-Ghali (1995), the standard or baseline conceptualization of peacekeeping- related definitions used by scholars and pol­icymakers. Beyond a traditional conception of peacekeeping, Boutros Ghali distinguishes it from two other forms of UN intervention: preventive diplomacy and peacemaking. Preventive diplomacy is ‘action to prevent dis­putes from arising between parties,' whereas peacemaking is ‘action to bring hostile parties to agreement.' The use of peace forces is only one possible strategy for these two approaches, and in practice, peace soldiers have rarely (e.g. preventive UN forces in Macedonia) been used in these fashions. More notable is the difference between peacekeeping and peace-building.

Boutros-Ghali speaks of ‘peace-building' as the ‘creation of a new environment,' not merely the cessation of hostilities facilitated by traditional peacekeeping. Although by no means a clear consensus in the literature, there are a series of characteristics or dimensions by which peace-building can be compared with other concepts, most notably traditional peacekeeping.

The first dimension concerns the goal(s) of peace-building.

The second dimension of peace-building includes the strategies and accompanying activities designed to achieve the goal(s). Not surprisingly, these vary somewhat accord­ing to whether one pursues goals broader than preventing conflict recurrence or not. A minimalist strategy of preventing con­flict recurrence adopts strategies consis­tent with conflict management. Thus, some peace-building activities include disarming warring parties, destroying weapons, and training indigenous security personnel. Facil­itating elections, repatriating refugees, and strengthening government institutions are peace-building activities consistent with this minimalist strategy (see Boutros-Ghali, 1995 for these and other examples). A broader conception of peace-building leads to some­what different strategies and sets of activities. Some (Cockell, 2000; Doyle and Sambanis, 2000; Cousens, 2001) see peace-building as addressing the ‘root causes of conflict.' Minimalists expect conflict to occur, but desire to manage it peacefully. In contrast, the maximalist strategy does not merely promote management, but conflict resolution as well (for a broad list of activities, see Cockell, 2000; Ramsbottom, 2000; Newman and Schnabel, 2002).

A third dimension concerns the timing of such activities. Most conceptions of peace­building envision its activities to occur following some type of peace settlement between warring parties. This is in contrast to other peace operations. Preventive diplomacy and its accompanying actions are supposed to be put in place before significant levels of violence occur.

A fourth dimension is the context in which peace-building should be carried out. Boutros-Ghali (1995) envisions that peace­building could occur following either inter­state or intrastate conflict; he also notes some activities appropriate to each context. De facto, however, most of the discussion of peace-building has assumed that it would be employed in a civil context, following an intrastate war, significant ethnic conflict, or even in a failed state (see Maley, Sampford, and Thakur, 2003). In practice, we should recognize that the distinctions between intrastate and interstate conflict break down when neighboring states intervene in civil conflicts, best illustrated by the Congo war starting in the 1990s.

The fifth and final dimension is the actors that will carry out the peace-building actions. As Pugh (2000) notes, peace-building seems to assume that external actors will play a significant, if not exclusive, role in this enter­prise. Again, an examination of the strategy and activities would not seem to preclude local actors, and indeed some elements (e.g. truth and reconciliation commissions) may be more successful when external actors are not the driving forces. Also implicit in the peace­building notion is that such actors will act in an impartial fashion for the greater good of the society, exercising some moral author­ity rather than pursuing private interests. Normatively, most regard peace-building as an altruistic enterprise, but as Pugh (2000) argues, such conceptions may still promote particular ideologies (e.g. democracy, neo­capitalism, and the like - see also Paris, 2004 for an elaboration).

Even accepting these conceptualizations does not mean that each type of peace opera­tion is completely different or that even those in the same category are homogenous. Thus, a number of scholars have devised taxonomies that go beyond conceptual distinctions to indi­cate variation on other dimensions. Among the first and most notable was that put forward by James (1990). Using a broad conception of peacekeeping and reviewing all operations prior to 1990, he classified them in four categories according to the relationship of the conflict to surrounding states. ‘Backyard problems' were those that took place within the sphere of a major power. ‘Clubhouse troubles' occur when a group of states organize an operation to deal with an ‘in-group' problem. ‘Neighborhood quarrels' are those conflicts that do not fall into the first two categories and largely remain localized, but do not reach the level of ‘dangerous crossroads,' which are those conflicts most prone to escalation and the greater involve­ment of external parties. Talentino (2004) divided peace-building operations into three categories: limited, extensive, and nation­building. She seems to make such distinctions based on the operation's enforcement powers, scope of reform activities, timing of deploy­ment, and degree of control over the local government machinery.

Actual operations may include missions that fall into multiple categories and missions sometimes mutate over their lifetimes. Thus, Diehl, Druckman, and Wall (1998) sought to measure the degrees of difference between missions. Unlike previous work, their tax­onomy was tied to a broader theoretical framework derived from the scholarly liter­ature on conflict management and resolution. They examined 16 different peace missions across 12 different dimensions suggested by those literatures and peacekeeping experts. The net result was an identification of four distinct clusters, scaled using traditional peacekeeping as a reference point: passive monitoring, damage limitation, restoring civil society, and coercive missions respectively.

The studies above were explicit attempts to conceptualize peacekeeping operations and their variants. Yet, a large set of studies implicitly engaged in classification exercises as they attempted to delineate the evolution of peace operations over time. Various scholars (Wiseman, 1987; James, 1990; Goulding, 1993; Segal, 1995; Malone and Wermester, 2000; Schmidl, 2000; Jakobsen, 2002; Bellamy, Williams, and Griffin, 2004; and Talentino, 2004) have divided peace­keeping history into different periods or epochs, largely according to the kinds of missions performed and thereby implicitly make conceptual distinctions between those missions. These treatments differ in a number of ways, but there is some general agree­ment. First, most treat the creation of the first traditional peacekeeping operation as sui generis, a significant break with past practice. There is also a notable break around 1988 or 1989, when peacekeeping operations increase dramatically in number and take on new missions. The terms ‘new' (Ratner, 1995) or ‘second-generation' (Mackinlay and Chopra, 1992) peacekeeping are among the labels used to describe these operations; in fact, their commonality is more that they differ from traditional missions rather than that they share substantial attributes among themselves.

Success and failure

As noted below, a plurality of peacekeeping studies focus on the conditions for success in those missions and attempt to draw lessons for future operations. From a research standpoint then, it is perhaps surprising that relatively little attention is given to defining what is meant by success. The abundance of attention is given to the independent variables in studies, and considerably less (if any at all) is given to the dependent variable. Most often, scholars delve into the reasons an operation succeeded or failed without detailing how that judgment was reached. Sometimes, the operational criteria are implicit in the discussions of the alleged influences, but explicit a priori conceptualiza­tions of success and operational indicators are relatively rare.

Determining what constitutes success or failure is a prerequisite for building knowl­edge about the factors associated with those conditions. Yet, there is considerable dis­agreement among even those scholars that have broached the subject matter (for a range of views, see Druckman and Stern, 1997). Beyond normative preferences, there are several concerns that make it difficult to find a consensus. First, when conceptualizing peace­keeping success, the question arises: success for whom?Although rarely addressed directly, there are several sets of stakeholders in peace operations, each of which might generate different ideas on success: the international community, the main protagonist states or groups, the local population, and the states contributing soldiers (Durch in Druckman and Stern, 1997). Although each stakeholder may share some common interests (e.g. limiting violence), their interests are not completely coterminous. For example, a contributing state may have as one of its goals limiting casualties to its soldiers. Succeeding in that goal, however, may necessitate actions that undermine the international community’s goal of human rights protection for the threatened population. Thus, there needs to be a clearer specification of whose success is assessed and recognition that different factors may affect success in multiple ways across stakeholders.

Second, defining success will vary according to whether one adopts a short versus long-term perspective (Weiss, 1994; Bellamy and Williams, 2005b). Success may be conceptualized as achievement of goals that occur during the course of a peace operation or in some time frame immediately following the withdrawal of the peacekeeping force. An example of the former is alleviation of starvation and improvement of medical conditions during a humanitarian operation; an example of the latter is the absence of violent conflict for several years following the operation (e.g. Enterline and Kang, 2003; Doyle and Sambanis, 2006). These are certainly valid conceptions and ones most amenable to the needs of policymakers. Yet, we know that a longer-term perspective, however, often leads to a different assessment of an operation’s success or failure. For example, various peacekeeping efforts in East Timor were almost universally considered a success in the immediate aftermath, only to prompt a reassessment when violence and instability returned in 2006. As with different stakeholders, there may be significant differences in the predictor and outcome variables for short- and long-term success. With respect to the latter, two problems arise. An initial problem is specifying how long a window should be considered in assessing peacekeeping outcomes. Given path dependency and other effects, peacekeeping may have consequences that extend for decades. Yet, extraordinarily long time frames make it impossible to assess ongoing and recently concluded operations (Bellamy and Williams, 2005b). Furthermore, the longer the time period that passes between the end of the operation and the assessment, the more difficult it will be to draw causal conclusions about the impact of the operation; intervening forces are likely to have as great or greater impact as the peace operations on future conditions.

A third consideration is developing a base­line against which to assess peacekeeping's effects. Some suggest that peacekeeping be compared to a situation in which no action was taken by the international community (e.g. Durch in Druckman and Stern, 1997). Related to this standard is one in which the conditions prior to deployment are compared to those during and following the operation (e.g. Kaysen and Rathjens, 1995; Heemskerk and Weller, 2002). This standard has the advantage of ‘normalizing' the baseline, as moderate levels of violence during peacekeeping may be considered progress in some contexts (e.g. deployment during full-scale civil war), but backsliding in others (e.g. deployment following a cease-fire). Yet, decision makers' choices are rarely between just peacekeeping and inaction (Diehl, 1994). Some scholars suggest that analysts consider opportunity costs imposed by the choice of peacekeeping (Ratner in Druckman and Stern, 1997). Problems with this standard, however, are that it requires an adequate specification of alternative policies and then an accurate counter-factual analysis of what would have happened if other alternatives had been selected. Needless to say, neither of these steps is straightforward and the validity of such efforts would be difficult to determine. Still others suggest comparing effectiveness across peacekeeping operations (Ratner, 1995; Stiles and McDonald, 1992, refer to this as a trend­based assessment), but this generates only comparative assessments of what may be dissimilar operations and provides no absolute assessment baseline.

Not surprisingly, most analysts advocate using guidelines provided in the opera­tion's mandate, the authorizing document (e.g. Security Council resolution) provided by the organization carrying out the mission (e.g. Durch, 1993; Ratner, 1995; Bratt, 1996; O'Neill and Rees, 2005; Bellamy and Williams, 2005b). Mandates often contain specific tasks to be completed or benchmarks that should be reached. In this way, the com­parison is made to a given standard specific to the mission; presumably the mandate was created with reference to extant conditions, but not necessarily. There may be standards created related to election supervision, troop withdrawal and disarmament, the creation of civil society institutions, and human rights among many others. Particular conceptions and indicators are far too numerous to list here. In one sense, this is appropriate as a particular mission is only judged according to the task with which it was assigned. On the other hand, there are a number of drawbacks associated with using mandates to define success. First, the mandates given operations are the products of political deliberation and compromise, and the result is that they are frequently vague. There is much room for debate on the scope and detail of the operation's mission; this alone makes it difficult to assess whether the designs of the mandate have been achieved (Diehl in Druckman and Stern, 1997). Second, mandates may be inflexible in the face of changing conflict conditions, and thus what peacekeepers are attempting to do may no longer reflect the standards present in the mandate (Bellamy and Williams, 2005b). Third, ‘mandate clarity' is associated with peacekeeping success, making the whole use of mandates in evaluations endogenous (Diehl in Druckman and Stern, 1997).

Regardless of mission or the phase of conflict in which the peacekeeping operation is deployed, there are some generic standards for success applicable to all missions. Perhaps the most common one in the literature is conflict abatement (e.g. Diehl, 1994; Bratt, 1996; Welch, 2000; Fortna, 2004a, 2004b; Mullenbach, 2006; Pushkina, 2006). All peacekeeping operations are supposed to discourage violent conflict (e.g. renewed war, organized crime), a prerequisite for any other mission that might be performed. Somewhat less common, although still prominent, is con­flict containment; this involves preventing the conflict from expanding to include additional internal actors, neighboring states, or major powers (Weinberger, 1983;Allan, 1996; Bratt, 1996; Jett, 2000; Ofuatey-Kodjoe, 2002; O'Neill and Rees, 2005; Pushkina, 2006). Finally, many analysts assign peacekeepers the role of creating an environment suitable for conflict resolution (Diehl, 1994; Allan, 1996; O'Neill and Rees, 2005). Yet, this would seem applicable only to those operations deployed prior to a peace agreement.

Even as the conceptual development of peacekeeping success has improved, oper­ational indicators for those concepts have lagged behind. There are few data sets devoted exclusively to peace operations, and therefore to the extent that scholars use operational indicators, they are borrowed from conflict studies in general. Thus, it is not surprising that a popular measure of conflict abatement is the ‘duration of peace,' the number of months or years from the time of peacekeeping deployment to the onset of renewed war or violence (e.g. Fortna, 2004a, 2004b; Mullenbach, 2006)

Theoretical frameworks

Peacekeeping research has historically been atheoretical or at best has involved what has been called ‘problem-solving' theory (Bellamy, 2004), the practical concern with developing strategies for conflict management and resolution. Much of this was a conscious choice as authors focused on improving peacekeeping performance. It was also the case that peacekeeping did not centrally fit into the Cold War or traditional military strategies, the centerpieces of most theorizing in international relations (Pugh, 2003). Yet, even as these efforts were attempts to build generalizations, these were largely empirical generalizations with little or no attempt to provide broad theoretical explanations. Diehl (1994) was perhaps the first to identify potential variables that influence peacekeep­ing success and then systematically test propositions across multiple cases. Although some theoretical logic underlay each of those factors, there was not an integrated model or theoretical argument tying these disparate elements together.

For over a decade, several scholars (Fetherston, 1994a, 2000; Paris, 2000; Pugh, 2003) have lamented the lack of theory in peacekeeping analyses, but progress has been remarkably slow. Some theoretical efforts aspired to create a ‘theory of peacekeeping,' attempts at mid-level theory. Fetherston (1994a; 1994b) adopted a con­tingency framework, derived from studies of conflict management, and based on the assumption that conflicts have subjective and objective elements. The central point is that different techniques of conflict man­agement, including peacekeeping, can be utilized to deal with different objective and subjective aspects of conflict. Success in this framework is designated as the achievement of ‘positive' peace, along a number of security and social dimensions. Ultimately, this provides a normative framework upon which to analyze peacekeeping rather than a theory of peacekeeping per se. Some­what similarly, Diehl, Druckman, and Wall (1998) adopt a framework from the con­flict management literature, the distributive- integrative bargaining distinction developed by Walton and McKersie (1965). Yet, this again is more a heuristic, for taxonomical purposes in their analyses, rather than a well articulated theoretical argument. Tethering peacekeeping studies to those in conflict management has the advantage of being able to draw upon an extensive and well-developed body of research and ideas, although no scholar has yet brought this application to fruition.

A number of other works have pro­vided scholars with suggested theoretical approaches, pointing the way to theoretical development, but stopping short of actual theory. Because peacekeeping research has always had a strong normative element, it is not surprising that suggested avenues for peacekeeping theory have emphasized this aspect. Critical theory approaches (Bellamy, 2004; Pugh, 2004) reject an objectivist view of global politics and peacekeeping (see Bellamy, 2004 for a comparison), and adopt the normative position that peace operations should promote human security and emancipation. Critical theory analysts often focus on how peace operations may be designed and executed to support the interests of the hegemon or leading states in the international system (Fetherston, 2000; Pugh, 2004; see also Gibbs, 1997, for a discussion), an effect that is to be resisted according to these theorists. Similarly, post­modern approaches such as that by Debrix (1999) also emphasize the subjectivist aspects of peacekeeping that are said to serve an unjust world order. Critical theory and related approaches have the advantages of raising fundamental questions about peacekeeping that would largely be ignored outside of this framework. Most peacekeeping analyses are myopic, focused only on the immediate and local effects of the operations. A critical theory framework forces scholars to inquire whose interests are served by the missions and what effects those operations have on the prevailing global order. One drawback, however, is that the normative orientation is not necessarily compatible with validating a theoretical argument through value-free social-science testing.

Much of critical theory responds to con­tention that peacekeeping is a component of the liberal world order. Liberalization (i.e. open markets, democracy) is supposed to be the remedy for armed conflict, a normative proposition as much or more than a theoretical one. Paris (2004) studies peace-building in the aftermath of civil conflicts from the liberal perspective, and assesses the extent to which peace-building serves those interests. Once again, this is a framework that leads the researcher to pose certain questions, rather than necessarily being a model or theory that provides answers to those questions.

Some of the most promising theoretical works on peacekeeping have not been those seeking to construct original theory. Instead, the modal approach has been to use extant theory to inform us about one element of peacekeeping operations, rather than as an attempt to build a theory of peacekeeping or to provide a comprehensive explanation for peacekeeping. In this vein, there have been a number of different theoretical applications to peacekeeping, reflecting the range of theoretical approaches in the international relations discipline.

Realist theory has traditionally considered actions of international organizations as epiphenomena, largely reflective of major states' interests. Using a realist lense on peacekeeping, Neack (1995) explores whether states contribute troops to peacekeep­ing operations for self-interested (consistent with realist motives) or more altruistic reasons. Consistent with state foreign policy behavior as the unit of analysis, Sandler and his colleagues (Khanna, Sandler, and Shimizu, 1998; Shimizu and Sandler, 2002) have analyzed peacekeeping contributions from a public goods perspective. The authors argue that peacekeeping is a public good, and the theory of those goods can give us insights into the willingness of states to contribute to those efforts, as well as their propensity to ‘free ride' on others.

Rational choice models have also been applied to peacekeeping (Smith and Stam, 2003). Such models often view war as an information problem, and the imposition of peacekeeping troops may interrupt the flow of information during warfare that would permit the warring parties to reach a settlement. Yet once a settlement is reached, peacekeepers may provide the credible commitment to guarantee implementation, allowing parties to overcome one of the key barriers to peace in the long run identified by rational choice theorists.

Peacekeeping has also drawn the attention of liberalism theorists, most notably those concerned with the so-called democratic peace. Lebovic (2004) looks at whether the level of democracy of a country makes it more likely to contribute troops to UN peacekeeping operations; he reports a strong positive relationship between democracy and peacekeeping contributions in post-Cold War operations.

Finally, constructivist approaches to peace­keeping include a broad range of issues, including identifying different kinds of discourse involved in the authorization and conduct of peacekeeping operations (Neack and Knudson, 1999). Most promi­nent, however, have been general theoretical approaches to international organizations in which peacekeeping operations are an important application. For example, Barnett and Finnemore (2004) use a sociological variant of constructivism to posit that the UN bureaucracy can exercise significant autonomy (quite in contrast to realist concep­tions) on international organization actions, including peacekeeping.