There is a renewed scholarly interest in peacekeeping with quantitative, systematic empirical studies figuring prominently. Arguably, this recent work has gone some way to address two puzzles that have consistently surrounded debates on peacekeeping. The first puzzle is that reporting on peacekeeping tends to be critical (if not outright hostile), and public opinion about intervention in far-away places – typical candidates for peacekeeping – is at best uninterested or even negative when costs become more apparent. Regardless, peacekeeping has become an important element of efforts by the international community to resolve conflict.
For example, recently UN peacekeepers have been deployed to the Central African Republic, (South) Sudan and Mali. Regional organizations like the African Union also regularly deploy peacekeepers. In Europe’s most recent and most serious conflict in the Ukraine, most parties, including Russia, want the OSCE to mediate and observe. Political leaders also regularly use peacekeeping to describe their military interventions; for example, Russia labeled its operations in Georgia and the Crimean as such, while the USA and NATO defined ISF in Afghanistan as a peacekeeping operation.
Peacekeeping has not only become more prominent, but its nature has also changed: nowadays peacekeepers tend to intervene as much in civil as in interstate wars. They often intervene early even before the situation has become resolved militarily or politically making it necessary to enforce rather than to keep the peace. Peacekeepers operate with a broader mandate; so called “new” or “integrated” peacekeeping monitors and observes, but also aims to protect civilians, to build capacity and to address the underlying factors that contributed to conflict (Ratner 1996). These changes may have not always been helpful from a practical perspective because they set unrealistic ambitions and open the door for widespread, unfocused popular skepticism as well as for abuse of the term peacekeepers by political leaders.
I argue that a further insight from recent scholarly work is that such broad and general understanding also hampers the analysis of the effectiveness of peacekeeping. In contrast, carefully designed and detailed empirical studies have helped to address to some extent a second puzzle of peacekeeping; namely, the contrast between quantitative, comparative studies and case studies in their assessment of the effectiveness of peacekeeping.
2 The impact of peacekeeping
Possibly because peacekeeping for a long time was a relatively minor undertaking that changed little over time, most analyses tended to rely on (comparative) case studies (Diehl, Reifschneider, and Hensel 1996; Durch et al. 2003; Paris 1997, 2004). The focus was on the legal framework of peacekeeping, in particular the difficulties in giving peacekeepers mandates that were appropriate and internationally acceptable, and on the management of peacekeeping operations, the “lessons learned.” As a consequence, analyses focused on failures: what peacekeepers had been unable to do because of the limited mandates or lack of capacity.
In this light, the findings of initial systematic and quantitative comparative studies were quite remarkable. To start with Doyle and Sambanis (2000, 2006), these studies almost invariably find a positive impact of peacekeeping. Moreover, Doyle and Sambanis (2006, 5) link the ability of peacekeepers to sustain peace to the broader agenda of new peacekeeping: “while the UN is very poor at “war,” imposing a settlement by force, it can be very good at “peace,” mediating and implementing a comprehensively negotiated peace.” Heldt (2004) finds only minor differences between the effectiveness of UN and non-UN peacekeeping. 1 Fortna (2003, 2004a,b, 2008), Sambanis and Doyle (2007), Gilligan and Sergenti (2008), Hegre, Hultman, and Nygård (2011) and Sambanis and Doyle (2007) have shown that the impact of peacekeeping is not simply a matter of selection bias: if anything, the UN selects “hard” cases – civil conflicts with high casualty levels that have been on-going and incompletely settled – making the positive record of peacekeeping even more remarkable. Fortna (2008) outlines why peacekeeping could matter from the perspective of (civil) war as bargaining failure: peacekeepers reduce uncertainty by providing information and increasing the credibility of any commitments made by warring parties; for example, by means of dealing with potential spoilers.
This rather impressive line of research has, however, done little to take away doubts about peacekeeping. Undeniably the findings of quantitative research contrast sharply with the conclusions reached by more qualitative, ethno-graphic research. The list of possible critiques of peacekeeping is long and fundamental. Below I summarize three main areas of concern.
Firstly, peacekeeping operations are often seen as dysfunctional. Many researchers observe a gap in expectations between what is hoped for in New York and what is actually delivered on the ground. There are serious concerns about the quality of peacekeepers provided. The limited willingness of countries that have sent peacekeepers to accept casualties compounds the lack of interest by major powers to sustain peacekeeping missions. Communication with New York is also often imperfect; it is slow and regularly fails to deliver the support needed on the ground. Howard (2008) notices that although the UN is capable of learning within missions, few lessons are learned across missions. Bellamy, Williams, and Griffin (2010) observe that the more recent emphasis on regional peacekeeping runs the risk of a divergence between low (mainly African) and high (Western) quality of peacekeeping.
Secondly, peacekeepers often have a very limited understanding of local conditions and (unsurprisingly given the need of a small number of peacekeepers to control a large area) limited presence on the ground. A further complaint is the frequent rotation of peacekeepers. Pouligny (2006) highlights the big difference between the official version of peacekeeping and local sentiments. Similarly, in Timor Leste I found the official version that peacekeepers build capacity of local policemen to differ markedly from the local version in which peacekeepers are good “taxi-drivers” (they could take you anywhere) and good at fixing computers. 2 Autesserre (2009, 2010) describes how in the Democratic Republic of the Congo, the international community, including the UN, tends to ignore local conflict dynamics and insists upon building peace from the top-down rather than from the bottom-up.
Thirdly, new peacekeeping is criticized as being part of the liberal governance agenda. Weinstein (2005) and Herbst (2003) express fundamental doubt about the feasibility of external intervention to succeed at peace- and state-building, and draw attention to the need of endogeneously supported processes. In an extremely entertaining reflection on his work in Timor Leste, Gordon Peake describes this as the failure of the “Tropical Bakery School of International Capacity Building”:
“Giving voice to a frustration never hinted at in the official cables he sent back to New York, he [a UN official, HD] threw up his arms in frustration and said, “If we (the United Nations) can’t be successful here, in this tiny place with not much crime and relatively decent political leadership, where can we be successful?”” (Peake 2013, 192).
Roland Paris (1997, 2004) has also extensively discussed the limits of the liberal democratic peace for post-conflict countries with historically weak states, where the true legacy of peace-building is often little more than providing quasi-authoritarian leaders to hold on to their position via quasi-democratic elections allowing them to retain the spoils of office.
3 Building an academic consensus of sorts
The situation may not be quite as bleak, and most recent research on peacekeeping is going some way to lessen the sharp contrast between the findings of quantitative and qualitative research. 3 This is obviously important since arguing on the basis of correlations that peacekeeping works will always remain unconvincing as long as instances where it doesn’t work seem to be so well documented.
3.1 Disaggregating peacekeeping operations
Whereas originally quantitative research used the country or mission as the main unit of analysis, there are now a number of studies that look at peacekeeping events (Ruggeri, Gizelis, and Dorussen 2011). The disaggregation of peacekeeping missions has led to a reassessment of a number of key questions: where are peacekeepers actually deployed, what are they doing and whose behavior is affected by these actions? These data correct some popular misperceptions; for example, peacekeepers are not spending their time “on the beach” but are actually deployed to conflict areas. At the same time, there are reasons for concern: deployment can be slow and the number of peacekeepers deployed locally is often small (Ruggeri, Dorussen, and Gizelis 2012).
Disaggregating peacekeeping also creates more precise measures of impact. For example, Gleditsch and Beardsley (2014) show that peacekeeping missions contain and fix conflict zones sub-nationally. Hultmann, Kathman, and Shannon (2013) show that peacekeepers protect civilians. In my own research with Ismene Gizelis and Andrea Ruggeri, I found the ability of peacekeepers to cooperate with local authorities to vary. Peacekeepers find it easier to cooperate with central (government) authorities compared to rebels. They also find it easier to engage in capacity building, and report more hostility when they take over full responsibility. Cooperation is also easier when peacekeepers operate from a position of strength and are perceived by the rebels as “protectors” (Dorussen and Gizelis 2013; Ruggeri, Dorussen, and Gizelis 2013b).
In some ways, this amends the findings of earlier quantitative research. There are some reasons to worry about possible selection bias sub-nationally. The UN does not always select the hardest route of action. It takes time to deploy to conflict areas and lack of infrastructure further hampers effective deployment. Peacekeepers also seem to engage more in interactions and activities where they perceive more opportunities for cooperation. Figure 1 shows that the majority of peacekeeping events in Africa involve central authorities with which relations are often cordial, while only about a third of events involve rebel groups where relations are often much more strained. Of course, disaggregate data also make it possible to correct for selection bias, and the good news is that peacekeeping still seems to matter. It may not necessarily deter conflict, but it seems to be able to contain it and making it last shorter (Ruggeri, Dorussen, and Gizelis 2013a).
3.2 Systematic and experimental field research
The practice of case studies and fieldwork has also changed markedly. Gordon Peake notes that the “Tropical Bakery” was also where researchers could be found: “The restaurant also served as a base camp for the many people who seemed to be conducting research about Timor Leste, albeit in probably one of the most culturally unrepresentative places in the country” (Peake 2013, 191). In contrast, Kate Roll carefully selected a random sample and travelled across the Timor Leste on motorbike in her research on veteran payments in Timor Leste. Moreover, she assessed the validity of local lists of veterans confirming suspicion that not all “veterans” had actually participated in the fight against Indonesian occupation (Roll 2014).
Particularly relevant here are the efforts to improve research design to address concerns of peacekeeping and the “peacekept” directly and to go beyond considering post-conflict situations more generally. Mvukiyehe and Samii (2012) report on randomized surveys of the experiences of ex-combatants and civilians in Sierra Leone, Liberia and Burundi. They consider where peacekeepers were deployed and focus on specific peacekeeping policies, such as demobilization and reintegration. Gilligan, Pasquale, and Samii (2014) have done so for Nepal with a focus on the rebels who were “cantonized” as part of the peace process. Gilligan et al. (2012) conducted randomized and anonymous interviews of the “peacekept” in Monrovia (Liberia); in this case, to assess any financial and sexual relationships of women with peacekeepers.
Concerns that local population may be hesitant to truthfully report on peacekeepers also motivate my own fieldwork in Timor Leste. Here, I use list experiments to study the perception of security situation and the role of the UN peacekeepers following their exit from Timor Leste in December 2012. List experiments are a way to create true anonymity, which hopefully leads respondents to respond unbiased rather than to provide socially or politically “correct” answers. In several respects, there are indeed some indications of reporting bias. With true anonymity, respondents occasionally reveal a more accurate understanding: for example, they acknowledge that the influence of UNPOL in border areas had declined considerably for some time. Another pattern is that, under anonymity, vulnerable groups are more willing to reveal their concerns about the exit of the UN. Women are more likely to report police abuse. Respondents with more trust in local justice systems as well as respondents who experienced more violence are less happy with the exit of the UN (Dorussen 2014).
The findings of systematic and experimental fieldwork show that peacekeeping can have some positive effects, but they also clearly echo many of the concerns of previous fieldwork. For example, Mvukiyehe and Samii (2012) report a limited impact of peacekeepers on the perception of security simply because often peacekeepers arrived only after the security situation had already improved locally. Gilligan, Pasquale, and Samii (2014) report notable disenchantment among demobilized rebels in Nepal. The findings on intimate relationships between women and peacekeepers in Monrovia are possibly even more disconcerting.
4 What next?
What could be fruitful avenues for future research? I highlight three “lessons learned” and will link them to the practice of peacekeeping as well as avenues for future research. First of all, there is a need to establish a proper baseline: what is the peace that peacekeepers are supposed to keep, and who are the peacekeepers? Secondly, assessing the impact to peacekeeping is ultimately akin to a form of policy evaluation: what policies fall under the rubric of peacekeeping, and how can we be certain that peacekeeping contributed to the observed outcomes? Thirdly, whose peace is being kept: how, if at all, does it matter that peacekeepers (and aid workers and researchers) tend to be foreigners, and what is the perspective of the locals on peacekeeping? 4
Establishing a proper baseline of expectations for peacekeeping is an on-going concern. The early decision of Doyle and Sambanis (2000, 2006) to focus on peace after five or ten years may seem arbitrary, but the same can be said for my own decision to look at conflict and cooperation during peacekeeping operations (Dorussen and Gizelis 2013). Limiting the assessment of peacekeeping to “negative” peace, regardless of whether this is the reduction of one-sided or conflict violence, is minimalistic. At the same time, requiring peacekeepers to establish a “positive” peace is unrealistic and akin to setting up a strawman. Patrick Burgess, the director of Human Rights unit in UNTAET/UNMISET and governor of Oecussie under UNTAET, expressed weariness about the high standards Western researchers often set for peacekeeping in Timor Leste. 5
The definition of who we consider to be peacekeepers is a related issue. Autesserre targets much of her criticism to the wider UN/NGO community, and is actually rather positive about the accomplishments of the “blue helmets.” Gordon Peake spares hardly anybody in his criticism of foreign meddling in Timor Leste, but clearly throws his hardest punches against UN and other foreign officials engaged in capacity building and development programs. Even though the emphasis on “good governance,” comprehensive peace and integrated peacekeeping operations makes good political sense, it may be less useful to identify the specific impact peacekeepers can have in a post-conflict situation. A more narrow definition of peacekeeping and peacekeepers helps to identify their specific, albeit possible limited, contribution.
Establishing that peacekeeping works is fundamentally policy evaluation, and the key insight of experimental research is that research design matters. First of all, there is an obvious risk of endogeneity and random assignment of the “treatment” (that is, peacekeeping) would be ideal. Secondly, it has to be clear what the exact treatment is: a simple “dummy” variable to measure the presence of peacekeepers is unlikely to capture essential information. Both points are obviously related: only if we consider specific peacekeeping policies is it possible to think about their application as randomly assigned.
This is an area where disaggregating peacekeeping and field experiments are most complimentary. So far, research on peacekeeping events focuses on establishing a link between peacekeeping and peace (or lack of conflict), but tells us less about why it would work: to test the theoretical arguments on why peacekeeping should work requires a clear understanding of what peacekeepers actually do or are at least perceived to be doing. More practically, like we may question whether every “peacekeeper” deserves the title, we also need to question whether everything that peacekeepers do should necessarily be called peacekeeping.
Finally, still relatively little is known about what the “locals” make of all of this. It seems plausible that peace needs to be built from the bottom as much, if not more, as that it needs to be made at the top (Autesserre 2010; Dorussen and Gizelis 2013). However, it remains an open question and, at a minimum, requires an assessment of the exposure and response of the local population to peacekeeping. The Asian foundation has done a series of high-quality surveys in Timor Leste about (perception of) security but did not include questions on peacekeepers. There have been calls for surveys to be held before, during (and after?) peacekeeping, which would definitely be a step forward.
The complexity of the relationship between the local population and foreigners (malae in Tetun) is hard to overestimate. Clearly, locals cannot be bothered with nice analytical distinctions between military and civil peacekeepers, peacekeepers and UN officials, official and non-official (NGO) capacity builders and aid-workers, or even researchers…they are all malae. And we share common features: we are present for a relatively short time, we are much better off – even when we are “roughing it,” we have a nice house and comfortable job to go back home to. Also, we tend to have only limited knowledge of the local language, and even if we make a commitment to learn it, in most cases we still stick out like a sore thumb. The use of malae reflects these ambiguities very nicely: it is both used as friendly greeting and attempt to engage in conversation, but also as outright abuse.
There is a widespread expectation that, as malae, we are bringing something. This is, of course, fair enough for peacekeepers who should deliver peace or aid-workers who should promote development. But what about us as researchers? I was warned repeatedly that locals were rather fed-up with being the objects of research. For them, a survey was supposedly either a test where answers had to be correct, or a tool that should deliver some immediate benefits. Personally I had no negative experiences and a refusal rate of about 15% is much better than one should expect in the UK. At the same time, it is fair enough to demand that research on peacekeeping brings not only benefits to us in the form of well-placed publications but also contributes to the practice of peacekeeping.
Autesserre, Séverine (2010), The Trouble with the Congo; Local Violence and the Failure of International Peacebuilding, Cambridge University Press, Cambridge.Google Scholar
Bellamy, Alex J., Paul D. Williams, Stuart Griffin (2010), Understanding Peacekeeping. 2nd Edition. Polity Press, Cambridge.Google Scholar
Dorussen, Han (2014), The Security Perception in Timor Leste Following the Departure of UN Peacekeepers. East-Asia Security Workshop, Kobe 8–9, 2014.Google Scholar
Doyle, Michael W., Nicholas Sambanis (2000), International Peace-Building: A Theoretical and Quantitative Analysis, American Political Science Review, vol. 94, no. 4, pp. 779–801.CrossrefGoogle Scholar
Doyle, Michael W., Nicholas Sambanis (2006), Making War and Building Peace. United Nations Peace Operations, Princeton University Press, Princeton.Google Scholar
Durch, William J., Victoria K. Holt, Caroline R. Earle, Moira K. Shanahan (2003), The Brahimi Report and the Future of UN Peace Operations, The Henry J. Stimson Center, Washington DC.Google Scholar
Fortna, Virginia P. (2004b), Does Peacekeeping Keep Peace? International Intervention and the Duration of Peace After Civil War, International Studies Quarterly, vol. 48, no. 2, pp. 269–292.CrossrefGoogle Scholar
Fortna, Virginia P. (2008), Does Peacekeeping Work? Shaping Belligerents’ Choices after Civil War, Princeton University Press, Princeton.Google Scholar
Gilligan, Michael J., Bernd Beber, Jenny Guardado, Sabrina Karim (2012), UN Peacekeeping and Transactional Sex in Monrovia, Liberia, Paper presented at the Folke Bernadotte Working Group on Peacekeeping, New York.Google Scholar
Gilligan, Michael J., Benjamin Pasquale, Cyrus Samii (2014), Civil War and Social Cohesion: Lab-in-the-Field Evidence from Nepal, American Journal of Political Science, vol. 58, no. 3, pp. 604–619.CrossrefGoogle Scholar
Gilligan, Michael J., Ernest J. Sergenti (2008), Do UN Interventions Cause Peace? Using Matching to Improve Causal Inference, Quarterly Journal of Political Science, vol. 3, no. 2, pp. 89–122.Web of ScienceGoogle Scholar
Gleditsch, Kristian, Kyle Beardsley (2014), Peacekeeping as Conflict Containment. Working Paper, University of Essex.Google Scholar
Hegre, Håvard, Lisa Hultman, Håvard Mokleiv Nygård (2011), Simulating the Effect of Peacekeeping Operations 2010–2035, in Sun-Ki Chai, John Salernoand Patricia L. Mabry, (eds.), Advances in Social Computing. Lecture Notes in Computer Science, Springer, Heidelberg, pp. 325–332.Google Scholar
Heldt, Birger (2004), UN-led or non-UN-led Peacekeeping Operations?, IRI Review, vol. 9, pp. 113–138.Google Scholar
Herbst, Jeffrey (2003), Let Them Fail: State Failure in Theory and Practice: Implications for Policy, in Robert I. Rotberg, (ed.), When States Fail: Causes and Consequences, Princeton University Press, Princeton, NJ.Google Scholar
Howard, Lise M. (2008), UN Peacekeeping in Civil Wars, Cambridge University Press, Cambridge.Google Scholar
Hultman, Lisa, Jacob Kathman, Megan Shannon (2013), United Nations Peacekeeping and Civilian Protection in Civil War. American Journal of Political Science, vol. 57, no. 4, pp. 875–891.Web of ScienceGoogle Scholar
Mvukiyehe, Eric, Cyrus Samii (2012), Does Peacekeeping Work from the Bottom Up? Negative Evidence from Liberia, New York University, Working Paper.Google Scholar
Paris, Roland (2004), At War’s End: Building Peace after Civil Conflict, Cambridge University Press, Cambridge.Google Scholar
Peake, Gordon (2013), Beloved Land. Stories, Struggles, and Secrets from Timor-Leste, Scribe Publications, Brunswick, Victoria.Google Scholar
Pouligny, Béatrice (2006), Peace Operations Seen from Below: UN Missions and Local People, Kumerian, London.Google Scholar
Ratner, Steven R. (1996), The New UN Peacekeeping. Building Peace in Lands of Conflict after the Cold War, St Martin’s, New York.Google Scholar
Roll, Kate (2014), Encountering Resistance: Qualitative Insights from the Quantitative Sampling of Ex-Combatants in Timor-Leste, PS: Political Science & Politics, vol. 47, pp. 485–489. doi:10.1017/S1049096514000420.Web of ScienceCrossrefGoogle Scholar
Ruggeri, Andrea, Han Dorussen, Theodora-Ismene Gizelis (2012), On the Frontline Every Day? Subnational Deployment of United Nations Peacekeepers, Paper presented at the Meeting of the International Studies Association, San Diego, April 2012.Google Scholar
Ruggeri, Andrea, Han Dorussen, Theodora-Ismene Gizelis (2013a), Winning the Peace Locally: UN Peacekeeping and Local Conflict, Paper presented at the Meeting of the International Studies Association, San Francisco, April 2013.Google Scholar
Ruggeri, Andrea, Han Dorussen, Theodora-Ismene Gizelis (2013b), Managing Mistrust: an Analysis of Cooperation with UN Peacekeeping in Africa, Journal of Conflict Resolution, vol. 57, no. 3, pp. 387–409.CrossrefWeb of ScienceGoogle Scholar
Ruggeri, Andrea, Ismene Gizelis, Han Dorussen (2011), Event data as Bismarck’s sausages? Intercoder Reliability, Coder Selection and Data Quality, International Interactions, vol. 37, no. 3, pp. 340–361.CrossrefWeb of ScienceGoogle Scholar
Sambanis, Nicholas, Johan Schulhofer-Wohl (2007), Evaluating Multilateral Interventions in Civil Wars: A Comparison of UN and Non-UN peace Operations, in D. Bourantonis, K. Ifantis, P. Tsakonis (eds.), Multilateralism and Security Institutions in an Era of Globalization, Routledge, London, pp. 252–287.Google Scholar
Weinstein, Jeremy (2005), Autonomous Recovery and International Intervention in Comparative Perspective, Working Paper 57, Center for Global Development, Washington, DC.Google Scholar
Based on keynote address delivered at the Annual Meeting of the Network of European Peace Scientists, The Hague, June 24 2014.