![Sample our Politics & International Relations journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5947/TOC-Subject-Sample-2021-Politics-International-Relations.jpg"})
ABSTRACT
Do UN peacekeeping forces protect civilians from harm in post-war environments? Current evidence suggests that the answer to this question is yes. But extant research mostly examines this relationship at the country-level and consequently has logical difficulty tracing decreases in civilian fatalities to actual peacekeeper activities. We would have more confidence in the ability of peacekeepers to limit harm and protect non-combatants if the reduction in violence occurred locally where blue helmets were positioned. Using original geocoded data of yearly UN deployments in four Sub-Saharan African conflicts (Sudan, South Sudan, Democratic Republic of Congo, and Ivory Coast), we find that peacekeeping units get locally deployed to violent post-war areas and they reduce the level of civilian harm almost immediately. But, in areas without violent clashes between government forces and rebels, we find peacekeeping units more responsive to civilian targeting by rebels, which indicates a reluctance among peacekeepers to confront government forces that target civilians. While host nation consent is crucial for the success of a peacekeeping mission, the findings from this study caution against nurturing illiberal regimes by failing to check government atrocities. The failure to confront government abuse can jeopardize long-term peace and reconciliation.
Disclosure Statement
No potential conflict of interest was reported by the authors.
Notes on contributors
Anup Phayal is Assistant Professor of Public and International Affairs at the University of North Carolina Wilmington.
Brandon Prins is Professor of Political Science at the University of Tennessee-Knoxville and a Global Security Fellow with the Howard Baker Center for Public Policy.
Notes
2 Gilligan and Sergenti, “Do UN Interventions Cause Peace?”
3 Beardsley and Gleditsch, “Peacekeeping as Conflict Containment.”
4 Hultman, Kathman, and Shannon, “United Nations Peacekeeping.”
5 Fortna, “Does Peacekeeping Keep Peace”; Phayal, “UN Troop Deployment”; Hultman, “UN Peace Operations.”
6 Hultman, Kathman, and Shannon, “United Nations Peacekeeping.”
7 We also, later in the manuscript, compare and contrast our theoretical argument and empirical results to a recent study published in International Organization by Fjelde, Hultman, and Nilsson, “Protection through Presence”, which is similar in several ways to our investigation.
8 Ruggeri, Dorussen, and Gizelis, “Winning the Peace Locally.”
9 Hultman, Kathman, and Shannon, “United Nations Peacekeeping.”
10 Costalli, “Does Peacekeeping Work.”
11 Hultman, Kathman, and Shannon, “United Nations Peacekeeping,” also examine peacekeeping effectiveness during civil war, similar to Costalli, Op. cit.
12 Walter, Committing to Peace, 26.
13 Gilligan and Sergenti, Op. cit.
14 Beardsley, “The UN at the Peacemaking.”
15 Allansson, Melander, and Themnér, “Organized Violence, 1989–2016.”
16 Diehl and Druckman, Evaluating Peace Operations.
17 Ruggeri, Dorussen, and Gizelis, “Winning the Peace Locally”; Beardsley and Gleditsch, Op. cit.; Costalli, Op. cit.
18 The recent publication by Fjelde et al., Op. cit., remains an exception.
19 Gilligan and Stedman, “Where Do the Peacekeepers Go,” 44.
20 Fortna, “Does Peacekeeping Keep Peace?”
21 Costalli, Op. cit.
22 Ruggeri, Dorussen, and Gizelis, “Winning the Peace Locally.”
23 Jett, Why Peacekeeping Fails, 50.
24 See https://www.theguardian.com/world/2015/sep/17/un-united-nations-peacekeepers-rwanda-bosnia. Also, see Duursma, “Obstruction and Intimidation of Peacekeepers.”
25 Of the 14 ongoing peacekeeping missions (as of July 2019), eight have the POC mandate. [Available: https://peacekeeping.un.org/en/where-we-operate, Accessed July 18, 2019].
26 For instance, following an event on July 2016, when UN Peacekeepers in Juba, South Sudan, were unable to protect civilians from government forces and other armed groups, there has been widespread criticism against the UN: https://www.theguardian.com/world/2016/aug/15/south-sudan-aid-worker-rape-attack-united-nations-un.
27 Macfarlane, Thielking and Weiss, “The Responsibility to Protect,” 988.
28 See http://world.time.com/2012/11/26/defining-peacekeeping-downward-the-u-n-debacle-in-eastern-congo/. When the rebel group M-23 captured Goma in Democratic Republic of Congo, angry locals protested against the UN and government inaction. It was reported that some locals even burned UN property in Kisangani (see https://www.bbc.com/news/world-africa-20405739).
29 See Letter from the Group of Experts on the DRC to the Chair of the Security Council Committee, S/2012/843.
30 The FIB consisted of military units from three countries from Southern African Development Community: South Africa, Tanzania and Malawi. Under the direct operational command of the MONUSCO Force Commander, its role was mainly to carry out offensive actions on its own or jointly with FARDC (See the Secretary General’s Special Report, S/2013/119).
33 Small Arms Survey Report, 2012 Available http://www.smallarmssurveysudan.org/fileadmin/docs/issue-briefs/HSBA-IB21-Inter-tribal_violence_in_Jonglei.pdf [Accessed March 11, 2018].
34 As argued by Costalli, Op. cit., 377, the best peacekeeping approach would be to prevent violence from occurring in the first place. But due to their smaller sizes, deploying peacekeeping units often becomes a reactionary measure.
35 de Carvalho and Lie “Chronicle of a Frustration Foretold” indicate that the Security Council leaves the decision to the field leadership to determine whether or not they have the capability to protect the civilians.
36 Nasu, “Operationalizing the Responsibility.”
37 de Carvalho and Lie, Op. cit., 344.
38 Holt, Taylor, and Kelly “Protecting Civillians.”
39 While reform efforts are ongoing, challenges persist. See Gorur and Sharland Prioritizing the Protection for a more recent report on the topic.
40 Nasu, Op. cit., 368.
41 See recent Security Council President’s statement S/PRST/2015/23 on civilian protection.
42 Holt, Taylor, and Kelly, Op. cit., 217.
43 Sebastian and Gorur, “U.N. Peacekeeping.”
44 See “Sudan expels two UN officials, 2014” Available: https://www.aljazeera.com/news/africa/2014/12/sudan-expels-two-un-officials-20141225192742271467.html [Accessed January 10, 2018].
45 Kalyvas, The Logic of Violence in Civil War.
46 Walter, Committing to Peace.
47 Ruggeri, Gizelis, and Dorussen, “Managing Mistrust.”
48 Fortna, Does Peacekeeping Work, 84, 86.
49 Fortna, “Does Peacekeeping Keep Peace”; Phayal, Op. cit.
50 Macfarlane, Thielking and Weiss, Op. cit.
51 There are grid-cells that lie in the border of South Sudan and DRC, and, South Sudan and Darfur, but without any incidents or UN deployments only in South Sudan. But in order to avoid double counting we removed 90 border grid-cells from DRC and 78 border grid-cells from Darfur, keeping only South Sudan grid-cells.
52 For e.g. see, Hultman, Kathman, and Shannon, “Beyond Keeping Peace”; Ruggeri, Dorussen, and Gizelis, “On the Frontline Every Day.”
53 There are generally three platoons in a company, therefore deployment of a platoon (around 30–50 personnel) is counted as .33 companies.
54 Raleigh et al., “Introducing ACLED.”
55 Fatalities due to clashes or the number of peacekeepers killed by belligerents are excluded.
56 Haussmann tests on our dataset suggest that a fixed effects model is superior to a pooled or a random effects model.
57 Angrist and Pischke, Mostly Harmless.
58 We use an ordinary least square fixed effects model for our first analyses. As a robustness check, we estimate using Poisson models with count of platoons as the DV in the supplementary information and the results are not different from the main models.
59 Allison and Waterman, “Fixed-effects”; Cameron and Trivedi, Regression Analysis, 53: 341–57; Wooldridge, Econometric Analysis, 755–8; Wooldridge, “Distribution-free Estimation.”
60 Guimaraes “The Fixed Effects” confirms Allison and Waterman’s, Ibid., assertion that a fixed effects negative binomial is “not a true fixed effects.”
61 Provided in supplementary information.
62 Changing this threshold to 10 (159 grid-cells) or 50 (75 grid-cells) does not change the results substantively.
63 Moreover, such a static measure gets rejected by our fixed effects model.
64 Bharti et al., “Remotely Measuring Populations during a Crisis by Overlaying Two Data Sources.”
65 Tollefsen, Strand, and Buhaug, “PRIO-GRID.”
66 A recent study by Fjelde et al., Op. cit., published in International Organization similarly focuses on the effect peacekeepers have on civilian victimization. It is reassuring that the findings of this study are similar to our own and our research certainly complements theirs. Still, there are several key differences. They find that the presence of peacekeepers is more likely to reduce one-sided violence against civilians by rebels, but not violence perpetrated by government forces. While our findings are similar, there are three key differences in our theoretical argument and methodological approach. First, we take into consideration armed clashes between government forces and rebels and how such renewed conflict between former rivals can affect peacekeeping deployments and civilian targeting. Since UN military peacekeepers are expected to act in response to high profile clashes, we expect that deployment is a function of civilian deaths conditioned by the occurrence of such armed clashes. Once military peacekeepers are deployed, violence against civilians peaks during renewed conflict between rebels and government forces. It is therefore imperative to consider the conditional effect of clashes while hypothesizing how deployments might affect civilian targeting. Second, Fjelde et al. use a pooled logit model with monthly level data to estimate the effect of deployment on civilian killings. One weakness of this approach is that it may not take into account time-invariant unobserved heterogeneity Wooldridge, Introductory Econometrics, 460. For our data, Hausmann tests reveal that a fixed effects model is more efficient than either pooled or random effects models. Finally, one of the key factors that tends to bias the results is grid-cell population, since both military deployments and civilian killings are likely to be in more populated areas. The grid-cell population data that the Fjelde et al. use remain temporally static. Instead, we use nightlight emissions (which vary over time) as a proxy to control for the population density of a grid-cell.
67 OIOS Report, “Inspection of the Performance,” 29.
68 See supplementary document for results with increased threshold for clash.
69 Some caution is required in how we interpret these findings. The OIOS report Op. cit., 22, notes that in some missions the primary responder to violence against civilians is the civilian component of the mission rather than the military component. We leave it for future research to determine whether that strategy is effective in combatting violence against civilians.
70 Note that the count model excludes observations that do not change over time. As a robustness check, see supplementary information for results from (1) negative binomial model that uses the grid-cell dummy as fixed effect, (2) negative binomial model with lag dependent variable, and (3) using ppmlhdfe-Fast Poisson Estimation with High-dimensional Fixed Effects Correia, Guimarães, and Zylkin, “PPMLHDFE.”
71 Similar to Fjelde et al., Op. cit.
72 is generated with margins in STATA, using the nu0 option that visualizes the effect when fixed effects is 0.
73 Hultman, Kathman, and Shannon, “Beyond Keeping Peace”; Phayal, Op. cit.
74 von Billerbeck and Tansey, “Enabling Autocracy.”
75 Variable, distance to nearest deployment, is not included in , but including it does not change the main results.