http://orcid.org/0000-0003-1261-3982
![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
This article explores the potential and limitations of epidemiological analyses of violence. We draw on an 18-month sample of Joint Mission Analysis Centres data to identify clusters of armed violence in Darfur and model the risk of armed clashes in space and time. We illustrate the merit of using methods from both descriptive epidemiology and analytical epidemiology to study armed conflict. We observe three interesting correlations. Firstly, that violence in one locality means it is more likely that there will be violence in a neighbouring locality in the next month. Secondly, that the presence of peacekeepers in a locality where violence has occurred means it is less likely that violence will occur in a neighbouring locality, than if peacekeepers were not present. Finally, our third observation is that the presence of peacekeepers in a given locality means it is more likely that violence will occur in that locality. Understanding how conflict occurs in space and time could contribute to the effectiveness of peacekeeping missions. This touches upon the major commonality between the efforts of peacekeeping missions and epidemiology: both are fundamentally concerned with the well-being of defined populations and both rely on data to design effective interventions.
Disclosure statement
No potential conflict of interest was reported by the authors.
About the authors
Allard Duursma completed his PhD in International Relations at the University of Oxford in 2015, focusing on international mediation efforts in civil wars in Africa. Upon completion of his PhD, Allard joined the HCRI as a research associate at the Making Peacekeeping Data Work Project. Allard's role within this project is to examine the incident data from the UN and African Union mission in Darfur through using GIS and conducting spatial analyses.
Róisín Read joined HCRI in 2014 as part of the Making Peacekeeping Data Work for the International Community project and is now a Lecturer in Peace and Conflict Studies. Her research sits at the intersection of peace and conflict and humanitarian studies and focuses on the politics of knowledge and representation. She is interested in exploring how critical approaches to knowledge might help us to better understand international interventions in conflict and post-conflict contexts, with a geographical focus on Sudan and South Sudan.
ORCID
Allard Duursma http://orcid.org/0000-0003-1261-3982
Róisín Read http://orcid.org/0000-0003-1782-3209
Notes
1 For some overviews of this trend, see: Gleditsch, Metternich, and Ruggeri, “Data and Progress in Peace and Conflict Research”; Clayton, “Quantitative and Econometric Methodologies”; Raleigh, Witmer and O’Loughlin, “Review and Assessment of Spatial Analysis.”
2 Taback and Coupland, The Science of Human Security, 7.
3 Pope, Resisting the Evidence.
4 UN DPKO, Policy Directive, POL/2006/3000/04, 1.
5 Shetler-Jones, “Intelligence in Integrated UN Peacekeeping,” 517.
6 Ramjoué, “Improving UN Intelligence,” 469.
7 These data have been provided by the African Union High-Level Panel on Darfur.
8 See Duursma in this volume: Duursma, “Counting Deaths While Keeping Peace.”
9 Clayton, “Quantitative and Econometric Methodologies.”
10 Wille, “The Six ‘Ws’ of Security,” 7.
11 de Waal et al., “Epidemiology of Lethal Violence.”
12 Roberts, “The Science of Human Security,” 18.
13 Gleditsch, Metternich, and Ruggeri, “Data and Progress in Peace and Conflict Research.”
15 Barnes and Wilson, “Big Data, Social Physics,” 3.
16 Toft and Zhukov, “Islamists and Nationalists”; Zhukov, “Roads and the Diffusion of Insurgent Violence; Weidmann and Ward, “Predicting Conflict in Space and Time”; Beardsley and Gleditsch, “Peacekeeping as Conflict Containment.”
17 Brody et al., Map-making and Myth-making.
18 Zwi and Ugalde, “Towards and Epidemiology of Political Violence.”
19 Ibid., 641.
20 Depoortere et al., “Violence and Mortality in West Darfur”; Degomme and Guha-Sapir, “Patterns of Mortality Rates.”
21 Taback and Coupland, “The Science of Human Security,” 4.
22 Murray et al., “Armed Conflict as a Public Health Problem,” 346.
23 Zwi and Ugalde, “Towards and Epidemiology of Political Violence,” 641.
24 For more on the Correlates of War project see http://www.correlatesofwar.org/.
25 Beer, “The Epidemiology of Peace and War.”
26 Ibid., 45.
27 Ibid., 46.
28 Kim, “Global Violence and a Just World Order,” 181.
29 Houweling and Siccama, “The Epidemiology of War, 1816–1980.”
30 For example, de Waal et al., “Epidemiology of Lethal Violence”; Burnham et al., “Mortality After 2003.”
31 de Waal et al., “Epidemiology of Lethal Violence,” 369.
32 Taback and Coupland, “The Science of Human Security.”
33 de Waal et al., “Epidemiology of Lethal Violence,”, 368.
34 Clayton, “Quantitative and Econometric Methodologies.”
35 Gleditsch and Ward, “War and Peace in Space and Time.”
36 Beardsley, “Peacekeeping and the Contagion of Armed Conflict.”
37 Gleditsch, Metternich, and Ruggeri, “Data and Progress in Peace and Conflict Research.”
38 Beardsley, Gleditsch, and Lo, “Roving Bandits?”
39 Beer, “The Epidemiology of War and Peace,” 59.
40 See, Ripley, Spatial Statistics.
41 A more formal way of assessing the interrelationship of points is to use the second-order method developed by Ripley. This method is a useful approach to describe patterns of points of which the incidence might well have been influenced by other points nearby. In essence, Ripley's second-order analysis of point patterns determines whether a given distribution of points is unusual by comparing it to a situation in which points are distributed in complete spatial randomness. We have examined spatial clustering of armed clashes in Darfur using the second-order method developed by Ripley. The results of this analysis are in line with what one can simply observe from the map in . Accordingly, the pattern of armed violence in Darfur exhibits spatial clustering at relatively lower distance scales, while it is exhibits dispersion at relatively greater distances scales. See, Ripley, Spatial Statistics.
42 Brody et al., Map-making and Myth-making.
43 Koch and Denike, Crediting His Critics’ Concerns, 1246.
44 Brody et al., Map-making and Myth-making in Broad Street, 64.
45 McLeod, Our Sense of Snow, 926.
46 de Waal et al., “The Epidemiology of Lethal Violence in Darfur,” 368–9.
47 Kalyvas, “The Ontology of “Political Violence’,” 484.
48 Ibid., 483.
49 Collier and Hoeffler, “Greed and Grievance in Civil War”; Fearon and Laitin, “Ethnicity, Insurgency, and Civil War.” Flint and de Waal also observe this in Darfur. See, Flint and de Waal, Darfur.
50 Beck and Tucker, “Taking Time Seriously.”
51 Available at: https://data.humdata.org/dataset/sudan-admin-level-2-boundaries.
52 Choi et al., “Evaluation of Bayesian Spatial-Temporal Latent Models.”.
53 Huang et al., “An Integrated Bayesian Model for Estimating the Long-Term Health Effects.”
54 For two other studies that also use a simple logit model with lagged variables that model spatial dependence, see, Buhaug and Gleditsch, “Contagion or Confusion?”; Beardsley, “Peacekeeping and the Contagion of Armed Conflict.”
55 The number of peacekeeping bases and the locations of these bases are coded on the basis of several deployment maps published in this period. We thank Fjelde et al. for kindly sharing their data, based on these deployment maps, with us. See, Fjelde et al., “Protection Through Presence.”
56 Walter, “Does Conflict Beget Conflict?”
57 Forsberg, Neighbors at Risk.
58 Flint and de Waal, Darfur.
59 Zhukov, “Roads and the Diffusion of Insurgent Violence.”
60 Beer, “The Epidemiology of Peace and War,” 60.
61 Ibid., 61.
62 Weidman and Ward, “Predicting Conflict in Space and Time.”
63 To put it in more technical terms, we would need a model that uses autoregressive priors to capture the spatial autocorrelation inherent in the JMAC conflict data.
64 Ruggeri et al., “Winning the Peace Locally.”
65 Ruggeri et al., “On the Frontline Every Day?”
66 Ruggeri et al., “Winning the Peace Locally.” Another option would be to use an instrumental variable approach. However, this would require the availability of a variable that influences the deployment of peacekeepers in Darfur, but not the effectiveness of the peacekeepers. It would be very difficult to find a variable that meets this requirement for peacekeeping in a single country.
67 This finding is in line with a finding by Beardsley and Gleditsch that the deployment of peacekeepers makes armed violence less dispersed. See, Beardsley and Gleditsch, “Peacekeeping as Conflict Containment.”
68 See, Weidmann and Ward, “Predicting Conflict in Space and Time”; Toft and Zhukov, “Islamists and Nationalists”; Buhaug and Gleditsch, “Contagion or Confusion?”